Help us keep the list up to date and submit new video software here.
Complete Version history / Release notes / Changelog / What's New for Topaz Video Studio
Topaz Video 1.6.0
Hi Everyone,
We have some more exciting updates this week! And to our AMD and Mac friends, this is a good one!
Improvements and Fixes
Features
Precise 2.5 update for AMD & Mac!
SDR-HDR Model: Hyperion 2 for AMD & Mac!
Topaz Plugin: Nuke
Additional Updates
Crop X & Y Update
Bug Fixes
Timeline issues with Image Sequences
Project loading fixes
and more!
Precise 2.5 update for AMD & Mac!
image
image
1927×1084 439 KB
image
image
1795×1183 392 KB
Starlight Precise 2.5 has made one heck of a splash! And like the title suggests, AMD and Mac are now supported for this model! And yes, this model available for everyone that meets the requirements to run it!
Model Recap
Starlight Precise 2.5 is the 5th variant to the desktop Starlight Model Family. Note, that previous iterations of this model exists on Astra. This model has been specialized for realistic human face and skin textures in addition to other areas of improvement such as material, texture, and text.
This model has become so popular, that you can now find Starlight Precise 2.5 model with it’s own icon for faster model access.
image
image
520×238 18 KB
About
This model works for Medium to High Quality content whether this is generative, live action or restoration videos.
Improves generated content that may have soft, plastic, or slight artificial looks, however the AI generated content should have solid structure and composition for ideal improvements.
There is a slight difference between the local version and the cloud version in order to fit on the desktop.
Greatly reduces flickering with smaller textures.
Recommended to use video inputs at or under 4K.
Outputs up to 4K.
Model access via the cloud for all subscription models.
Cloud Export is supported on both Mac and Windows.
Expect long render times for this diffusion model. Appears to have faster processing compared to Mini and HQ.
Please note that Starlight Precise 2.5 model had to be slightly modified to fit the MAC hardware requirements. Mac users may experience slower renders, however outputs have maintained results that are reasonably similar to Nvidia with great quality results.
This Model Is Less Ideal For
Shots with heavy motion blur may be rendered as texture.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Not ideal for content below 480p
image
image
2048×739 348 KB
Technical Things To Know
Now available for AMD, Mac & Nvidia
Available for all licenses to run locally! Yes, you read this right!
Nvidia Requirements:
Requires an absolute minimum of 12GB VRAM with 32GB CPU RAM, though 16GB VRAM is recommended.
Targeted for GPU compute compatibility of 8.9. This is also the recommendation.
Please note, though there might be some success with a CC of 8.6 cards, this is not
This model is included locally, using neuroserver.
AMD Requirements:
Requires an absolute minimum of 12GB VRAM with 32GB RAM, though 16GB VRAM is recommended.
Mac Requirements:
Requires an absolute minimum of 16GB though 24GB is recommended.
All Apple Silicon should be supported.
image
image
1869×1447 352 KB
SDR-HDR Model: Hyperion 2 for AMD & Mac!
image
image
1671×1020 377 KB
Hyperion 2 is finally available for AMD and Mac for local rendering! Please note the individual requirements needed pending your machine’s hardware below.
Model Recap
Hyperion 2, is a great update to our previous iteration. This version of the model expands your color and nit value range even further! This is fantastic for wider spectrum manipulation for VFX Compositing and for modern viewing on HDR supportive displays!
image
image
2868×1615 555 KB
About
Converts 8-bit GenAI or archival video into HDR, high bit-depth media suitable for professional finishing pipelines.
Expands tonal range and color precision to provide significantly greater grading latitude in highlights, shadows, and midtones.
Prevents highlight and shadow clamping often seen in standard range and native GenAI video sources.
Expands existing color opportunities/steps for professional compositing and color manipulation workflows.
The additional color opportunities greatly reduces visible banding and compression artifacts common in 8 bit renders.
Has great details with text and textures even during high motion blur areas.
Export ProRes video or EXR (Pro) image sequences.
Support ACES color output with EXR (Pro) renders for better integration into industry-standard film and television pipelines.
This works best for medium to high quality content.
This model has some built in enhancements to revitalize and improve content.
This version is an easy “one click” solution. There are no parameters to adjust.
Max input and output resolution for this model is 3840x3840.
image
image
1920×1432 1.58 MB
Technical things to know
Nvidia Requirements:
Requires 16GB VRAM with 32GB CPU RAM, though 24GB VRAM is recommended.
Targeted for GPU compute compatibility of 8.9. This is also the recommendation.
Please note, though there might be some success with a CC of 8.6 cards, this is not technically supported and some may encounter a less ideal experience.
AMD Requirements:
Requires an absolute minimum of 12GB VRAM with 32GB CPU RAM, though 24GB VRAM is recommended.
Mac Requirements:
Requires an absolute minimum of 16GB RAM though 24GB is recommended.
All Apple Silicon should be supported.
image
image
2464×1831 679 KB
Topaz Plugin: Nuke!
The very long awaited plugin is finally here! The Topaz plugin for Nuke is now a reality! As a Compositor or VFX artists, one is now able to take advantage of some of Topaz’ famous models right within the the Nuke working environment.
Upon installation, the OFX file will be available to be downloaded. If there are any issues with the initial installation. The plugin is available within the Topaz Video application in the upper left side under the “Plugin” drop down.
Within Nuke, there are two places to access the Topaz plugin.
image
image
799×799 24.7 KB
Via the left panel of icons, select the Topaz plugin icon to reveal the four plugin options.
image
image
1054×418 64.5 KB
Via the “Tab”, search for
Enhance [Topaz Labs]
Frame Interpolation [Topaz Labs]
Interlaced Enhance [Topaz Labs]
Motion Deblur [Topaz Labs]
Or with the “Tab” search, search instead for “[topaz” to reveal all the Topaz plugins.
image
image
1084×304 40.8 KB
Have fun experimenting with the different model options with your footage!
Note: Having previous Alpha, Beta, or Release versions of the Topaz OFX plugin alongside the current release can lead to crashes in Nuke. Please ensure that only the latest Topaz Video bundle is installed in your OFX directory, which can be found at C:\Program Files\Common Files\OFX\Plugins on Windows or /Library/OFX/Plugins on MacOS.
Additional Updates to Note!
Crop X & Y
The X & Y transforms are now back for Crop. This gives one the ability to input specific placement coordinates for the upper left corner of the crop shape. This provides accurate pixel values that you may need for your crop.
image
image
3099×1898 698 KB
Bug Fixes
Fixed image sequence in/out ranges appearing incorrectly on the timeline when updating fps.
Fixed project load failing when sources are moved or deleted.
Topaz Video 1.5.0
Hi Everyone,
We have another really cool update this week! This includes a new model, a panel and some bug fixes.
Improvements and Fixes
Features
New SDR-HDR Model: Hyperion 2!
New Topaz Panel for Premiere!
Additional Updates
Starlight Precise 2.5 Updates
Bug Fixes
Fixed Starlight Precise 2.5 last frame drop
Addressed Neuroserver image sequence failures
Refined Starlight Precise cloud resolution issues
Fixed broken feedback link
and more!
New SDR-HDR Model: Hyperion 2!!
image
image
1920×1083 983 KB
Hyperion 2, is a great update to our previous iteration. This version of the model now expands your color and nit value range even further! This is fantastic for wider spectrum manipulation for VFX Compositing and for modern viewing on HDR supportive displays!
About
Converts 8-bit GenAI or archival video into HDR, high bit-depth media suitable for professional finishing pipelines.
Expands tonal range and color precision to provide significantly greater grading latitude in highlights, shadows, and midtones.
Prevents highlight and shadow clamping often seen in standard range and native GenAI video sources.
Expands existing color opportunities/steps for professional compositing and color manipulation workflows.
The additional color opportunities greatly reduces visible banding and compression artifacts common in 8 bit renders.
Has great details with text and textures even during high motion blur areas.
Export ProRes video or EXR (Pro) image sequences.
Support ACES color output with EXR (Pro) renders for better integration into industry-standard film and television pipelines.
This works best for medium to high quality content.
This model has some built in enhancements to revitalize and improve content.
This version is an easy “one click” solution. There are no parameters to adjust.
Max input and output resolution for this model is 3840x3840.
image
image
1920×1432 1.58 MB
Technical things to know
Nvidia Requirements:
Currently Nvidia Only
Requires 16GB VRAM though 24GB VRAM is recommended.
Targeted for GPU compute compatibility of 8.9. This is also the recommendation.
Please note, though there might be some success with a CC of 8.6 cards, this is not technically supported and some may encounter a less ideal experience.
Colorspace Outputs
ProRes Primaries: BT.2020 PQ
ProRes Transfer Function: BT.709
EXR (Pro only) Image Sequence: Aces 2065-1
Things to Consider
SDR to HDR does not operate well with interlaced footage. If you do have interlaced footage, it’s recommended that you deinterlace first before applying this tool.
Though this model has some enhancement benefits additional enhancements with this SDR to HDR Model can be a great combo to reduce issues from source such as compression and artifacts for better results.
Extreme cases of video with very muddy/dark imagery or very bright/blown out content, may produce limited results with the limited source range to build upon.
An HDR display will be needed in order to view the results properly. Otherwise results will look washed out.
Known Issues
Cloud Export is not yet supported on both Mac and Windows. Coming soon!
Limited codec options for local rendering. Looking to expand the current options!
New Topaz Panel for Premiere!
We have built a UXP panel for Adobe Premiere users! Topaz Labs for Premiere is new utility that introduces a “send to cloud” workflow to batch upscale video clips and images directly from the Premiere.
Simply drag video or image clips from your bin or timeline into the panel, choose an AI enhancement model, and upscale to HD, 4K, or even 8K. Better yet, this process does not take up system resources, allowing you to continue editing. Rendered files are automatically imported to your project and placed in timeline.
Topaz Video 1.4.0
Hi Everyone,
We have a short but extra sweet release!
The team’s been busy and working hard to get this latest update for you!
I’m pleased to announce our most powerful model release in a while! Starlight Precise 2.5 for local desktop rendering!
Improvements and Fixes
Features
Starlight Fast Precise 2.5 model
Neuroserver
Starlight Precise 2.5 Local
image
image
1920×1076 1.39 MB
image
image
1869×1447 473 KB
Starlight Precise 2.5 made its debut in our Cloud app, Astra, with much excitement and positivity. In the meantime, we were also working hard to get this wonderful model for local desktop rendering! I’m happy to announce that the team has just done that! The best part, we’re making this model available for everyone that meets the requirements to run it! This makes Starlight Precise 2.5 the 5th variant to the desktop Starlight Model Family. Note, that previous iterations of this model exists on Astra. This model has been specialized for realistic human face and skin textures in addition to other areas of improvement such as material, texture, and text.
image
image
857×1659 131 KB
About:
This model works for Medium to High Quality content whether this is generative, live action or restoration videos.
Improves generated content that may have soft, plastic, or slight artificial looks, however the AI generated content should have solid structure and composition for ideal improvements.
There is a slight difference between the local version and the cloud version in order to fit on the desktop.
Greatly reduces flickering with smaller textures.
Recommended to use video inputs at or under 4K.
Outputs up to 4K.
Model access via the cloud for all subscription models.
Cloud Export is supported on both Mac and Windows.
Expect long render times for this diffusion model. Appears to have faster processing compared to Mini and HQ.
image
image
3826×1370 737 KB
Technical things to know:
Nvidia Only
Available for all licenses to run locally! Yes, you read this right!
Requires a minimum of 12GB VRAM though 16GB VRAM is recommended.
Targeted for GPU compute compatibility of 8.9. This is also the recommendation.
This model is included locally, using neuroserver.
image
image
3849×1384 1000 KB
This model is less ideal for content with:
Shots with heavy motion blur may be rendered as texture.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Not ideal for content below 480p
image
image
1920×656 1 MB
image
image
3129×1371 815 KB
Additional Updates to Note!
Neuroserver
Neuroserver installation happens through the installer with a checkbox. An in-app update is recommended.
image
image
761×564 17.4 KB
All diffusion models (Starlight Family) now require Neuroserver: Starlight Mini, Sharp, Fast 2, and HQ.
image
image
1120×105 20.2 KB
Quality for the Starlight Models is expected to be the same.
If for some reason the Neuroserver is not working as expected, reinstall the app to run the “Repair” mode.
Topaz Video 1.3.1
Hi Everyone!
We have an update that addresses a few things from the 1.3.0 release.
Improvements and Fixes
Features
Starlight Fast 2 model -Founders access update
Starlight HQ model -Founders access update
Bug fixes:
AMD/Mac Starlight Mini issue fix
Fixed “Update Installation” pop up issues
Fixed compatibility check for Starlight Fast 2
Starlight Fast 2 Local
image
image
1920×1326 1.23 MB
Updated for Pro and Founders.
Recap about the model:
Introducing the new Starlight Fast 2, for local desktop rendering! Starlight Fast 2 is the third variant to the desktop Starlight Model Family. Note that Starlight Fast exists on Astra. This model is great to get good diffusion results at a much faster processing time. Note that in order to make this model both available locally and still be speedy (for a diffusion model) there is a slight difference in quality for local renders.
image
image
769×1228 96.6 KB
Things to know:
Nvidia Only
Requires a Pro or Founders license to run locally.
Best results come from 2x and above upscaling.
Great option for faster local diffusion based rendering.
Recommended to use video inputs at or under 4K.
Outputs between HD - 4K.
This model works for Medium to High Quality content whether this is generative or live action plates/videos
speed can be 3-5x faster than Starlight Mini pending on your GPU and available VRAM.
Provides better treatment for flickering/stuttering frames (especially when camera moves)
Requires a minimum of 16GB VRAM though 32GB VRAM is recommended.
There is a slight quality difference between the renders with 16GB vs 32GB VRAM. The latter being a tad more detailed.
Model access via the cloud for all subscription models.
Cloud Export supported
image
image
1920×1594 2 MB
Starlight HQ Local
image
image
1920×595 804 KB
Updated for Pro and Founders.
Recap about the model:
Introducing Starlight HQ, for local desktop rendering! Starlight HQ is the fourth variant to the desktop Starlight Model Family. This diffusion model has a great balance of detail and sharpness while keeping the intent of the original video source. Provides higher quality outputs for a wide range of videos, though not necessarily intended for archival content it can still provide good results. Though addressing similar intentions as Starlight Fast2, outputs have generally been higher quality and consistent with this model’s longer processing time.
image
image
630×1012 110 KB
Things to know:
Nvidia Only
Requires a Pro or Founders license to run locally.
Processing speed is very similar to Starlight Mini.
Supports 480p to 4k inputs.
Outputs between HD - 4K.
This model works for Medium to High Quality content whether this is generative or live action plates/videos
Requires a minimum of 10GB VRAM though 12GB VRAM and up is recommended.
Model access via the cloud for all subscription models.
Cloud Export supported
image
image
3197×1093 1.2 MB
This model is less ideal for content with:
haloing issues common with vintage digital videos
hair quality can vary in detail when content is on the lower medium spectrum.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Small repeated patterns can flicker
Bug fixes
AMD/Mac Starlight Mini issue with notification of “export disabled …requiring model download before local export.”
Fixed “Install now” update installation from app closing.
Fixed the cancel, close or install buttons (from the “Update Installation” floating window) from being unresponsive when there is active exports.
Fixed compatibility check for GPU minimum values to prevent failures for machines that can not run Starlight Fast 2.
Topaz Video 1.3.0
Hi Everyone,
We have a pretty sizable release!
The team’s been busy and we now have several exciting features for you!
Behind the scenes, the Starlight Mini model (being the first diffusion model for the local desktop experience) has a unique architecture that’s unlike any of our other models. We essentially brought the best diffusion/restoration model to fit in the local app. Diffusion models are computational beasts and are very slow when processing. Since it’s original announcement, we have been trying to find way to make Starlight Mini faster while keeping the quality the same. However, over the months, this has been challenging. Ultimately, we do not want to sacrifice quality nor the type of treatment you get from starlight mini in order to get faster outputs. Though, we will continue to work and refine the Starlight Mini experience, we have brought two new exciting Starlight family models locally.
For a more speedy diffusion render experience we have Starlight Fast 2.
Those interested in a model that offers higher quality renders for inputs up to 4K, we now have Starlight HQ.
Additionally, I’m pleased to announce Gaia 2, a new animation model that I’m sure you’ll enjoy!
Features
[PRO ONLY] Starlight Fast 2 model
[PRO ONLY] Starlight HQ model
Gaia 2, the new Animation model
New project settings dialog
Telecine migration
New Transforms + Crop area
Horizontal + Vertical flip
Bug fixes:
Non-Square PAR Issue Fix
MXF Encoder Restrictions Fix
Image Sequence Timeline Refinement
and more!
Pro: Starlight Fast 2 Local
image
image
1920×1326 1.23 MB
Introducing the new Starlight Fast 2, for local desktop rendering! Starlight Fast 2 is the third variant to the desktop Starlight Model Family. Note that Starlight Fast exists on Astra. This model is great to get good diffusion results at a much faster processing time. Note that in order to make this model both available locally and still be speedy (for a diffusion model) there is a slight difference in quality for local renders.
image
image
769×1228 96.6 KB
Things to know:
Nvidia Only
Requires a Pro license to run locally.
Best results come from 2x and above upscaling.
Great option for faster local diffusion based rendering.
Recommended to use video inputs at or under 4K.
Outputs between HD - 4K.
This model works for Medium to High Quality content whether this is generative or live action plates/videos
speed can be 3-5x faster than Starlight Mini pending on your GPU and available VRAM.
Provides better treatment for flickering/stuttering frames (especially when camera moves)
Requires a minimum of 16GB VRAM though 32GB VRAM is recommended.
There is a slight quality difference between the renders with 16GB vs 32GB VRAM. The latter being a tad more detailed.
Model access via the cloud for all subscription models.
Cloud Export supported
image
image
1920×1594 2 MB
This model is less ideal for content with:
haloing issues common with vintage digital videos
hair quality can vary in detail when content is on the lower medium spectrum.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Small repeated patterns can flicker
Pro: Starlight HQ Local
image
image
1920×595 804 KB
Introducing Starlight HQ, for local desktop rendering! Starlight HQ is the fourth variant to the desktop Starlight Model Family. This diffusion model has a great balance of detail and sharpness while keeping the intent of the original video source. Provides higher quality outputs for a wide range of videos, though not necessarily intended for archival content it can still provide good results. Though addressing similar intentions as Starlight Fast2, outputs have generally been higher quality and consistent with this model’s longer processing time.
image
image
630×1012 110 KB
Things to know:
Nvidia Only
Requires a Pro license to run locally.
Processing speed is very similar to Starlight Mini.
Supports 480p to 4k inputs.
Outputs between HD - 4K.
This model works for Medium to High Quality content whether this is generative or live action plates/videos
Requires a minimum of 10GB VRAM though 12GB VRAM and up is recommended.
Model access via the cloud for all subscription models.
Cloud Export supported
image
image
3197×1093 1.2 MB
This model is less ideal for content with:
haloing issues common with vintage digital videos
hair quality can vary in detail when content is on the lower medium spectrum.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Small repeated patterns can flicker
Gaia 2
image
image
1920×758 853 KB
It’s about darn time that we got a new animation model, Gaia 2! This model works on a frame by frame basis, similar to an image model, which is perfect for traditional types of animation. The unique method is why this model is in it’s own category. Continuing with the Gaia initiative this model has a great handle on animation. This model addresses aliased edges, compression and sharpens without “over doing it”. This is not a restoration model, so this will not remove film hairs and scratches or generate details. As a conservative model, it aims to preserve the original source intent.
Here are the types of content that this model best performs with:
Traditional Animation
Motion Graphics
Stop Motion Animation
Note that the model is not limited to the content above, so experiment and have fun with this one!
image
image
580×588 37.4 KB
Things to know:
Available for both PC and Mac
Processing speed can be faster or equivalent to Proteus!
Supports several input resolutions.
Support a wide range of output resolutions.
Best results come from 2x and above upscaling.
No modifiable settings.
Stack this model by adding a second enhancement with another Gaia2 to get sharper results.
Cloud Export supported
image
image
1920×642 276 KB
This model is less ideal for content with:
This model works for Medium to High Quality content and it is worth noting that it is not a restoration model.
Haloing issues common with vintage digital videos
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
If there’s heavy grain you may want to denoise first. Some grain particulates may be seen as part of the texture. Pending personal preference.
Project Settings Dialog
image
image
1446×1320 120 KB
The Project Settings Dialog can be found from the File menu item and replaces old input settings panel. Some of the options for settings include project-wide input colorspace, black + white, timecode options, and image sequence FPS.
image
image
2311×973 171 KB
Telecine Migration
image
image
513×1048 53.9 KB
Telecine has been moved to the right panel, under “Interlaced” enhancement for more visibility. Moving this under a category that makes a little more sense will help one find this option faster rather than remembering to look into the settings menu.
Transforms + Crop area
image
image
639×321 14.2 KB
We now have a new area that includes video Transform options and Crop in the same location. This area is identified with the same crop symbol with rotations.
Horizontal + Vertical flip
image
image
1375×763 74.4 KB
Sometimes capturing footage can get all flipped around and you may want a quick way to fix this. You now have the ability to flip your footage vertically or horizontally.
image
image
1075×1876 1.16 MB
Additional Updates to Note!
Improved Performance
Improved some VFR and dropped framed issues.
Bug fixes
Fixed issues regarding previewing, cropping, and rotating videos with non-square pixel aspect ratios. Please let us know if you continue to run into issues with specific formats.
Fixed MXF render failures with DNxHR codec.
Fixed timeline related issues with image sequence formats.
Added re-download option for cloud exports.
Fixed links on the “About Topaz Video” pop up menu
NeuroStream
As some of you have heard the exciting news that we are implementing NeuroStream with Topaz Photo. This is something we are looking into for Video as well. However, while this is being worked on and tested, more details will be provided at a later and more appropriate time. Thank you!
Topaz Video 1.2.1
Changes to authentication
We are excited to announce that we have reworked authentication to make using the app a more pleasant experience. We have relaxed restrictions on offline app usage to match our other apps, allowing up to 30 days (or until the end of your subscription) of offline usage after authentication.
Other fixes:
fixed incorrect “Invalid username or password” issues.
prevented seeing unexpected paywalls popping up while the app is open.
reduced delay for seeing credits show up for cloud exports.
Added support for the denoise filter in the cloud.
Added more error messaging for out of memory issues.
Resolved odd timeline behavior for image sequences.
loosened restrictions for encoding to Material Exchange Format (.mxf).
Topaz Video 1.2.0
Hi Everyone,
Today we have a new release with a good collection of updates!
See more details below:
Proteus Natural, is the next installment to our popular Proteus ai model. Like Proteus, this is intended as a general up-scaler for most types of content and can render locally and on the cloud.
Great for most video conditions and types
Addresses compression and general digital artifacts
Reduces halos common with older content or dark sharpening halos from digital zooming
Respects the original source material -doesn’t overly enhance/modify source inputs
Clean upscale that does not distort or warp original source details
Addresses aliased edges
Great model to mix and match with a second enhancement
image
image
1920×1569 1.45 MB
What’s the Difference between Proteus and Proteus Natural?
image
image
1920×1836 1.39 MB
This model takes the best out of Proteus and improves it with a more natural output that doesn’t over sharpen or “over enhance” everything.
Though the original Proteus is quite robust, at times the background texture can appear extra sharp or crunchy. Natural takes this into account and minimizes these types of issues.
Additionally, smaller faces (background people -small in scale) can also have distorted faces as a result of Proteus trying to understand a face from pixels of low quality video data. Proteus Natural, reduces this issue and allows for a more subtle touch up rather than an over estimation of the face.
image
image
1920×939 707 KB
Things to Note
Unlike Proteus, this does not have modifiable settings. It’s a “click and play” type of model.
Due to the more natural intention with this model, it is worth noting that it is not the sharpest model available and this is by design. If additional sharpening is desired, it is recommended to couple this model with a second enhancement for the desired look.
Though speed can be similar to Proteus, it can take a little longer.
Designed as a 2x model similar to Nyx. Like Nyx and Rhea (4x), these are not limited to the upscale multiplier. You can still output the model for any resolution that your codec can support.
Also available as a second enhancement.
Available for Windows (Nvidia & AMD) and Mac
Additional Updates to Note!
Nyx XL Fix
We have resolved the issue that some have experienced with dropped frames associated with this model.
Aion Mac Fix
Those on Mac, may have had issues with the interpolation Aion model causing a color shift and burn on the video. This has been addressed.
UI Refinements
image
image
1150×588 82.4 KB
Drive-root paths (e.g., F:) are displayed in workspace/export/model directories settings.
Export queue scrolling won’t snap to the bottom when selecting “Newest First.”
image
image
1552×980 819 KB
Welcome/Project page clips the projects grid and no longer overlaps the reference videos if you have multiple projects and need to scroll down.
Processing Update
Videos with non square pixels can be cropped and exported without error or change to the aspect ratio.
image
image
1845×667 85.2 KB
Previews with telecine turned on will display in the timeline. Previously doing a preview while Telecine enabled did not display the green visual notification of the processed preview on the timeline.
Batch Exporting Improvements
When batch exporting with Quick export, this now targets the default directory instead of the last used path.
image
image
1512×1129 159 KB
Clicking “Start Export” button (while batch exporting) now closes the Export Settings pop up menu. Previously this would stay open and require the “X” button on the upper screen right corner to be pressed.
Cloud Export Fix
Cloud downloads keep their progress regardless if you change the sort order.
Cloud output resolutions are honored. Few have received a 4K output rather than the expected minimum resolution for Starlight.
Additional Known Issues
Starlight Mini: Mac
For those with Mac that have updated to MacOS 26, there is an issue with blank or overly smeared color outputs. We are currently working with our friends at Apple to find a solution with this recent OS update. They have assured us that we will have a resolution to this soon.
Announcements
CLI Update
We want to share an update regarding Topaz Video: we’ll be gradually phasing out the Command-Line Interface (CLI) and its support.
This was a thoughtful and deliberate decision, driven by our goal of delivering a more reliable, flexible, and powerful experience over the long term. Moving away from the CLI allows us to:
Enable more customizable and modern workflows
Remove technical constraints that limit advanced features
Improve error detection, validation, and recovery
Reduce turnaround time when addressing processing issues
Provide greater flexibility for performance optimization
We recognize that this change may require some adjustment, and we truly appreciate your patience as we make this transition. Our intention is to focus on an application experience that is more consistent, dependable, and capable as the product continues to evolve.
To clarify, this change applies only to the CLI and support for CLI. It does not affect API access or custom solutions for enterprise customers. Note, that existing perpetual licenses will retain CLI.
Thank you for your understanding and continued support.
Verification Checks Update
In the coming weeks there will be a patch or official update that will reduce the amount of verification checks for subscribers of the Topaz Video App. This will align with our Photo and Gigapixel apps. For Topaz Video, the more frequent checks are no longer necessary and should allow for a more enjoyable experience with less interruptions while rendering. In addition, this should better support those dealing with areas having less than ideal internet service. Further details to come with the update.
Topaz Video 1.1.1
Fixes:
Fixed Starlight mini exports not adhering to resolution selection on Mac.
Updated Starlight Sharp cloud export resolution options to take into account maximum resolution.
Fixed second enhancement failing with Nyx XL as the first enhancement.
Fixed exports failing when preserving alpha layer and applying crop.
Resolve issues with “Export as” option and batch exporting on Mac.
Prevent overwriting original and existing files when using “Export as”.
Fix bit-depth related issues in Adobe After Effects plugin when using 16-bit color.
More precise reasons given for disabled previews.
More accurate metadata when exporting with Rhea/Rhea XL.
Topaz Video 1.1.0
Improvements and Fixes
[PRO ONLY] Starlight Mini: You asked, we listened. The Starlight Mini name is back!
[PRO ONLY] Starlight Mini: AMD and Mac improvements
Grain: greatly improved!
New Denoise Filter!
Plugin Updates: AE & Davinci Resolve
Fixed Starlight cloud resolution mismatch on the cloud for all platforms
Engine Refinements
Codec Update
And More!
Pro: Starlight Mini - Boomerang of a name!
You asked, we listened. The Starlight Mini name is back.
image
image
1075×331 64.9 KB
Some Historical Context:
Starlight Mini had dropped the “Mini” after the original Project Starlight was renamed in ASTRA (our cloud app) as Precision Quality. Precision Quality, could only render on the cloud due to the complexity and size of the model.
Starlight “Mini” was originally named for it’s more efficient model size that still keeps a high quality output.
With Project Starlight renamed, it was thought best to promote the Starlight Mini model’s name to just “Starlight” for the desktop experience. This model produces outputs that rivals the original Project Starlight while adding some unique properties that is more natural to the original content. So, the “mini” part of the name was removed.
Unfortunately, this only caused confusion. We’ve listened and learned. It gives us great pleasure to return the “Mini” back to this model’s name. By a large request, the Starlight Mini name is back to stay!
Moving forward, Starlight, represents a series of AI models. Models such as Starlight Mini, Starlight Sharp, etc are variations of this series with specified properties to enhance and improve your footage.
Pro: Starlight Mini: AMD & Mac
We have improved the quality and performance of Starlight Mini for AMD and Mac by addressing the following issues that some have encountered:
RGB channel shifts (color smears such as chromatic aberration)
Export fails even if when matching the minimum requirements.
A few reminders for Starlight Mini
You will need at least a Pro account to render locally.
A requirement of an NVIDIA GPU with at least 10GB of video memory, but we strongly recommend a GPU with 16GB or more memory for optimal performance and quality.
One of the minimum NVIDIA GPU requirements, is that the GPU needs the compute capability of at least 7.0, which means all the legacy GPUs with compute capability smaller than this version is not supported, including GTX 10 series and Quadro P series.
In addition to these GPU requirements, Starlight also requires a minimum of 32GB of system RAM.
It is recommended for AMD, to have 20GB of VRAM.
The recommendation for Mac is to have a minimum 36GB of RAM.
This is an AI model that is both complex and resource heavy. Meaning, that it could take hours to process.
It is not recommended to run Preview or Frame Preview with this model. For Nvidia users it will not work. AMD and Mac users this may also not work. It is recommended to export a few seconds for Preview purposes.
Per Cloud Render, there is a 5 minute maximum for Starlight.
Grain: Cinematic and Natural!
image
image
1920×870 877 KB
I am absolutely happy to announce the very very long awaited update for our Grain is finally here! Though our previous grain was good in it’s own right, it always felt a bit like an overlay and not scientifically accurate nor natural when compared to actual film stock or digital captures.
When light travels through a camera’s lens, it hits either the film that causes a chemical reaction or the sensor with ISO that has to be pushed further to interpret light, especially in low light situations. It is at this point for either the film or sensor interaction with light, where the grain or noise gets generated and affects your output.
What’s New?
We have overhauled the grain to take the following considerations into account:
image
image
507×262 12.8 KB
Luminosity levels (happens in the background)
RGB Channel wavelengths causes different strengths for grain (happens in the background)
3 Grain types (select-able)
Size (modifiable) - overall dimension of the grain.
Density (modifiable) - is the strength of the grain.
Amount (modifiable) - The quantity of the grain.
Grain Types
Gaussian: This is grain with color.
Grey: Similar to Gaussian, however there is no saturation.
Silver Rich: This is a more organic grain, something a little more akin to older film stock. The parameters don’t need to be as aggressive. More conservative adjustments is recommended from the default settings.
Denoise: Pure denoiser for High Quality content!
image
image
1920×767 1.06 MB
Nyx has been our leading denoise option for Topaz. However, Nyx and it’s variations are also enhancement models, meaning that although they reduce noise, the models intend to sharpen and modify the original content.
Denoise (powered by Nyx High-Fidelity) was made with a singular purpose, remove grain from high quality content without adding any additional enhancements. This is perfect for VFX teams that have film plates that need grain removal without any additional changes to their source.
image
image
499×207 12.2 KB
Things to Note
Please note, that there are no modifiable properties to this model.
Option to extract grain – You must initiate this option by clicking on the check box for “Extract noise on export.” This option will take the denoised output and input and create a noise map tiff sequence in the same folder of the export.
There is no option to access the noise map from the export directly.
Additional Updates to Note!
After Effects and Davinci Resolve Plugins
AE – fixed freezing/hang issues experienced on some machines
Overall speed for Davinci Resolve and After Effects are faster for enhancements
Quality of the enhancements are better than previous plugin iteration
Starlight Cloud Rendering
Fixed the Starlight cloud resolution mismatch on all platforms.
Additionally, add info in cloud dialog in edge cases.
Engine Refinements
Models should run, export and do frame previews for Windows AMD and Nvidia RTX 20 series cards with greatly reduced crashes. Overall giving a performance improvement for Windows AMD and Nvidia.
This also includes better general performance for 2x & 4x upscales.
Codec Update
image
image
613×724 52.9 KB
We have an additional Codec, DNxHR which includes the following Profiles
LB
SQ
HQ
HQX
444
Additional Updates
Fixed issue that requires double clicking for play
Fixed Starlight Mini and Starlight Sharp failures experienced by some users with some encoders.
Fixed issue of the application not fully closing at times.
Note:
If you experience any issues, please ensure that your drivers are up to date.
Topaz Video 1.0.4 - Patch
Fixes:
Fixed incorrect error message during network connection issues when logging in.
Fixed issue where messages not showing properly during EXR file importing when there’s less than 10 frames.
Fixed preset menu button for showing custom presets.
Skipped showing redundant model version dropdown menu.
Fixed missing icons in dialogs.
Topaz Video 1.0.3 - Patch
Hello everyone,
Today we have a sizable patch addressing the following updates below.
Be on the look out for more coming to Topaz Video soon!
Fixes:
Resolved label issue for trial button.
Rotation if functioning properly.
Login getting stuck when there are no seats.1
Raised minimum number of frames to 10.
During installation when a resource path is set, this is no longer getting reset.
Removed “crop to fill” for specific models that do not support it.
UI elements remains the same size, even when switching to bitrate mode.
Starlight Sharp will now fit within the download indicator.
Frame interpolation models no longer displays redundant model if there are no other iterations.
Topaz Video will no longer crash when cancelling an upload to the cloud
Repaired the issue with the occasional encoder mix-ups when switching back and forth between image sequence and video codec settings.
Always display the trimmer popup for starlight models.
Updated the grid/list selection preference to not be title case.
Removed the crop from being “off” on “letterbox”
Starlight resolution options have been addressed. Includes: minimum (original), 2x, 3x and 4x as long as outputs are between 720p and 4K.
Fixed a few source pane related bugs including “Close all sources” from failing.
Resolved frame number calculation error.
Refined the occasional issue of displaying source in the cloud “preview window” instead of the output within the cloud export page.
Fixed the issue of of copy/paste the interlaced/interlaced progressive model settings from not switching over properly.
There is now a warning tool tip for audio codec AC3 if an issue occurs.
Enjoy the patch updates!
Topaz Video 1.0.2 - Patch with Plugins
Hello everyone,
We’ve just released v1.0.2 with a couple of important updates based on your feedback:
Fixes:
Plugins Are Back
The plugins that didn’t make it into Topaz Video 1.0.0 have been restored to their previous state. They’re fully functional again.
We’re also working on a rewrite for these plugins behind the scenes to improve both performance and quality, so stay tuned for those improvements in a future update, coming soon.
Topaz Video 1.0.1 - Patch
Hello everyone,
Today we have a small patch addressing the following updates below.
Windows
Fixes:
Addressed Legacy toggles for EXR/DPX.
Starlight (Mini) cloud error refinements.
Fixed Starlight (Mini) download loop that some users experienced.
Repaired Starlight (Mini) RLM model issues.
Starlight family can now support 4K ProRes.
Seat fixes for Studio Pro users.
Removed frame preview pixelation.
Known Issues
Nyx XL runs slow on Mac for inputs above 1080p.
Topaz Video 1.0.0 (New Studio Release)
We’re excited to announce Topaz Video with an incredible new set of AI models and capabilities as a part of our larger Topaz Studio bundle of products and some of our most exciting features to date.
Here’s a quick preview of what’s in this release:
Brand new Nyx XL denoising AI model for all platforms. Denoise while preserving significantly more detail than ever before.
New AI Model Variant: Starlight Sharp: One of our most powerful diffusion-based video restoration models gets even more detailed. Starlight Sharp can restore with an extra level of crispness not present in any other offering. Super demanding, currently available on Windows NVIDIA platforms for local rendering, or on the cloud.
Live Frame Preview is a game-changer that allows you to instantly few updates as you adjust model parameters and settings in the app. This makes your workflow so much better. Note this feature is in Beta and doesn’t currently jive with some models like Starlight.
New General Enhancement Preset + Tuning: Not just a new enhancement preset that works in most scenarios, but also a single slider that dynamically adjusts multiple model settings as once to get the best results possible. A huge time-saver! You’re going to love this one.
Right Panel Updates: We’re continuing to refine the right panel to make it easier to do what you need to do as fast as possible. Presets are now easier to access than ever, and we’ve simplified the way we present models and settings. Clean!
Get the latest version now!
v1.0.0
Mac: Download DMG | Download PKG
Windows: Download
Released September 16, 2025
There’s so much more coming, and I’ll talk a bit more about that at the end of the post. But I also know that you probably have questions about Topaz Video and our Studio release!
What is Studio?
Topaz Studio is our collection of the latest edition of our desktop apps (Topaz Video, Topaz Gigapixel, Topaz Photo) as well as our web applications (Bloom, Astra, Express), mobile iOS app, and more! In short - it’s everything that Topaz has to offer to enhance your images and videos.
How does Topaz Video fit into Studio?
Topaz Video is available as a part of our new Studio Launch. With this launch, Topaz Video will only be available under a subscription license. Topaz Video will install into a new directory so you can try it out, and of course you’ll still have access to Video AI forever under the legacy perpetual license. Note that new features and new AI models will be coming only to Topaz Video, along with the other products in the Studio suite.
What if I already own Video AI or other Topaz products?
Purchased a Topaz product within the last year? Then you get to try Studio for free for the length of your remaining license. You’ll also get some extra benefits.
If you are a Topaz customer with an active license and you maintain your new studio subscription, you’ll be considered a Topaz Founding Customer which will give you access to some of our more powerful Pro models such as Starlight and other Pro features at no extra cost as well with your Individual license of Topaz video.
Can I buy the individual products or a perpetual license?
Topaz Studio is an incredible value - all for one low monthly price. You can still purchase subscriptions to the individual products as you wish, but you’ll get the best value with the Studio or Desktop collection bundles.
We will no longer sell perpetual licensed products moving forward, but we will continue to update them to maintain compatibility with major OS updates for as long as possible.
When will I get my free studio upgrade and access?
Please note that the transition plan is underway and may take a few days - but if you’ve already purchased and hold an active license for Video AI, or if you own more than 2 Topaz Products on an active license, you’ll have automatic access to Topaz Video at no extra cost once the transition is complete. We’ll keep you updated via email, and you can check the status of your account on the My Account page.
More questions?
If you have questions about the transition to Studio - please post them in this forum thread where Eric talks about the intention behind the release and our future plans:
Topaz Studio, the new subscription collection of all Topaz apps, begins rolling out today. Topaz Studio includes 7 products, over 100 AI image/video models, unlimited cloud image processing, and unlimited local processing. [studio-header] In 2025, we’ve released dozens of new AI models like Astra, Bloom, Recover, and Starlight. Previously, you had to individually purchase the ones you wanted. Now, you get access to everything at the cost of just two individual products. In that spirit, the S…
We know many of you are adjusting to the new subscription model, and our goal with this update is to make Topaz Video more valuable and accessible than ever — so you can spend less time worrying about limitations and more time creating.
Now - on to the latest and greatest updates available in Topaz Video!
New Starlight Model Variant: Starlight Sharp
A new addition to the Starlight family of models, Starlight Sharp excels at processing low-resolution footage, delivering superior results for even the most challenging content.
image
image
1920×680 76.7 KB
image
image
1920×911 148 KB
The model provides improved handling of smaller faces and facial details for enhanced clarity, while producing noticeably sharper and more detailed enhancement results compared to previous offerings.
Dual Deployment Options: Available on both cloud and local platforms for maximum flexibility
Multi-GPU Support: Leverages multiple NVIDIA GPUs in parallel for enhanced processing capabilities
To access Starlight Sharp - click the “Starlight” model family and select the Starlight Sharp model variant.
image (83)
image (83)
814×1273 130 KB
Starlight Sharp, can generate native 1080p outputs out of the box, or scaled 4K outputs using a special second pass. 4k outputs are currently using a Nyx variant under the hood for the second stage upscaling process.
Please note that Starlight Sharp is currently only available for local rendering to Windows users with NVIDIA GPUs, and is quite demanding in terms of VRAM, requiring 16GB of VRAM in order to run the model effectively.
Important Notes
Local rendering is only available to Founder and Pro users
Cloud Rendering options are available for all users
16GB VRAM required to run this model
Currently available only on Windows + NVIDIA GPUs
Longer render times expected due to increased model complexity
1080p or 4k output options available
Standalone enhancement and cannot be combined with other filters
Use square-pixel video footage at or under 1920x1080 to avoid aspect ratio shifts
If working with interlaced videos, we recommend de-interlacing first before enhancement
Text that is too small, blurry, or inconsistent may not be processed correctly
New Nyx XL Model
Yet another new AI model, but this time added to the Nyx family of models for incredible denoising.
a20eae3771a7bebc2d0f5cbf9fda91e6467a7c13_2_1275x523
a20eae3771a7bebc2d0f5cbf9fda91e6467a7c13_2_1275x523
1275×523 78.7 KB
The original Nyx is primarily a denoise model tuned for preserving detail in high-resolution input footage. Nyx XL takes this to the next level by learning the frames and outputting a clean and sharp output.
Nyx XL is optimized for sharpening and denoising a wide variety of footage in many lighting conditions.
This model is geared for high quality and high resolution media.
Nyx XL lends itself as a great chaining model, when merging with a different enhancement. A great results occurred when coupling Proteus with Nyx XL as the second enhancement. What kind of chaining combo can you make with Nyx XL?!
Notes
This is a more complex and heavier variant of the Nyx Model series. Please note, that this will currently produce slow render times. This is something that we are looking to make more efficient.
Currently available on both Mac and Windows.
Please note that previews can take quite a while to render.
It’s not recommended to use with Seek Render due to long processing.
This is a one size fits all type of model, so there are no modifiable settings.
Though Nyx XL is not limited by a resolution size, this variation of the model is not an up-scaler. Essentially, it will function at the resolution you input and render out the same. Though you can output at a higher resolution, the details and quality will be based on the original input resolution.
Please keep the long processing in mind with your machine’s capabilities/resources when calculating larger resolution sizes.
Live Frame Preview
3c529bd19334b4a971de408abead3ff80fd53e34_2_1275x721
3c529bd19334b4a971de408abead3ff80fd53e34_2_1275x721
1275×721 207 KB
You can now generate Live Frame Previews inside of Topaz Video, just by moving the playhead around with your current model settings.
Essentially, when you stop on any frame with an enhancement, the frame you are on will automatically load with an instant preview of said enhancement. Note, that this will not pre-render a range, just the single frame. The purpose for this is to get immediate feedback as you modify settings to find the results you are looking for.
image (84)
image (84)
466×235 39.9 KB
You have the option to disable this feature by switching the toggle “on” or “off” above the screen right side of the timeline. Currently, this is disabled by default.
Notes
This will still function if you have live render in preview enabled. If enabled, the live render will update if the video is played.
Seek render supports one or two enhancements, motion deblur and frame interpolation at this time.
Starlight AMD & Mac Support
image_720
image_720
719×319 110 KB
At long last - Starlight - one of our most powerful video restoration models ever, now runs on Mac and Windows systems with AMD GPUs in addition to the current Windows NVIDIA support. We will continue to optimize performance and quality of this model, but we’re excited to give you early access.
The model is also now a part of the Starlight “family” model selector, and loses the “Mini” designation, as it actually outperforms the base model in most scenarios. We will be building future versions of Starlight with new architectures over time - Starlight Sharp being the first in the series of next gen variants.
Please note that these Mac and AMD models are in Beta, being converted specifically for these platforms. As a result of platform architecture and the way that floating point operations are handled, there will be a small quality difference as well as a difference in system requirements (particularly VRAM) for these versions of the model. More in the notes below.
Notes
Local rendering is only available to Founder and Pro users
Currently listed as “Starlight” on the AI Model list, no longer “Starlight Mini”
This iteration of the AI model is both complex and resource heavy. Depending on the hardware you are using (AMD, MAC or NVIDIA) the performance can vary and take hours to process pending the media input. We will work to improve performance over time.
It is recommended for AMD, to have 20GB of VRAM.
The recommendation for Mac is to have 36GB of RAM
12GB of system RAM required to run this model
New General Enhancement Preset with Tuning
image (79)
image (79)
436×409 18.1 KB
Try the all new Tunable Preset: General Enhancement! The first of it’s kind, this specialized preset is built on three setting into one. As you move along the slider, this will adjust the preset parameters automatically so that you can determine how light or aggressive you want to modify the preset. At default, the modifier is centered between the two extremes.
General Enhancement has the following properties that it’s addressing: sharpness, compression and detail. This preset can also be utilized as an up scaler.
Right Panel Refresh
a43ed643f280a200712a4d50f770bd5433bda139_2_1275x1104
a43ed643f280a200712a4d50f770bd5433bda139_2_1275x1104
1275×1104 77.4 KB
We’ve made a number of updates to the right panel in Video AI to make it more accessible, especially for new users.
image
image
298×394 25 KB
In user interviews, we found that many users had difficulty finding essential tools like Starlight Mini, de-interlacing models, or presets. Others mentioned that the UI felt inconsistent and hard to navigate, especially when trying to understand how different enhancement settings worked together.
We’ve taken time to do a refresh of the right panel to make powerful tools more accessible, and ensure a better first-time experience that increases success with the application.
Clearer organization of enhancement options, with more intuitive grouping and consistent naming.
Improved discoverability for models like Starlight Mini and de-interlacing.
Presets are now easier to find and manage, making it faster to get back to your favorite workflows.
Refined parameter labels and controls, helping reduce confusion on usage and prevent mistakes.
Advanced Tuning is where you can chose your mode for the enhancement and modify accordingly.
We believe these changes will reduce friction for new users, streamline the editing experience for existing users, and ultimately help everyone get better results faster.
Topaz Video AI 7.1.4
Added Topaz Video studio download path
Topaz Video AI 7.1.3
Fixes:
Fix for the In-app updater causing a crash on some Mac systems
Fix for floating license enterprise export validations
Fixed issue where projects with missing input files did not open properly
Topaz Video AI 7.1.1
We have some updates to share in our latest updates for Video AI 7.1.1.
Bug Fixes:
Fixed issue where ffmpeg processes crash when watermark is applied.
Fixed issue of incorrect audio concatenation when scene detection is applied.
Fixed some auth-related issues.
Topaz Video AI 7.1.0
Hi everyone,
We are excited to share some of our latest updates for Topaz Video AI 7.1!
See more details below:
Video AI 7.1
Released July 17, 2025
Windows Download
Mac Download (DMG)
Mac Download (PKG)
Major Features
Cross-Scene Artifact Suppression for Frame Interpolation
image
image
1920×896 315 KB
With Scene Detection enabled, the Slow Motion function for videos, is better equipped to handle videos with cuts. This enabled feature will treat each shot/clip between cuts as independent entities. This will prevent blending/smearing of frames from the previous or next shot from your full video.
image
image
502×874 72.9 KB
As noted when you enable this feature, scene detection won’t live render.
Things to consider
It is recommended to have at least 8 frames per cut within your video to prevent potential stalled frames.
While rendering, this will split the scenes as individual parts. Ref Image: (Green) Original Source, (Red Arrow) Scene parts, (Yellow Arrow) processing scene.
image
image
979×463 77 KB
Within the Render Queue, if a process errors and the “Retry Failed Scenes” message populates under the Status section of the render Queue. Simply Click the highlighted “Retry Failed Scenes” button to continue the render.
image
image
611×171 23.1 KB
image
image
image
368×103 11.7 KB
If a crash does occur, the individual parts will remain so that you can continue from where you left off rather than having to re-render the entire video.
Once the entirety of the render completes, the individual scenes will be stitched back into a single video. Ref Image: (Green) Original Source, (Blue Arrow) Rendered Video.
image
image
958×393 52.2 KB
If any stalled frames occur, it is recommended to “Enable externally previewable exports” via the Crash Recovery section within Preferences
image
image
1161×735 89.4 KB
At this time, you will need to export locally to take advantage of this feature.
image
image
522×429 53.3 KB
Presets for Film Stock
The following Film Stock Presets will require a second pass enhancement. At default this feature is not enabled. Please refer to the Second Pass Enhancement web doc for enabling and more information.
Film Stock 4K Light
image
image
1920×742 164 KB
This setup is for film or progressive older footage that need a light touch to bring out some details.
4K preset
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Much of this setting can be used as a base and further modified to your liking.
Film Stock 4K Medium
image
image
1920×783 112 KB
This setup is for film or progressive older footage that need an extra push from the light touch version. This uses a different set of filters for the overall treatment.
4K preset
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Works best with high contrast black & white films like the example above. Please ensure that “Black & White” is checked (within the Input Settings -Gear Button next to Presets), if your film is grayscale.
Much of this setting can be used as a base and can be partially modified to your liking.
Film Stock 4K Strong
image
image
1920×779 103 KB
This setup is for film or progressive older footage with lots of grain/digital noise.
4K preset
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Great results for 8mm/Super 8mm films digitally captured with products such as Wolverine Data MovieMaker PRO, Kodak Reels Digitizer and Eyesen Film Scanner Digitizer.
Also good results with grainy 16mm telecine digital files.
Much of this setting can be used as a base and can be partially modified to your liking.
Preset for MiniDV
This setup is for MiniDV or digitally captured videos that made use of Interlaced footage (horizontal fields) for the time. MiniDV saw a large range of quality variations and conditions throughout it’s reign beginning around 1995 and ending around the late 2000’s. There was a huge use case scenarios with home movies to actual television shows and feature films.
MiniDV Int HD Basic
image
image
1920×788 122 KB
Due to the large quality and use case scenarios, the settings is a basic starting point for interlaced MiniDV footage. The intention of this preset is to get an average result that looks good in many cases. This setting is therefore intended to be modified.
HD Preset. Can be adjusted to 4k.
To make use of the Interlaced footage the settings are prepared with the Interlaced Video Type category.
The resolution is set to custom in order to keep the aspect ratio and pixel type as Square Pixel. Change accordingly.
This setting is intentionally and fully modifiable as a solid base for further refinement.
Disable Live Export
image
image
2653×1339 231 KB
You can now disable or enable the Live Export viewer.
You can find this option within the Preferences menu under the Advanced Features section.
Please note that this is enabled by default.
Disable Live Render
image
image
1920×1628 149 KB
You can now disable or enable the Live Render function.
You can find this option within the Preferences menu under the Advanced Features section.
When disabled, areas that do not have a rendered preview will automatically play the original source.
Please note that this is enabled by default.
EXR Alpha Support
image
image
1920×585 88.3 KB
EXRs with an Alpha channel, can now be supported by giving you the option to enable Alpha channel outputs via the codec settings.
image
image
658×340 45.2 KB
*Note, that at this time EXRs with Alpha may show transparency of the alpha as a checkerboard pattern in the preview. Outputs do not perform any matte operations, so the full of the RGB image will remain intact. Refer to the example below:
image
image
1657×1102 128 KB
DPX Non-Conversion
image
image
1920×631 83.7 KB
No longer will DPX files need to be converted to Tiffs when working with Video AI. Now you can use DPX files like any other image sequence.
Starlight Mini: 10x Faster in the Cloud!
image
Starlight Mini, the second diffusion-based upscaling model Topaz Labs has released, is now up to 10x faster in the Cloud when compared to its original iteration. Starlight Mini brings what makes the Original Starlight so special, into a more friendly and cost effective alternative on the Cloud.
Some Starlight Mini Reminders below
Starlight Mini is still a large resource intensive AI Model, so long renders are expected.
Recommended to use square-pixel video footage at or under 1920x1080 to avoid any aspect ratio shifts
Starlight Mini renders up to 4K resolution. Note: Starlight Mini output resolution options are limited.
Starlight Mini is a standalone enhancement and can’t be combined with other filters.
Please note that if you have text that is too small, blurry or inconsistent, the result may not be “sp3ll3d” correctly. If you can barely make it out, so will the AI model. It may assume that the text is a texture or pattern.
If you have interlaced videos, it is recommended that you deinterlace first before enhancing with Starlight Mini.
Learn more about Starlight Mini: Here
Pro GPU: Selection
Those with Pro accounts that have multiple GPUs, will be able to select the specific GPU for processing.
VP9 16K Renders
Revisited the VP9 codec to fully support 16K outputs.
Known Issues
For Cross-Scene Artifact Suppression the preview may display a frame of overlap. However, it will produce successful renders.
Enjoy the updates with 7.1!
Topaz Video AI 7.0.2
Hi everyone,
Today we have a new release with some great updates that we have been working on!
See more details below:
Improvements and Fixes
Film Stock Presets: Light, Medium & Strong
Requires 2nd enhancement enabled
MiniDV Preset
A great starting point for interlaced MiniDV videos
Added option to disable live export in preferences
Added option to disable live render in preferences
EXR Alpha Support
DPX no longer converts to TIFF
Fixed cloud renders from duplicating frames when using frame interpolation.
Presets for Film Stock
The following Film Stock Presets will require a second pass enhancement. At default this feature is not enabled. Please refer to the Second Pass Enhancement web doc for enabling and more information.
Film Stock 4K Light
image
image
1920×742 164 KB
This setup is for film or progressive older footage that need a light touch to bring out some details.
4K preset
This preset requires the Second Pass Enhancement to be enabled.
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Much of this setting can be used as a base and further modified to your liking.
Film Stock 4K Medium
image
image
1920×783 112 KB
This setup is for film or progressive older footage that need an extra push from the light touch version. This uses a different set of filters for the overall treatment.
4K preset
This preset requires the Second Pass Enhancement to be enabled.
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Works best with high contrast black & white films like the example above. Please ensure that “Black & White” is checked (within the Input Settings -Gear Button next to Presets), if your film is grayscale.
Much of this setting can be used as a base and can be partially modified to your liking.
Film Stock 4K Strong
image
image
1920×779 103 KB
This setup is for film or progressive older footage with lots of grain/digital noise.
4K preset
This preset requires the Second Pass Enhancement to be enabled.
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Great results for 8mm/Super 8mm films digitally captured with products such as Wolverine Data MovieMaker PRO, Kodak Reels Digitizer and Eyesen Film Scanner Digitizer.
Also good results with grainy 16mm telecine digital files.
Much of this setting can be used as a base and can be partially modified to your liking.
Preset for MiniDV
This setup is for MiniDV or digitally captured videos that made use of Interlaced footage (horizontal fields) for the time. MiniDV saw a large range of quality variations and conditions throughout it’s reign beginning around 1995 and ending around the late 2000’s. There was a huge use case scenarios with home movies to actual television shows and feature films.
MiniDV Int HD Basic
image
image
1920×788 122 KB
Due to the large quality and use case scenarios, the settings is a basic starting point for interlaced MiniDV footage. The intention of this preset is to get an average result that looks good in many cases. This setting is therefore intended to be modified.
HD Preset. Can be adjusted to 4k.
To make use of the Interlaced footage the settings are prepared with the Interlaced Video Type category.
The resolution is set to custom in order to keep the aspect ratio and pixel type as Square Pixel. Change accordingly.
This setting is intentionally and fully modifiable as a solid base for further refinement.
Disable Live Export
image
image
2653×1339 231 KB
You can now disable or enable the Live Export viewer.
You can find this option within the Preferences menu under the Advanced Features section.
Please note that this is enabled by default.
Disable Live Render
image
image
1920×1628 149 KB
You can now disable or enable the Live Render function.
You can find this option within the Preferences menu under the Advanced Features section.
When disabled, areas that do not have a rendered preview will automatically play the original source.
Please note that this is enabled by default.
EXR Alpha Support
image
image
1920×585 88.3 KB
EXRs with an Alpha channel, can now be supported by giving you the option to enable Alpha channel outputs via the codec settings.
image
image
658×340 45.2 KB
*Note, that at this time EXRs with Alpha may show transparency of the alpha as a checkerboard pattern in the preview. Outputs do not perform any matte operations, so the full of the RGB image will remain intact. Refer to the example below:
image
image
1657×1102 128 KB
DPX Non-Conversion
image
image
1920×631 83.7 KB
No longer will DPX files need to be converted to Tiffs when working with Video AI. Now you can use DPX files like any other image sequence.
Have fun with the 7.0.2 release and give these new and updated features a try!
Topaz Video AI 7.0.1
Features
Added option to disable live export in preferences
Improvements and Fixes
Applying rotation no longer highlights the export queue
For Mac, fixed navigation with arrow keys, such as using the keys to select a file
Fixed disconnected drives from being picked up on app relaunch
Improved starlight-mini downloader reliability
Updated Starlight-mini!
Resolves loading model hanging
Improved performance on higher end hardware
Fixed certain videos producing black bars
Video AI 7.0 - NEW Starlight Mini (Local) AI Model
Introducing Video AI 7
We’re excited to announce the biggest update for Video AI in a long time with version 7.0 releasing today, giving you access to a version of Starlight which can run locally (!!!) on Windows machines with NVIDIA GPUs.
That’s right. Starlight can now run on your local machine with Starlight Mini!
Please note we’ll be working on rapidly expanding support to other Windows GPUs and Mac platforms in the future. If your system doesn’t match the specs to run Starlight Mini, you’ll have to use cloud rendering for now.
Starlight MINI, you say? Skeptical about the results? Trust us - Starlight Mini is no slouch. It’s one of the best video restoration models we’ve ever built. You’re going to be blown away by what’s possible. Here’s an example:
Keep reading below to learn more about Starlight Mini … and everything else in this release of Video AI!
Video AI 7.0
Released May 13, 2025
Windows Download
Mac Download (DMG)
Mac Download (PKG)
Please review the system specs below to run Starlight Mini:
Operating System: Windows
GPU: NVIDIA GPU with at least 8GB of video memory, but we strongly recommend a GPU with 16GB or more memory for optimal performance and quality
RAM: Minimum of 12GB of system RAM
Starlight Mini local rendering will not currently work on AMD GPU/Intel GPU/macOS systems, though cloud rendering is fully accessible at this time. We’re working on making this possible! :slight_smile:
Major Features
image
image
1681×520 41.7 KB
The First-Ever Video Enhancement Diffusion Model that Renders Locally On Your Machine
Starlight Mini is the second diffusion-based upscaling model Topaz Labs has released, and it is the first to be offered as a standard in-app update for all existing and new users. Starlight mini brings what makes the Original Starlight so special, into a more friendly size for local processing.
The results of Starlight Mini can be just as astounding as the Original Starlight. Take a look:
Like its predecessor, it is capable of enhancing footage that may otherwise be a challenge with our more traditional AI Models. This can be anything from tricky 8/16mm film to mini-DV footage.
It even deals with interlaced frames, right out of the box :exploding_head:
System requirements for this model are more strict than the overall application.
System Specs to run Starlight Mini locally:
Operating System: Windows
GPU: NVIDIA GPU with at least 8GB of video memory, but we strongly recommend a GPU with 16GB or more memory for optimal performance and quality
RAM: Minimum of 12GB of system RAM
We’ll run a little check when you load up Starlight Mini locally for the first time:
Some notes on Starlight Mini below:
Starlight Mini is roughly 3x cheaper and will be up to 4x faster than the original version of Starlight to render in the cloud! However, the cloud version is not yet optimized in this release.
Recommended to use square-pixel video footage at or under 1920x1080 to avoid any aspect ratio shifts
Starlight Mini renders locally at up to 4K resolution. Note: Starlight Mini output resolution options are currently limited, and we’ll be improving these options in an upcoming release.
Starlight Mini is a standalone enhancement and currently can’t be combined with other filters. You’ll want to use the “reimport” function to add additional filters such as frame interpolation.
tpz-video-7-4
tpz-video-7-4
1840×2300 180 KB
If you’d like to review the Starlight Mini workflow in more detail, please take a look at our Docs page here which contains examples of exactly how to access the model and best practices for using it.
Starlight Mini local shows different options for resolution output from the model, based on the size of the input video.
image
image
1054×853 61.3 KB
Starlight Mini will also respect the max GPU memory settings you have in preferences as a percentage of total available VRAM:
image (51)
image (51)
1554×998 117 KB
Live Export Previews
Live previews are now visible in the export queue! This is a game-changer.
Before, there was no easy way to compare exports or cloud renders, or even to preview results inside the application.
Now, you can confirm that your outputs are matching your intended look without having to wait for a complete render.
You can also compare rendering/completed output against source video:
image
image
1920×1016 136 KB
Additional Improvements
EXR Support
Video professionals have been asking for an improvement to EXR support, and it is finally here!
image (55)
image (55)
1920×899 132 KB
TIFF conversions for EXRs are no longer necessary
32 bit image sequences are now fully supported
Video AI provides greater color support for ranges outside of the standard 0-1 range
Hear the Difference with Volume Control!
Take full advantage of controlling your video’s volume. Though you may need to hear and see an enhancement with sound, you no longer need to do this with blasting sound waves!
image
image
471×197 9.99 KB
Sometimes it’s the little things that matter.
Pro Account Panel Updates
Account-level Generative AI Controls: Video AI Pro admins can now toggle generative AI features on and off for all sub-accounts. If generative AI is turned off, the Starlight capabilities will be hidden from the interface.
Account-level Cloud Render Controls: Similar to the controls above, admins can also toggle Cloud Rendering features on and off for all sub-accounts. If the toggle is off for cloud rendering, you won’t see those options in the UI, including Starlight cloud options.
Enjoy the unique qualities from this “first of it’s kind” Starlight Mini diffusion model on the local desktop and all the updated features of 7.0!
Known Issues
Live Export player may display over-exposed results for HDR model outputs in some scenarios
https://community.topazlabs.com/t/video-ai-7-0-new-starlight-mini-local-ai-model/90631
6.2.2 released on May 8, 2025
6.2.2 update was released to fix crashes on Mac
6.2.2 update also included fixes to login flows on all platforms
Fixed Topaz Labs plugins crashing Davinci Resolve
The two plugins are now combined a single plugin (Enhance)
To chain different models, chain multiple instances of the plugin
Fixed Topaz Labs plugins crashing After Effects
Fixed ffmpeg AV1 encoding for RTX 5000s
Added volume slider to the preview controls
6.2.1 released on May 6, 2025
Fixed Topaz Labs plugins crashing Davinci Resolve
The two plugins are now combined a single plugin (Enhance)
To chain different models, chain multiple instances of the plugin
Fixed Topaz Labs plugins crashing After Effects
Fixed ffmpeg AV1 encoding for RTX 5000s
Added volume slider to the preview controls
Topaz Video AI 6.2.0
Login flow updated.
TensorRT 10 update for 5000 series NVIDIA GPUs.
Improved export error messaging.
Disabled app actions on Welcome screen.
Fixed EXR to TIFF conversion sometimes rendering invalid TIFFs.
Fixed FPS counter in Cloud export view.
Pass crop to fit properly in cloud.
Fixed opening area for preset dialog.
Scale CPU usage utilization reduced for 2x/4x outputs.
Topaz Video AI 6.1.3
UPDATE: 6.1.3 patch now available with fixes for image sequence exports, thanks to all who reported this quickly!
Topaz Video AI 6.1.2
Added 4K Starlight renders.
Added Cloud export history for renders <7 days.
Added detail to metadata when HDR model is used.
Added unique ID for each OFX plugin.
Added trackpad pinch zoom for preview and timeline.
Added dynamic compression settings for AV1 (NVIDIA).
Fixed trim intermittently not being applied for Starlight.
Fixed cases where image sequence naming contains no less than 10 digits.
Fixed warped thumbnails for non-square PAR input videos.
Fixed reset timeline zoom on input video info changed.
Fixed image sequence frame count when duration changes.
Fixed focus reset when becoming disabled via exports panel.
Fixed wrong frame/black frame on switching back to input with previews.
Fixed single view audio error.
Fixed hidden installer window on app opening.
Fixed scaling for alpha layer with aspect ratio change for pillarbox/letterbox and crop to fill.
Fixed video thumbnail being included with image sequence exports.
Fixed trim hanging for cloud exports.
Fixed image sequence start number setting if input has its own start number.
Fixed Pro user reclaiming seats when they have expired personal license.
Fixed auth token for foreign characters.
Topaz Video AI 6.1
Today’s release is a very important moment in the history of Topaz Video AI. We are releasing Project Starlight 41, the first-ever diffusion model for video enhancement.
This is a significant evolution over our existing models, and it’s the single largest increase in model capability since Video AI launched.
Even the most challenging footage can be enhanced with Project Starlight. So please, try out the “bad” footage—footage that may have previously produced artifacts with other models.
Because of its size, this model is significantly slower and more expensive to run than previous models, and requires cloud processing on server-grade graphics hardware. But this is just the beginning. Our mission as a product and research organization is to advance visual quality to the furthest extent possible—it just takes a little more processing power to get there right now.
We see a clear path towards expanding access to Starlight and other models of this generation to very high end desktop GPUs, and Project Starlight is just the start of a new series of models.
Project Starlight Research Preview
Active Video AI users can currently process three 10-second weekly previews of Starlight for free. Results will be viewable through unlisted, shareable links with the new “Compare” button in Video AI.
These previews take about 20 minutes to render if servers are available immediately, and all exports are delivered at 1080p.
Starlight Early Access
Video AI users can also render up to 9000 frames of video using Starlight (5 minutes at 30 fps).
The current introductory under-cost pricing of 90 credits/minute will change over time as our team expands access and server capacity.
Our research team is eager to receive feedback and discuss this step-change in AI video upscaling. We appreciate your involvement on this journey to Video AI’s next frontier.
Project Starlight announcement on X 8
Topaz Labs CEO Eric Yang has also written a new blog post in our Roadmap section to talk more about Project Starlight. 18
Want to learn more about Enterprise deployments for Starlight and other models? Our new page details different options for bringing Video AI to your team. 5
Project Starlight is also available in the Video AI web app.
https://app.topazlabs.com/starlight
Topaz Video AI 6.0.4
Preview queue added! Monitor all active previews in one menu.
Several improvements to OpenFX performance/stability.
Use matching parameter values from Enhancement filter for cloud exports.
Cloud exports now use the Codec settings selected in-app.
Removed “Force CFR” control, replaced with fixed VFR exports.
Fix batch estimates for certain queues.
Fix cloud download paths.
Fix a crash on closing app.
Fix frame interpolation toggling on when telecine is enabled in input settings.
Allow cloud uploads and downloads with utf-8 characters in path to files.
Preserve zoom setting on both players when player dimension changes.
Make scrollbar visible on the Exports panel.
Fixed snooze update not applied.
In-app video player now always uses input tab’s audio track
Cloud exports list now view auto scrolls to most recent item.
Fixed cross-input queued preview injection.
Prevent lone export item from disappearing.
Only show PAR type + crop setting options under Enhancement in batch if batch items require these controls.
image
Topaz Video AI 6.0.3
Enabled Enhancement in UI when combo box for output resolution is changed from original.
Enabled HLG transfer function for SDR HDR.
Renamed focus fix tooltip to match buttons.
Updated warning text for forcing CFR.
Speed improvements to Frame Interpolation plugin for OpenFX.
Fixed Davinci Resolve plugin failing to load.
Fixed Davinci Resolve Frame Interpolation plugin sometimes not installing.
Fixed Davinci Resolve crashes when running Auto mode.
Fixed naming of duplicate tabs in tab options.
Fixed default preset behavior – will now create a new tab with the preset applied.
Fixed zoom resetting on play/pause.
Prevented preset tab from setting audio mode to none.
Keep manual parameters when viewing codec settings.
Always show second enhancement if active.
Allow network path imports and exports with Video AI Cloud.
Fixed Export priority handling.
Allow setting audio settings in batch exports.
Add zoom in/out keyboard shortcuts (Ctrl/Cmd +/-).
Accept more naming schemes for image sequences.
Improved accuracy of Cloud Credit balance.
Reset enhancement params to 0 if set to auto.
Keep bitrate export settings when switching between video + image sequence export options.
Keep file name during EXR conversion.
Prevent incorrect values for scaling with non-square PAR inputs.
Topaz Video AI 6.0.1
Included a “force CFR” (Constant Frame Rate) setting in encoder options.
Added a new feature to compare tabs in single view mode via a tab option.
Fixed stabilization filter bug where enabling jittery motion does not update filter correctly.
Prevented model parameters from resetting when navigating to the export page.
Added a display for elapsed time in export items.
Brought back tooltips for view mode buttons.
Moved the status label and added an estimated time for preview.
Fixed crop display bounds during zooming by ensuring the display area always aligns with the video view.
Known issues:
Enabling “force CFR” option may cause issues in the exported videos such as duplicate frames, missing frames, or frames in wrong order.
As always, thanks for being part of the Topaz Video AI Community.
6.0
This release represents a major revamp of preview comparisons, with a tab-based interface that allows for workspace customization. Using tabs makes comparing different model settings much simpler, and brings more screen real estate back to the playback windows.
Here’s a video overview of what’s new in v6!
Introducing Topaz Video AI 6: Smarter, Smoother, Stronger!
Introducing Topaz Video AI 6: Smarter, Smoother, Stronger!
Highlights
Tabs
We’ve heard your feedback and we want to make comparison, easier, better, and more intuitive. Tabs are self contained groups of settings. You can create them from scratch or use presets. They replace for our previous preview queue and A/B comparison mode. Now rendered previews are contained in each tab, and it’s easy to compare two sets of previews, or the original. Just drag the tabs to create the view you want.
Video player
Our upgraded video player has been available in Windows and is now coming to Mac. The video player supports HDR, audio, and smooth 8K playback. Getting high quality previews in app, has never been better.
Note: HDR support on Mac is still in progress
Exports
We’ve revised the export queue to take up less space, offer more information about your exports, and make it easier to track exactly what you’re rendering.
Sources
We’ve moved the sources panel to create more work area. You’ll find it right below the timeline on the right.
Presets
The preset system has been refreshed to make it easier to find your favorite settings and quickly apply them across projects. You can select a preset directly when starting a new project. Presets are tightly integrated with tabs, you can create a tab from a preset, or save a tab to a preset.
Timeline
We gave the timeline a lot of love in this update. The upgraded timeline is faster, more responsive and has thumbnails to help you find what you’re looking for.
Live render
We’re making huge improvements to the live render system. You can render specific length chunks using the render button, and now it’s easier than ever to compare two sets of settings. Just switch into Side by Side or Split View, and press the play button to start rendering.
Other updates:
A new Preferences toggle is now available to enable/disable Cloud processing in Video AI:
CleanShot 2024-12-03 at 19.13.39@2x
CleanShot 2024-12-03 at 19.13.39@2x
1920×1232 63.6 KB
Notes:
Stabilization is currently not supported as a Live Render filter in previews. Stabilization previews can be rendered using the Render button.
5.5.1
Using Topaz Cloud Credits 277, Video AI users can now offload heavy processing tasks to our servers for faster processing.
Use cases:
Processing large batches of videos in parallel: There’s no limit to how many simultaneous jobs you can queue on the Cloud Rendering service, which can greatly improve your overall output of clips.
Processing long videos on low-power machines: Cloud Rendering excels when working with videos that are 30 minutes or longer, which allows us to parallelize across more GPU instances. For users on laptops with less-capable GPUs, you can sleep your machine without worrying about renders being paused or cancelled.
Known issues:
H.265 and AV1 are the only enabled codecs for now. We plan to expand the list of supported codecs as we determine system stability. Cloud renders larger than 8K are also disabled for this release.
Video inputs under 10MB are not currently supported for Cloud Rendering.
Hyperion is not currently enabled for Cloud Rendering.
We’re excited to work together on furthering the efficiency of Cloud Rendering, and as always we appreciate your involvement and feedback!
New in 5.5.1:
Adds support for Intel ARC ‘Battlemage’ B580 and B570 GPUs
5.5
Using Topaz Cloud Credits 7, Video AI users can now offload heavy processing tasks to our servers for faster processing.
Use cases:
Processing large batches of videos in parallel: There’s no limit to how many simultaneous jobs you can queue on the Cloud Rendering service, which can greatly improve your overall output of clips.
Processing long videos on low-power machines: Cloud Rendering excels when working with videos that are 30 minutes or longer, which allows us to parallelize across more GPU instances. For users on laptops with less-capable GPUs, you can sleep your machine without worrying about renders being paused or cancelled.
Known issues:
H.265 and AV1 are the only enabled codecs for now. We plan to expand the list of supported codecs as we determine system stability. Cloud renders larger than 8K are also disabled for this release.
Video inputs under 10MB are not currently supported for Cloud Rendering.
Hyperion is not currently enabled for Cloud Rendering.
5.4
1. All-new, GPU-accelerated video player
Video AI 5.4 uses an mpv 10-based video player that unlocks incredible improvements to playback performance. Scrub through highly-compressed input videos, or massive uncompressed media with near-instant responsiveness. This new player also supports HDR displays and audio playback! The old live preview behavior with the “Render” button is also restored. Right now, the new player is available on Windows only. We are working on bringing this player to macOS very soon.
Playback of 8K video with old player:
Playback of 8K video with new player:
2. Hyperion SDR-to-HDR model
This model is designed to perform inverse tonemapping by increasing the bit depth, expanding the color gamut, and brightening the highlights of input videos. It improves contrast and recovers detail in highlight and shadow areas of standard dynamic range inputs, and produces HDR10 (BT.2020/PQ) output.
Controls:
Exposure adjustment slider: This slider allows the user to increase the exposure of the scene by up to 1 stop. This adjustment primarily affects the brightness of the shadows and midtones.
Highlight threshold slider: This slider explicitly tells the model the lowest intensity level to be considered a highlight. This slider should be set to a value higher than the brightest skin-tone (or any other color where accuracy is vital) present in the video. In practical terms, we recommend starting with the default setting and raising the value if undesirable color shifts appear.
Saturation boost slider: The HDR color space has a larger color gamut and supports more saturated colors. Turning this slider up results in more saturated outputs in HDR without loss of detail. This slider will, however, lead to a color shift.
3. Rhea XL (RXL) Enhancement model
RXL is an even larger variant of Rhea, and is designed to handle complex textures in a more refined way.
The model requires significantly more VRAM and GPU power than previous models, so expect some longer processing times compared to Proteus or Iris.
Here’s a link to an interactive player window with a sample of Rhea and RXL’s output with the same settings applied. 51
RXL is currently listed as a “model variant” under Rhea in the Enhancement filter menu. This may change in future updates based on feedback and model fine-tuning.
Known issues:
Hyperion SDR-to-HDR does not currently support Intel-based macOS systems.
5.3.6
Fixed playback stalling from corrupt projects.
Fixed copy/paste settings not working.
5.3.5
Fixed preset name being lost on input switch and view switch
Fixed interlaced + interlaced progressive models not saving to presets
Fixed copy/paste settings from export to input
Fixed default preset causing project to not load + authentication to reset
Fixed crop+focus fix causing wrong crop area on preview and exports
Fixed combo boxes being scrollable while open
5.3.4
Fixed exports not closing in certain conditions
Fixed export jumping to top of list despite sort order
Fixed playhead cursor resetting to middle of timeline on pause
Fixed Live Render sometimes not caching (being able to seek into live rendered areas and see the preview)
Fixed Input filenames falling out of box and not eliding
Fixed audio codec disappearing in preferences when PCM is selected
Fixed copy-paste overriding fps settings
Fixed empty directories in main drive
Fixed CPE failing from missing directories
Fixed batch processing ignoring codec settings, depending on preset
Fixed bug where using batch would reset your codec settings
5.3.3
Updated batch processing
Updated flow
If you dont change settings with multiple sources selected, batch processing will use each videos unique settings.
If you do change settings with multiple sources selected, batch processing will propagate the changes you made to every input video.
Fixed multiple resolutions and model settings applying across wrong videos.
Fixed batch processing resetting AI model.
Fixed timeline playhead centering issue when quickly playing and pausing.
Fixed bug where timeline is missing major tick.
Fixed timeline padding issue.
Fixed several issues with the hovertip of input items:
FPS in hovertip now will update with the change in input settings for image sequences.
Fixed precision for FPS field.
Added feature to recommend models based on input video height.
Enabled selection of recommended models automatically on turning on Enhancement filter.
Fixed - Last major tick now only visible if minor ticks follow.
Fixed issue where playhead will disappear when going out of bound of timeline view: now reappears at the beginning of the timeline or at the smallest timestamp it can reach without causing the timeline to go out of bounds.
Fixed AV1 appearing for some unsupported GPUs.
Fixed incorrect end trim marker when placed on the frame before last.
Removed autozoom of timeline on rendering a preview.
Modified “Reset timeline zoom” button functionality to match with the older versions.
Fixed bug where image sequence start frame number text field not working.
Updated End User License Agreement
5.3.2
Added zooming into timeline with mouse wheel.
Updated timeline to center around cursor head on pause.
Updated End User License Agreement.
Fixed issue where video views lose rotation on project load.
Fixed bug where changing image sequence fps values would not update the playback fps.
Fixed bug where changing image sequence fps values would not update timeline.
Quick export now correctly identifies how many jobs it will be exporting within batch processing.
Export as now correctly disabled for batch processing.
5.3.1
Fixed a bug where crop settings no longer displayed under resolutions that changed aspect ratio
5.3.0
Control panel updates
This update includes a new Model Discovery system in the right panel, which shows examples of when to use each Enhancement model:
In addition to the new carousel, users can also switch to a list view for more compact controls:
This new model selection tool will help users navigate our growing collection of AI models and find the right fit for any project. We’ve also added a star icon for models that have very broad applications.
Timeline updates
Video AI’s timeline now offers much more functionality, with frame-accurate timeline markers and a new system for timeline zoom:
This change will make working with very long videos much more manageable, and offers precise seeking for trimming.
5.2.3
Fixed network share locations causing an app crash on export on some systems.
Fixed certain Radeon cards showing the AV1 export option when not supported in hardware.
Fixed issue where batch processing overwrites output resolution of all exports to the current opened video’s export resolution.
Fixed issue where focus-fix option enabled will lock output resolution options while Enhancement is off.
Updated minimum window size to avoid UI elements overlapping.
Improved login flow for Standard+Pro license owners.
5.2.2
Fixed AI Processor reverting to Auto
Fixed Low Power Mode not appearing on macOS systems
Improved error messaging – fewer errors will be incorrectly reported as “Colorspace error”
Improved alpha layer preservation, image sequence input image sequence output now works across TIFF & PNG
Fixed constant bitrate exports for NVENC - now uses -rc cbr for setting rate control in constant mode
Fixed bug where out trim marker was not moving for short inputs
Fixed bug where filling intervals less than 5 frames would error out
Renamed Instant Render to Live Render
Fixed preview intervals merging - will not re-render processed areas but will also not merge the two intervals
Preview render region will always start where specified + fill in gaps until end of preview region
Update 07/11: Video AI 5.2.1 is now released with a fix for videos stored on network shares sometimes not importing correctly.
Today we’re excited to release Video AI 5.2!
This release includes the new Rhea Enhancement model, Pro seat management, Frame Interpolation for After Effects, Alpha layer support, and many UI and backend fixes.
Rhea Enhancement Model
Rhea represents a combination of Proteus and Iris. The model is intended to be more accurate in preserving fine details across a wider range of subjects, while also handling text in a less destructive way.
This model and Rhea internally processes inputs at 4x scale and then downscales to your selected output resolution.
We’re very excited to see your results with Rhea, here are some example frames from this model:
Frame Interpolation for After Effects
Today we are also launching Frame Interpolation in Adobe After Effects. This plugin addition includes access to all Frame Interpolation models for up to 16x slow motion conversion.
For more information on using the After Effects Frame Interpolation plugin, see this post in the Video AI Plug-ins forum section. 2
Alpha (transparency) layer support
Video AI can now copy alpha layers from inputs and merge them back with AI model output videos. Alpha layer copying is supported in the following codecs:
QuickTime Animation
TIFF
PNG
image
Pro seat management
After a pre-launch period with several Pro teams around the world, Video AI Pro will be officially launching later this week. This new license tier offers seat management, commercial usage rights, and multi-GPU support for teams working with Video AI.
As part of this transition in licensing, we will be discontinuing the previous implementation of multi-GPU rendering for the standard license. Our team found that less than 2% of users made use of the multi-GPU setting, and further development will be focused on multi-GPU optimization for Pro teams.
If you are affected by this change and would like to discuss further options, please contact enterprise@topazlabs.com
Bug fixes & other improvements:
Preview set saving functionality restored.
UI lag greatly reduced, specifically for long operations.
Ensures colorspace correctness for input videos with a wider list of colorspace related flags.
– Uses the setparams filter instead of -colorspace, ... flags to explicitly override the input metadata to ensure only the metadata tags are updated without changing the video data itself.
Timeline width now adjusts correctly between different video inputs.
Export preferences now save consistently.
Added pause/resume controls to right-click menu.
Fixed crop settings auto-populating after first digit.
Time remaining/elapsed toggle now available for both Previews and Exports.
Fixed “Open in Explorer/Finder” button for previews and exports.
Disabled automatic generation of TIFF sequences for Previews. These temporary files are no longer in use and the app now reads preview files in the selected Export format.
5.1.4
Stabilization no longer has potential to hang app on swapping between ‘Previews’ and ‘Exports’ tab
In-to-out no longer disables the render button if your cursor is near end of timeline
Fixed Certain inputs making the timeline UI bugged
Fixed certain files not responding to ‘Show in explorer’
Fixed default export area not remembering your choice
Fixed default profile export setting not remembering
Fixed clicking render preview sometimes not putting you in processed area
Fixed many scrollable areas not hiding scroll bar when not needed
Fixed zoom percentage not updating when using shortcuts
Fixed buggy UI for image sequence Video Settings option dialog
Fixed broken shortcuts for all player controls in Welcome page
Fixed crop snapping to min/max before you finish typing
Fixed clicking cancel on preset creation still creating the preset
Fixed bug where you could edit video input settings in batch mode
Fixed ‘Include Preview & Export Settings’ in preset dialog closing the dialog
Added time elapsed/remaining to previews section
Added ‘Close Project’ button under ‘Save project as’
Added “Remove Noise” preset to defaults.
5.1.3
Fixed render in-to-out range.
Fixed player looping over same beginning region when saving with a trim in projects.
Fixed various UI icon related problems
Fixed 'None' option showing twice in 'default audio settings' area in preferences
Fixed Instant Render rotation showing stretched output
Fixed EXR failing to convert / import
Fixed image sequence exports being lost intermittently
Fixed some Instant Render related crashing
Fixed a bug where toggling Instant Render only applies to the currently selected view
Previews no longer overlap when generating and will fill in gaps instead for optimal performance
Added full app version to the header
5.1.2
Fixed exports not appearing.
Fixed image sequence related crashing.
5.1.1
Fixed slow performance of exports.
Fixed Low Power Mode missing in preferences.
5.1.0
1. Instant Rendering (Experimental)
Video AI now renders previews as soon as you click ‘Play’.
Instant renders make the timeline more interactive and allow you to directly compare models with shorter wait times.
2. Frame Interpolation for DaVinci Resolve OFX (macOS + Windows)
Access Apollo, Aion, and Chronos directly from DaVinci Resolve. Convert to slow motion at up to 16x interpolation and make use of Resolve’s retiming controls for smooth speed-ramping.
To use Frame Interpolation in Resolve, be sure to follow the guide posted here. 7
3. New multi-GPU options (Video AI Pro)
Video AI now includes two multi-GPU modes. The existing “All GPUs” option now supports NVIDIA cards running TensorRT models, leading to major utilization increases for systems with 2+ NVIDIA cards. This mode is now named “Single video” under “GPU Settings”.
The second mode, “Multiple videos”, optimizes for large export queues when running on systems with multiple GPUs. We’ll have more to share about this Video AI Pro feature when it launches in June.
4. New Welcome Screen
Quickly reopen projects and resume renders using the new Welcome Page. This is a home for recent projects and includes the ability to select favorites for easy access.
5. Preferences UI Refresh
Navigate settings using the new preferences sidebar with expanded categories and new tooltips.
6. Colorspace settings added to Video Input Options
Set custom colorspace, color primaries, color trc, and color range for video inputs with metadata issues.
5.0.4
Changelog from 5.0.3
Fixed a bug where large file names take up all the space in export panel
Updated wording for Output Resolution area in Video Settings
Added project name to Windows title bar
Updated app icon for Windows (higher resolution)
Fixed empty time display mode in export panel
Added the 'hide filters' button to the Windows title bar
Fixed failing import of EXR/DPX conversions
Updated presets for old models (Proteus v3 -> v4, Chronos Fast -> Apollo)
Upgraded to FFmpeg 7 + OpenVINO 2024.1
Known Issues
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
5.0.3
Changelog from 5.0.2
Added back reset button for timeline zoom.
Added .SER as an importable file type.
Fixed certain input tiff sequences failing to load with projects.
Fixed 0s elapsed bug when finishing export.
Fixed stabilization not reloading in app (loading with error).
Fixed benchmarking failing to load.
Fixed trim + crop + rotation persistence across projects.
Fixed EXR+DPX failing to reconvert.
Known Issues
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
5.0.2
Changelog from 5.0.1:
Fixed variable frame rate exports.
Fixed rotation previews in Workspaces.
Fixed overriding FPS values in Workspaces exports.
Fixed some Workspaces related corruption of project.tvai.
Fixed main selected video sometimes not being remembered.
Fixed rotation/crop/trim related loss of thumbnails.
5.0.1
Fixed trimmed exports with audio not functioning as expected.
Fixed intermittent crashes caused by previews.
5.0
1. Workspaces
Video AI now keeps a library of active previews and exports, allowing for project-style organization.
Your app workspace will be saved and reloaded to pick up sessions exactly where you left off using the same preview files.
2. Pause/Resume for Exports
You can now pause and resume exports in Video AI.
Prioritize different queue items on-the-fly, start up a different project, or even switch to a GPU-intensive game without losing render progress.
Even after system reboots, power loss, or unexpected crashes, your export sessions can be restored.
Paused exports are saved to disk, freeing up all CPU and GPU utilization.
3. Wonder UI Refresh
New colors & icons across all Topaz Labs apps for a unified, streamlined look.
4. Nyx v3
Nyx 3 improves the consistency of denoising across a wider variety of videos.
The accuracy of the auto-parameter estimation has been improved, and the sharpen control has been restored.
The model no longer sharpens by default, significantly reducing the unnatural grainy texture ( orange peel artifacts 6). The figure below illustrates this issue in Nyx 2 and the result from Nyx 3.
fewer Moire pattern artifacts
image
image
1920×1080 169 KB
Input (Left) Nyx v2 (Center) Nyx v3 (Right) [Gamma increased for comparison]
5. DaVinci Resolve/OFX plugin updates
Dione Dehalo & Dione Robust added to Enhancement models.
Motion Deblur moved to a separate section of the plugin control panel.
Caveats / Notes:
The following settings must be enabled for pause/resume:
CleanShot 2024-04-02 at 17.09.25@2x
CleanShot 2024-04-02 at 17.09.25@2x
1002×1200 80.7 KB
Pause/resume is not currently available for exports using Stabilization.
4.2.2
Changelog from 4.2.1
Added 2x & 4x Upscaling to DaVinci Resolve OpenFX plugin.
Fixed preset issue with deinterlace 2x being overwritten to 1x.
Fixed problem where batch processing will only export one video.
Updated minimum trim length from 1sec to 6 frames of time.
Fixed inconsistent model labels in enhancement and second enhancement.
Fixed issue with ProRes reporting two Codec labels.
Fixed bug where opening preferences windows would stack infinitely.
Removed incompatible settings with mkv container.
Fixed interlace model resetting to Prob4
Fixed jittery output with Apollo Fast
Updated OpenVINO: compatible Intel devices will perform more reliably.
How to use 2x and 4x upscaling in DaVinci Resolve OFX Plugin:
Create a new timeline that matches the desired output resolution for your video.
Set the timeline to "center crop without resizing in the Timeline Settings panel.
Apply the Enhance effect to any clips that require upscaling to the Edit panel in Resolve. (The plugin will not show scale modes in a Color node.)
Select 2x or 4x scaling, an AI model, and a parameter mode.
If your desired output resolution is not exactly 200% or 400% the size of the input video, use Resolve’s transform tools to crop/fit the video to the timeline.
Known Issues
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
4.2.1
Changelog from 4.2.0
Added plugin installation selection to Windows .msi package.
Fixed batch processing issues with interlaced/mixed FPS inputs.
Fixed previews overwriting exports when using "Export As".
Fixed preview sets / views losing their selection across inputs.
Fixed MXF trim errors.
Fixed non-square PAR videos not fitting to preview window.
Fixed input-specific playback errors.
Fixed some cases of content scrolling beyond edge of containing window.
Added Export Error Resolver to macOS.
Filter states per view + selected view are now preserved across swapping inputs.
Resolve Plugin Updates:
Faster processing, reduced initialization time, resolved some memory access crashes.
Updated to Proteus v4
"Estimate" button fixed (macOS)
After Effects Plugin:
Plugin can now be installed to beta releases of After Effects.
Known Issues
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
4.2.0
Changelog from 4.1.2
First public beta plugin release for Davinci Resolve
Video AI is now available within Davinci Resolve as an OpenFX plugin. This first release includes all Enhancement models + Motion Deblur for both Windows and macOS.
The OpenFX plugin can be installed either during app installation (automatic on Windows and an optional checkbox on macOS) OR after app installation from the “Plugins” button in the app’s top menu bar.
image
The OpenFX plugin can be used in both the Edit workspace and as a Node within the Color workspace
image
image
1920×489 47.5 KB
Aion Frame Interpolation model
More accurate motion estimation for high-resolution inputs.
Reduced artifacting and ghosting when compared with Apollo and Chronos.
True 16x slow motion without duplicate frames.
Improved temporal consistency.
Can reduce tiling artifacts depending on available VRAM.
No TensorRT support in this beta release.
3D LUT option in Export panel
Process log footage directly in Video AI by selecting a .cube file to apply during exports
image
Low Disk Space Warning System
When queuing items for export, Video AI will now warn if the selected output disk is low on space. This threshold is fully customizable from the Preferences Advanced menu.
image
Other Changes & Fixes
Changed Previews/Exports sections to use tab buttons to toggle display.
Fixed bug where renamed FHD presets still have the previous preset as a duplicate.
Fixed bug where some codec names caused the export button to overlap.
Fixed bug where file names containing '#' were truncated in "Export As" dialog.
Fixed bug where FPS is doubled when reading interlaced video files and caused errors for Frame Interpolation.
Added options to append original filename and model names to image sequence export settings.
Fixed various rotation issues.
Fixed bug where cropping a rotated image sequence did not display the full frame.
Fixed bug where Auto Letterbox/Pillarbox crop was done incorrectly for rotated videos.
Fixed bug where switching inputs would reset rotation in preview views.
Fixed bug where input video thumbnails did not correspond to changed rotation.
Fixed bug where rotation metadata was not read when files were imported.
Known Issues
Resolve-specific
On first launch, you may need to manually enable the Video AI OpenFX plugin from Resolve's preferences menu:
image
image
1413×631 34.9 KB
TensorRT is not currently supported on NVIDIA cards.
It is recommended to enable "Show All Video Frames" in the Edit workspace when using the Enhance plugin.
image
Nyx performance is significantly slower compared to the in-app model.
Video AI app
Batch processing selection can overwrite FPS output.
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
v4.1.2
Released February 13th, 2024
Properly closes ongoing preview/export and makes sure background processes are killed.
Added toggle button for displaying remaining time/elapsed time for ongoing preview/export processes.
Added sharpen value to metadata on Nyx Fast exports.
Renamed HD presets to FHD (1080p)
Added three new QT codecs: QuickTime v210, r210, and QTRLE (Animation).
AV1 codec now allowed in .webm container.
Enabled a preference option in 'Advanced' for disabling the conversion from DPX->16bit TIFF..
Fixed some inaccurate labeling in UI (SAR-> PAR and Tooltip updates).
Fixed a bug where the remaining time of ongoing processes could show negative values.
Fixed bug with time remaining text for finished processes showing 0s.
Known Issues
Batch processing selection can overwrite FPS output.
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will intermittently be over-rotated.
Looping previews has a stutter.
Live preview can have frame de-sync.
Auto-parameter estimation moved to GPU for non-AVX2 CPUs.
v4.1.1
Released January 30th, 2024
Changelog from 4.1.0
Fix crashing on certain specific inputs.
Fix crashing on previewing for certain inputs.
Fix certain EXR/DPX inputs not converting to 16bit Tiff correctly.
Updates to OpenVINO to improve Intel ARC stability and performance
Applying frame interpolation related batch processing commands will correctly apply the FPS relative to the original.
Fix top part of window being hidden for some users.
Fix bug where auto params being lost when switching to Theia.
Fix bug with generating temp files for Exports.
Fix many preset related problems/bugs.
Add toggle functionality to switch between displaying the remaining time and the elapsed time for export processes.
v4.1.0
Released January 16th, 2024
Changelog from 4.0.9
Model Updates
Proteus v4
Better handles noise and compression artifacts.
Works on interlaced and interlaced progressive videos as well.
Improves faces better than Proteus v3.
Reduces ghosting of the previous frame at scene changes.
Better results on Auto.
More temporal consistency.
Faster than Proteus v3 in most cases.
Iris LQ v2
Improvements in detail, especially, for non-face inputs.
Improves temporal consistency (reduces 'marching ant' artifacts).
Reduces ghosting of the previous frame at scene changes (same as Proteus v4).
Reduces color distortion.
Adds TensorRT models for faster inference on certain NVIDIA machines.
Nyx Fast v1
Improvements to processing speed -- twice as fast as Nyx v2.
Reduced oversharpening and texture artifacts compared to Nyx v2.
Allows tweaking Sharpen slider value (Nyx Fast only).
Full changelog
AVX2 requirement removed and replaced with a warning about possible errors.
In a future update, auto-parameter estimation will be moved to the GPU to handle cases where Video AI is running on an incompatible CPU.
image
Rebuilt EXR + DPX support in-app
EXR and DPX inputs will now be correctly downsampled to 16-bit TIFF for processing, and exported at 32-bit linear color.
This function requires a large amount of available disk space on the drive selected for exports.
Support for ACES and other color formats is still underway, and our team is actively seeking different use cases for EXR and DPX to further improve compatibility.
Added focus fix feature to enhancement.
image
Improve missed focus in videos with the new "focus fix" mode for the Enhancement filter.
This feature changes the way models affect input videos by downscaling inputs before upscaling.
Two different modes can be used to control the strength of this effect:
Standard mode first downscales input media to 50% and then upscales to 2x.
Strong mode first downscales input media to 25% and then upscales to 4x.
image
image
1920×748 65.9 KB
Added Full Disk Access prompt to resolve disk access errors on macOS.
Improved some UI elements in filters panel.
Fixed incorrect frame interpolation related bugs when applying presets.
Fixed bug where SAR information was not retained across inputs.
Fixed bug where switching between Relative to Auto and Auto modes would not reset sliders.
Fixed bug where Nyx is resetting "Recover Original Detail" to 0.
Fixed window display related errors.
Fixed fullscreen UI issues.
Fixed bug where the app would hold a filesystem lock on input videos after app close.
Fixed bug where audio codec getting reset during trim change.
Fixed bug with metadata loss on MKV exports.
Fixed bug where Export Error Resolver tries to resolve preview related errors.
Known Issues
Batch processing selection can overwrite FPS output.
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will intermittently be over-rotated.
Looping previews has a stutter.
Live preview can have frame de-sync.
v4.0.9
Released January 2nd, 2024
Updated EULA and added it to help menu for viewing.
These changes only affect Enterprise users working with Video AI in certain venues such as movie theaters and premium streaming services.
Adobe After Effects plug-in now available for macOS.
Disabled audio automatically when there are no audio tracks in input.
Batch processing now preserves more from video input settings on export.
Setting priority on outputs now respects your priority decisions.
Fix rotation bug where it won't rotate around 180 degree axis of current selected rotation.
Batch processing crop will only apply to cropped input videos, not all the selected ones.
Fixed incorrect version labels in UI.
Added "Recover Original Detail" slider details to labeling.
Known Issues
Batch processing selection will overwrite individual video filter settings.
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will intermittently be over-rotated.
Looping previews has a stutter.
Live preview can have frame de-sync.
v4.0.8
Released December 19th, 2023
Added an Export Error Resolver that helps people automatically resolve export related errors.
When there is an export error, a dialog will pop up to assist in automatically resolving the error.
This will be specifically helpful for large batch processing related problems.
Added further support for batch processing.
Fix problems with non-square PAR related Image Sequence inputs.
Fix rotation not being respected for non-preview related views.
Fix many thumbnail related mismatches.
Fix timeline related bugginess surrounding Image Sequences.
Video Player now supports frame by frame seeking inside frame-interpolated previews.
Removed few-pixel-wide black overlay when in Comparison Mode.
Model Manager will now attempt download multiple times before reporting connection failure.
Known Issues
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will intermittently be over-rotated.
Slow-motion playback intermittently inaccurate across two slow-motion previews.
Looping previews has a stutter.
Live preview has some frame de-sync.
v4.0.7
Released December 12th, 2023
Changelog from 4.0.6
Model Manager GUI has now been added. Access with File -> Model Manager.
Reintroduced live preview to the new 4.x system.
Added a way to toggle Comparison Mode UI. You can now lock the left view to Original.
Improved frame desync related issues in player.
Improved 'Export' tab UI to show more details.
NLE Plugin Update
Added multi-frame rendering for After Effects.
Known Issues
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will not be rotated.
Slow-motion playback intermittently inaccurate across two previews with different slow-motion factors playing at the same time.
Looping previews and Play button has a stutter.
Live preview has some frame de-sync.
v4.0.6
Released December 5th, 2023
Changelog from 4.0.5
Updated the video player.
Improved speed/responsiveness of player in general workflow.
Video player no longer stutters through previews playback.
Video player no longer stutters when exiting previews.
Video player can handle larger preview load in timeline.
Known Issues
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will not be rotated.
Slow-motion playback intermittently inaccurate across two previews with different slow-motion factors playing at the same time.
Looping previews has a stutter.
Traded some speed on play-button-click-to-playing to improve speed on preview playback.
v4.0.5
Released November 28th, 2023
Changelog from 4.0.4
Fixes for default constant bitrate values in preferences.
Fix for “open in external player” failing.
Fix issue where entering custom values into text boxes doesn’t hold.
Simplify presets UI.
Improvements to frame de-duplication in frame interpolation.
Fixes frame count for interpolated output.
Fixes initial frames for enhancements.
AE Plugin Changelog
Added ‘Auto’ & ‘Relative to Auto’ support.
Known Issues
Inconsistent “Preview X frames” enable status.
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Stuttering on entrance-to/leaving previews / looping.
v4.0.4
Released November 14th, 2023
Crop is visually applied to the input and previews.
Adds description to preferences for Import/Export explaining that they will be applied the next time a video is opened.
Hides telecine option for image sequences (previously did nothing).
Removes confusing “Processing” label on exports that was appearing while they are waiting in queue.
Fixes zoom and upscaling issues with non-square SAR videos.
Fixes inconsitencies in crop tool.
Fixes issue in split view causing video to scale down in size while dragging pane handle.
Fixes issues editing previously applied crop.
Known Issues
Inconsistent “Preview X frames” enable status.
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Stuttering on entrance-to/leaving previews / looping.
v4.0.3
Released November 7th, 2023
Improved accuracy and consistency of frame interpolation.
Fixed issue that sometimes caused Nyx to use 2x upscaling instead of 1x.
Known Issues
Inconsistent “Preview X frames” enable status.
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Stuttering on entrance-to/leaving previews / looping.
v4.0.2
Released October 31st, 2023
Adds related preview ranges underneath preview group, including processing status and stats for each preview.
Adds looping of preview ranges by clicking the “Loop” button in the preview list.
Adds: jump directly to a preview in the timeline by clicking on it in the list.
Adds: mouse scroll wheel zooms timeline when hovered.
Adds: choose under “preview” button if you would like to process both views or just current (default current).
Fix: only show message about Nyx scaling when using old nyx-1 model.
Fix: do not reprocess previews starting at the same starting timeframe as an existing preview.
Fix: looping on in-out range.
Fix: “Auto”, “Relative” or “Manual” showing in preview info for enhancement models that don’t use this feature.
Fixes: various minor UI issues.
Known Issues
Overlapping preview ranges may have strange playback effects. This will be addressed in the next few releases.
Inconsistent “Preview X frames” range selection (example: selecting 10 frame preview may generate 9 frames)
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Some users may see a slight “stutter” during playback of rendered previews. This is file-dependent and does not affect output videos.
Cropped videos will show the full input dimensions during Preview, but the applied Crop setting will be applied to Exports.
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration.
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v4.0.1
Released October 24th, 2023
Restored ability to enter custom slow-motion values (like 2.5x).
Better UI/UX for Direct Comparison view – in-progress previews are now dimmed to show their processing status.
Fixed timeline zoom precision related issues.
Fixed UX in tutorial slideshow.
Fixed blank GPU options in preferences.
Known Issues
Inconsistent “Preview X frames” range selection (example: selecting 10 frame preview may generate 9 frames)
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Reprocessing overlapping preview areas.
Some users may see a slight “stutter” during playback of rendered previews. This is file-dependent and does not affect output videos.
Cropped videos will show the full input dimensions during Preview, but the applied Crop setting will be applied to Exports.
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration.
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v4.0.0
Released October 17th, 2023
Changelog from 3.5.4:
Comparison views: apply different filters to Views 1 & 2 to compare them. Freely switch between reviewing input clips and previews to find the best settings possible for your video.
Preview history: previews are tracked based on filter settings and shown grouped together on the video timeline.
Zoomable timeline: take advantage of timeline zoom controls to hone in on smaller areas of a large video.
In & Out Trim: replaced the “trim” edit mode with an In & Out option comparable to other video editing applications.
Refreshed UI and filter icons: tighter more informational display.
Nyx v2: improve results and 2x scaling mode.
Significant performance improvements for the in-app video player.
Crop controls moved to video input panel.
Known Issues:
Some users may see a slight “stutter” during playback of rendered previews. This is file-dependent and does not affect output videos.
Cropped videos will show the full input dimensions during Preview, but the applied Crop setting will be applied to Exports.
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration.
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v3.5.4
Released October 12th, 2023
Changelog from 3.5.3:
Fixes issue making it impossible to set custom resolution based on height or scale when locking aspect ratio.
Fixes issue with custom resolution presets based on height not being applied.
Known Issues:
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v3.5.3
Released October 10th, 2023
Changelog from 3.5.2:
Adds “Black and White” toggle to video input options.
This option will desaturate output to prevent models from adding color to Black & White videos.
Adds “Telecined” toggle to video input options.
This option will reverse the telecine process and deinterlace your video to 23.976 fps before applying AI models.
Adds detailed tooltips to Output Settings.
Enables ProRes, H264/5 and AV1 encoders for Linux.
Fixes many color-banding issues seen in the in-app player.
Fixes update issue related to frame interpolation / ffmpeg that some users experienced after updating to 3.5.1 and was not fully addressed by 3.5.2.
Fixes tooltips so they never go off-screen.
Fixes UI issues in video settings when information displayed is multiple lines.
Changes filter tooltips so they are now also seen when you hover on the filter name and not just the icon.
Changes “os” field in video-encoders.json to a regular expression (regex) evaluation to give more flexibility.
Known Issues:
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v3.5.2
Released October 3rd, 2023
Changelog from 3.5.1:
Adds Iris options to select between Medium/Low Quality. Medium will use Iris-v2, Low will use Iris-v1.
Adds 1x & 2x deinterlace options for Interlaced videos when keeping original fps.
Adds detailed tooltips for all preferences.
Adds info about Iris & Nyx models to Progressive models tooltip.
Adds detailed metadata for Iris & Nyx enhance models to exported video file.
Adds “Noise” value for enhancement models to video metadata when not zero.
Fixes ffmpeg issue causing exported video to sometimes get stuck in playback.
Fixes for rotation issues for files with rotation metadata. This was causing many of the issues like “file opened upside down.”
Fixes issues selecting “original” fps option in fps selector.
Fixes FPS display to drop extra "0"s when the fps is a whole number.
Fixes filter order when deinterlacing which should lead to higher quality visual results.
Fixes metadata for exported video when using Iris to only say “Deinterlaced” when the original video was actually Interlaced.
Iris LQ/MQ
After reviewing user feedback, our research team has decided to split Iris into two models depending on input quality. Iris v1 will now be called “Iris LQ” and is recommended for very low resolution videos and highly compressed content. Iris v2 is showing better performance on medium-resolution videos (more than 480, less than 1080) and will now be labelled “Iris MQ”.
We believe this is the best way to communicate the specific strengths of each model, and we appreciate all the feedback that came in after the initial release!
Known Issues:
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.5.1
Released September 26th, 2023
Fixes the frame interpolation models repeating frames when changing FPS.
Reduces model loading time for Nyx, Proteus, Iris Auto and Relative to Auto.
Fixes stabilization not running on lower memory machines.
Fixes crashes related to image sequence fps change.
Adds warning when rotating after setting crop dimensions.
Adds 10-bit for NVENC AV1.
Automatic parameter estimation now runs on 8 frames instead of 20.
Fixes for Nyx as second enhancement: will now automatically make the first enhancement perform upscaling.
Iris-v2 is supported on Intel ARC.
Known Issues:
Intel ARC is currently unsupported.
For Interlaced and Interlaced-Progressive videos, setting FPS back to Original may cause issues. To get around this, re-import the file. This only happens after manually changing the FPS.
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Some vertical videos will be rotated an additional 90 degrees when imported
v3.5.0
Released September 19th, 2023
Adds “Nyx” Enhancement model.
Adds Iris v2 model.
Adds support for rotating videos 90, 180, and 270 degrees.
Adds crash logs to “logs for support” and changes format to zip file.
Adds change moving “Enable Previous Model Versions” preference to advanced section of settings.
Adds change renaming “Second Pass Enhancement” to just “Second Enhancement” to reduce any potential confusion with two-pass encoding.
Fixes bug causing logs to be inaccessible on Windows.
Fixes special characters in file path causing export failure when using Export As.
Fixes crash on app open for some machines.
Fixes trim accuracy issues for ProRes.
Fixes bug causing inaccuracy with video duration after resetting trim range.
Nyx:
A new model for denoising high-quality video.
The focus of the model is on footage from high-quality cameras captured in less-than-ideal lighting conditions.
Processes videos at 1x scale to maintain the level of detail in the input video while removing noise.
Iris v2:
An improved version of Iris v1, offerring more details and sharpness especially to faces.
It uses TensorRT models for Nvidia GPUs and could result in a 50% speed-up.
It also reduces the “previous frame’s ghost” artifacts.
It reduces “marching ant” (diagonal movement of texture) artifacts, especially, for 1x and 2x.
Attention:
The format for support logs is being changed to include license information (auth.tpz). While this file is encrypted and only useful on your machine, please try to avoid posting it or your support logs here by mistake. This would not likely lead to a problem, but better to be safe than sorry. Please submit any logs directly to our support team or devs
Known Issues:
Intel ARC is currently unsupported.
If using Frame Interpolation to increase fps over 2x, output may contain duplicate frames.
For Interlaced and Interlaced-Progressive videos, setting FPS back to Original may cause issues. To get around this, re-import the file. This only happens after manually changing the FPS.
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.4.4
Released September 12th, 2023
Adds AV1 encoder for AMD 7000 series and 700 iGPUs.
Adds option to reset all preferences to defaults.
Adds new “Advanced” tab section to preferences. More items will be added to this in the future.
Adds change to copy all subtitle tracks for MKV exports. Support for more containers will be added in the future.
Adds improvements to file size for dynamic high setting without sacrificing visual quality (targets 95 VMAF).
Adds fix vastly improving responsiveness for live preview for MKV & WebM formats.
Changes minimum preview duration to 0.5 seconds, which fixes some bugs that could happen with <10 fps videos.
Fixes NVIDIA Tesla GPUs not showing encoder options.
Fixes videos encoded with H265 from AMD GPUs not playing on iOS and macOS.
Fixes output bitrate appearing on an output item when using a codec that does not use bitrate (ex. ProRes).
Fixes bug making it impossible to seek by typing frame number after trimming.
Fixes target bitrate in preferences displaying incorrect number.
Known Issues:
Intel ARC is currently unsupported due to Intel driver issues.
For Interlaced and Interlaced-Progressive videos, setting FPS back to Original may cause issues. To get around this, re-import the file. This only happens after manually changing the FPS.
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.4.3
Released September 5th, 2023
Adds MOV container option for FFV1 encoder.
Adds dynamic bitrate options for Apple Silicon computers.
Adds “Import” section to preferences and moves “Default Import File Type” to this section.
Adds preference for default FPS for image sequence import. Changing fps on an image sequence that has been imported will no longer be “sticky” - set your default here instead.
Fix for aspect ratio input to reset to Custom after applying reset or auto crop.
Known Issues:
Intel ARC is currently unsupported.
For Interlaced and Interlaced-Progressive videos, setting FPS back to Original may cause issues. To get around this, re-import the file. This only happens after manually changing the FPS.
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.4.2
Released August 29th, 2023
Fixes many issues related to in-app player seeking and trim related precision.
Fixes for deinterlaced video playback and processing feedback.
Adds FFV1 encoder options.
Adds option to preferences to default to searching for image sequences during import.
Adds “Advanced” section to application settings and moved “Show Second Enhance Control” here.
Removes “Show Intermediate Resolution Control” - this will always be shown when Second Enhance is enabled.
Fixes bug that could cause processing failures on first app launch after updating.
Fixes for Preview button - enabled for all valid trim ranges.
Fixes to Export As to respect the selected folder in preferences.
Fixes issue causing one additional frame to be processed during preview
v3.4.1
Released August 22nd, 2023
Added second enhancement
Added new user guidance
Added preference to turn off new user guidance
Video player playback and trim functionality is more reliable and accurate
Improved dynamic bitrate usability
Improved tooltips
v3.4.0
Released August 15th, 2023
Added Dynamic Compression Options (H264 and H265).
Added ability to create presets from anywhere.
Added new user experience tips (with preference to disable all).
Added “refresh login” button to update screen that will allow you to update your license by signing in again.
Added quality Improvement for first few frames while using enhancement models (except Gaia).
Added Iris to benchmark tool.
Added speed improvements for Iris on macOS Ventura (1x and 2x upscale).
Added capability to run Apollo on M1/M2 with HD or above input videos.
Added processing change: Minimum number of frames needed is now 4 instead of 5.
Fixes first few frames repeating in frame interpolation.
Fixes login to work with proxy settings.
Fixes processing on Intel CPU machines (12th, 13th gen). No longer slows down when app is minimized or out of focus.
Fixes image sequence issues when non ASCII characters are in the path.
Fixes ANE process using up all Memory on Mac.
Fixes for Tesla GPUs. Now show NVENC H264 and H265 encoders.
Fixes for log files. Can now be opened when app is open (Ctrl+G, Cmd+G).
Potential Fixes
Processing failures when exporting large number of frames.
Intel Arc/iGPU processing issues.
v3.3.10
Released August 8th, 2023
Changes from v3.3.9:
Fixed Color Correction Parameters
Fixed Preferences UI Bug
Added more Tooltips
Improved Logging
Improved Recover Original Details Tooltip
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Seeking/preview may not work well with DPX inputs
v3.3.9
Released August 1st, 2023
Reworked custom output resolution workflow to better support presets
Added tooltips for video output settings panel
Fixed preset migration for video output settings
Fixing Theia to Interlaced Enhancement UI bug
Fixing Proteus and other sliders text box bug
Added tooltip for automatic pillarbox/letterbox removal in crop tool
Fixed textbox focus for output resolution
Fixed Interlaced Iris naming
Fixed integer overflow issue for start frame number
Changing toggle box to checkbox in UI
Added version labels for Proteus and Theia
Exports / Previews now show filename when hovering over thumbnail
Fixed video setting handling of SAR values for custom resolution
Fixed “proteus” showing up in FFmpeg command and added additional preset migration safeguard
Updated the live preview checkbox in preferences
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Seeking/preview may not work well with DPX inputs
Topaz Video AI v3.3.8
Released July 26th, 2023
Changes from v3.3.7:
Fixed image sequence previews overwriting or deleting previously completed exports
Topaz Video AI v3.3.7
Released July 26th, 2023
Changes from 3.3.6:
Iris Interlaced now has Recover Original Detail slider
Topaz Video AI v3.3.6
Changelog from 3.3.5:
Reworked Preset creation, modification, delete, and apply user workflows.
Reworked custom output resolution flow.
Removed (Modified) text from presets.
Fixed video scan type changing bug when switching presets.
Fixed Preview switching Iris.
Fixed auto crop stabilization preset issue.
Fixed Proteus defaulting to V1 instead of V3 in some cases.
Output FPS combo box does not duplicate the selection.
Enhancement filter will now stay enabled for custom output resolution.
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.3.5
Released July 18th, 2023
Changes from v3.3.4
Fixed invalid state for start frame number when passed an empty string
Fixed Start Frame Number getting invalid number when deleting value to modify it on Windows
Fixed values close to max available for Start Frame Number getting random or negative values
Start Frame Number & Same as Input locations have been swapped
Fixed Start Frame Number not resetting to original when selecting “Same as Input” in Preferences
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Export settings from old presets might not be applied correctly. Try creating a new preset if this happens to you.
Deleting custom Start Frame Number value may place that setting into an invalid state on Windows. As a workaround, select the value and modify it
v3.3.4
Released July 11th, 2023
Changes from v3.3.3
Adds option to export settings allowing a custom start frame number to be set for image sequence exports.
Adds a new preference for default start frame number when exporting an image sequence and not matching the input (ex. if you want first frame to be 0000.png vs. 0001.png or 0042.png)
Adds “.3gp” as an available file type when importing videos through the file opener.
Fixes bug causing image sequences to be re-exported with start frame off by one when matching start frame from import.
Fixes color space issues when importing image sequences and exporting as video. Color space will be converted to BT709 when input color space is RGB & exporting to YUV pixel format.
Fixes invalid output when exporting PNG or DPX to VP9
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Export settings from old presets might not be applied correctly. Try creating a new preset if this happens to you.
Deleting custom Start Frame Number value may place that setting into an invalid state on Windows. As a workaround, select the value and modify it
v3.3.3
Released July 4th, 2023
Changes from v3.3.2
Adds DPX as import option for Image Sequences.
Adds - when export is pressed, an output item will always be added even if there is an error. This should improve batch processing workflows in case of errors.
Adds more clear options to export preferences for “live preview” behavior. This replaces the “video export” option. Behavior will not change from your previous setting for this option.
Adds detailed tooltips in-app to explain “live preview” options.
Adds toggle for “live preview” to export settings to easily turn on/off live preview for exports.
Adds error handling when trying to copy TrueHD audio to MP4/MOV container.
Fixes “Encoder Profile” option in export preferences not showing your saved option when opening preferences.
Fixes “Audio Codec” list in export preferences to be filtered by chosen “Extension” option.
Fixes “videoai” metadata tag missing on exported videos.
Fixes error copying AAC audio with ADTS headers from M2TS video source.
Fixes “back to input” - will always return you to the input video corresponding to the current output video.
Known Issues:
The output settings of old presets may not be applied. Please modify them if needed
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.3.2
Released June 27th, 2023
Changes from v3.3.1
Fixes frames missing from end of video when exporting with ProRes.
Fixes app sometimes trimming off one frame when exporting full video.
Improves playback speed for output
Adds ProRes Proxy profile for exports
Fixes Iris artifacts on M2
Disables audio for image sequence inputs
Syncs license with server more often if connected to internet
Misc updates
Known Issues:
The output settings of old presets may not be applied. Please modify them if needed
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.3.1
Released June 20th, 2023
Changes from v3.3.0
New export settings UI and options (details below).
Large list of “encoders” has been replaced with separate selectors for encoders – ProRes, H264, H265, VP9, AV1 – an associated profile/codec, and a GPU option if applicable (ex. you have both a NVIDIA and AMD GPU mounted).
Image Sequence options are given through a tab option now to separate UX flows for Video and Image Sequence exports. Similar to video encoders, we have added Bit Depth options for Image Sequence exports.
Removed “Auto” option for Audio Mode, default behavior is “Copy”. The UI should show an error if this is not possible for your video for compatibility reasons. (ex. PCM audio cannot fit in MP4).
Changed “Transcode” audio mode name to “Convert”.
When Converting audio, you have the option between AAC, AC3, PCM codecs. For MP4 container, you will only be presented with AAC & AC3. When exporting with VP9 + WebM container, the only codec option is Vorbis.
For AAC & AC3 you have the option to set a bitrate for the audio. AAC default is 320k, AC3 is 448k
Saved preferences from previous versions should be migrated to new preferences. ex. if you have a preference set to export as “TIFF lossless 8-bit”, your preferences in the new version should show “TIFF” for encoder and “8-bit” for Bit Depth and this should still be applied when opening a new video.
Added MKV container to extension options for VP9 exports.
Known Issues:
The output settings of old presets may not be applied. Please modify them if needed
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.3.0
Released June 13th, 2023
Changes from v3.2.9
Adds a new enhancement model Iris for face enhancement and/or improving low-to-medium quality progressive/interlaced videos
Adds Recover Original Detail option for Proteus, Iris, and Artemis models
Adds DPX support
Maps all audio streams with metadata to output
Fixes issue with estimate not working for certain image sequences
Misc bug fixes and improvements
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Submit files here: Dropbox 35
Please take a look at the Video Roadmap Update 734 for more context on what we’re focusing on right now. Thanks in advance for your feedback!
Enhancement model - Iris
Primary aim of the model is to improve human faces
This model is also designed to enhance low- to medium-quality video, particularly, with noise and compression artifacts. Note, the model tends to lose details on high-quality videos
This model is trained to work on all three categories - progressive, interlaced, interlaced progressive
Please look at some examples in the roadmap post 734
Recover Original Detail
Purpose of this slider is to retain details from original source if the model’s output is overly smoothed
It adds the source texture details to the model’s output to bring back the original details
Hi Everyone,
We have some more exciting updates this week! And to our AMD and Mac friends, this is a good one!
Improvements and Fixes
Features
Precise 2.5 update for AMD & Mac!
SDR-HDR Model: Hyperion 2 for AMD & Mac!
Topaz Plugin: Nuke
Additional Updates
Crop X & Y Update
Bug Fixes
Timeline issues with Image Sequences
Project loading fixes
and more!
Precise 2.5 update for AMD & Mac!
image
image
1927×1084 439 KB
image
image
1795×1183 392 KB
Starlight Precise 2.5 has made one heck of a splash! And like the title suggests, AMD and Mac are now supported for this model! And yes, this model available for everyone that meets the requirements to run it!
Model Recap
Starlight Precise 2.5 is the 5th variant to the desktop Starlight Model Family. Note, that previous iterations of this model exists on Astra. This model has been specialized for realistic human face and skin textures in addition to other areas of improvement such as material, texture, and text.
This model has become so popular, that you can now find Starlight Precise 2.5 model with it’s own icon for faster model access.
image
image
520×238 18 KB
About
This model works for Medium to High Quality content whether this is generative, live action or restoration videos.
Improves generated content that may have soft, plastic, or slight artificial looks, however the AI generated content should have solid structure and composition for ideal improvements.
There is a slight difference between the local version and the cloud version in order to fit on the desktop.
Greatly reduces flickering with smaller textures.
Recommended to use video inputs at or under 4K.
Outputs up to 4K.
Model access via the cloud for all subscription models.
Cloud Export is supported on both Mac and Windows.
Expect long render times for this diffusion model. Appears to have faster processing compared to Mini and HQ.
Please note that Starlight Precise 2.5 model had to be slightly modified to fit the MAC hardware requirements. Mac users may experience slower renders, however outputs have maintained results that are reasonably similar to Nvidia with great quality results.
This Model Is Less Ideal For
Shots with heavy motion blur may be rendered as texture.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Not ideal for content below 480p
image
image
2048×739 348 KB
Technical Things To Know
Now available for AMD, Mac & Nvidia
Available for all licenses to run locally! Yes, you read this right!
Nvidia Requirements:
Requires an absolute minimum of 12GB VRAM with 32GB CPU RAM, though 16GB VRAM is recommended.
Targeted for GPU compute compatibility of 8.9. This is also the recommendation.
Please note, though there might be some success with a CC of 8.6 cards, this is not
This model is included locally, using neuroserver.
AMD Requirements:
Requires an absolute minimum of 12GB VRAM with 32GB RAM, though 16GB VRAM is recommended.
Mac Requirements:
Requires an absolute minimum of 16GB though 24GB is recommended.
All Apple Silicon should be supported.
image
image
1869×1447 352 KB
SDR-HDR Model: Hyperion 2 for AMD & Mac!
image
image
1671×1020 377 KB
Hyperion 2 is finally available for AMD and Mac for local rendering! Please note the individual requirements needed pending your machine’s hardware below.
Model Recap
Hyperion 2, is a great update to our previous iteration. This version of the model expands your color and nit value range even further! This is fantastic for wider spectrum manipulation for VFX Compositing and for modern viewing on HDR supportive displays!
image
image
2868×1615 555 KB
About
Converts 8-bit GenAI or archival video into HDR, high bit-depth media suitable for professional finishing pipelines.
Expands tonal range and color precision to provide significantly greater grading latitude in highlights, shadows, and midtones.
Prevents highlight and shadow clamping often seen in standard range and native GenAI video sources.
Expands existing color opportunities/steps for professional compositing and color manipulation workflows.
The additional color opportunities greatly reduces visible banding and compression artifacts common in 8 bit renders.
Has great details with text and textures even during high motion blur areas.
Export ProRes video or EXR (Pro) image sequences.
Support ACES color output with EXR (Pro) renders for better integration into industry-standard film and television pipelines.
This works best for medium to high quality content.
This model has some built in enhancements to revitalize and improve content.
This version is an easy “one click” solution. There are no parameters to adjust.
Max input and output resolution for this model is 3840x3840.
image
image
1920×1432 1.58 MB
Technical things to know
Nvidia Requirements:
Requires 16GB VRAM with 32GB CPU RAM, though 24GB VRAM is recommended.
Targeted for GPU compute compatibility of 8.9. This is also the recommendation.
Please note, though there might be some success with a CC of 8.6 cards, this is not technically supported and some may encounter a less ideal experience.
AMD Requirements:
Requires an absolute minimum of 12GB VRAM with 32GB CPU RAM, though 24GB VRAM is recommended.
Mac Requirements:
Requires an absolute minimum of 16GB RAM though 24GB is recommended.
All Apple Silicon should be supported.
image
image
2464×1831 679 KB
Topaz Plugin: Nuke!
The very long awaited plugin is finally here! The Topaz plugin for Nuke is now a reality! As a Compositor or VFX artists, one is now able to take advantage of some of Topaz’ famous models right within the the Nuke working environment.
Upon installation, the OFX file will be available to be downloaded. If there are any issues with the initial installation. The plugin is available within the Topaz Video application in the upper left side under the “Plugin” drop down.
Within Nuke, there are two places to access the Topaz plugin.
image
image
799×799 24.7 KB
Via the left panel of icons, select the Topaz plugin icon to reveal the four plugin options.
image
image
1054×418 64.5 KB
Via the “Tab”, search for
Enhance [Topaz Labs]
Frame Interpolation [Topaz Labs]
Interlaced Enhance [Topaz Labs]
Motion Deblur [Topaz Labs]
Or with the “Tab” search, search instead for “[topaz” to reveal all the Topaz plugins.
image
image
1084×304 40.8 KB
Have fun experimenting with the different model options with your footage!
Note: Having previous Alpha, Beta, or Release versions of the Topaz OFX plugin alongside the current release can lead to crashes in Nuke. Please ensure that only the latest Topaz Video bundle is installed in your OFX directory, which can be found at C:\Program Files\Common Files\OFX\Plugins on Windows or /Library/OFX/Plugins on MacOS.
Additional Updates to Note!
Crop X & Y
The X & Y transforms are now back for Crop. This gives one the ability to input specific placement coordinates for the upper left corner of the crop shape. This provides accurate pixel values that you may need for your crop.
image
image
3099×1898 698 KB
Bug Fixes
Fixed image sequence in/out ranges appearing incorrectly on the timeline when updating fps.
Fixed project load failing when sources are moved or deleted.
Topaz Video 1.5.0
Hi Everyone,
We have another really cool update this week! This includes a new model, a panel and some bug fixes.
Improvements and Fixes
Features
New SDR-HDR Model: Hyperion 2!
New Topaz Panel for Premiere!
Additional Updates
Starlight Precise 2.5 Updates
Bug Fixes
Fixed Starlight Precise 2.5 last frame drop
Addressed Neuroserver image sequence failures
Refined Starlight Precise cloud resolution issues
Fixed broken feedback link
and more!
New SDR-HDR Model: Hyperion 2!!
image
image
1920×1083 983 KB
Hyperion 2, is a great update to our previous iteration. This version of the model now expands your color and nit value range even further! This is fantastic for wider spectrum manipulation for VFX Compositing and for modern viewing on HDR supportive displays!
About
Converts 8-bit GenAI or archival video into HDR, high bit-depth media suitable for professional finishing pipelines.
Expands tonal range and color precision to provide significantly greater grading latitude in highlights, shadows, and midtones.
Prevents highlight and shadow clamping often seen in standard range and native GenAI video sources.
Expands existing color opportunities/steps for professional compositing and color manipulation workflows.
The additional color opportunities greatly reduces visible banding and compression artifacts common in 8 bit renders.
Has great details with text and textures even during high motion blur areas.
Export ProRes video or EXR (Pro) image sequences.
Support ACES color output with EXR (Pro) renders for better integration into industry-standard film and television pipelines.
This works best for medium to high quality content.
This model has some built in enhancements to revitalize and improve content.
This version is an easy “one click” solution. There are no parameters to adjust.
Max input and output resolution for this model is 3840x3840.
image
image
1920×1432 1.58 MB
Technical things to know
Nvidia Requirements:
Currently Nvidia Only
Requires 16GB VRAM though 24GB VRAM is recommended.
Targeted for GPU compute compatibility of 8.9. This is also the recommendation.
Please note, though there might be some success with a CC of 8.6 cards, this is not technically supported and some may encounter a less ideal experience.
Colorspace Outputs
ProRes Primaries: BT.2020 PQ
ProRes Transfer Function: BT.709
EXR (Pro only) Image Sequence: Aces 2065-1
Things to Consider
SDR to HDR does not operate well with interlaced footage. If you do have interlaced footage, it’s recommended that you deinterlace first before applying this tool.
Though this model has some enhancement benefits additional enhancements with this SDR to HDR Model can be a great combo to reduce issues from source such as compression and artifacts for better results.
Extreme cases of video with very muddy/dark imagery or very bright/blown out content, may produce limited results with the limited source range to build upon.
An HDR display will be needed in order to view the results properly. Otherwise results will look washed out.
Known Issues
Cloud Export is not yet supported on both Mac and Windows. Coming soon!
Limited codec options for local rendering. Looking to expand the current options!
New Topaz Panel for Premiere!
We have built a UXP panel for Adobe Premiere users! Topaz Labs for Premiere is new utility that introduces a “send to cloud” workflow to batch upscale video clips and images directly from the Premiere.
Simply drag video or image clips from your bin or timeline into the panel, choose an AI enhancement model, and upscale to HD, 4K, or even 8K. Better yet, this process does not take up system resources, allowing you to continue editing. Rendered files are automatically imported to your project and placed in timeline.
Topaz Video 1.4.0
Hi Everyone,
We have a short but extra sweet release!
The team’s been busy and working hard to get this latest update for you!
I’m pleased to announce our most powerful model release in a while! Starlight Precise 2.5 for local desktop rendering!
Improvements and Fixes
Features
Starlight Fast Precise 2.5 model
Neuroserver
Starlight Precise 2.5 Local
image
image
1920×1076 1.39 MB
image
image
1869×1447 473 KB
Starlight Precise 2.5 made its debut in our Cloud app, Astra, with much excitement and positivity. In the meantime, we were also working hard to get this wonderful model for local desktop rendering! I’m happy to announce that the team has just done that! The best part, we’re making this model available for everyone that meets the requirements to run it! This makes Starlight Precise 2.5 the 5th variant to the desktop Starlight Model Family. Note, that previous iterations of this model exists on Astra. This model has been specialized for realistic human face and skin textures in addition to other areas of improvement such as material, texture, and text.
image
image
857×1659 131 KB
About:
This model works for Medium to High Quality content whether this is generative, live action or restoration videos.
Improves generated content that may have soft, plastic, or slight artificial looks, however the AI generated content should have solid structure and composition for ideal improvements.
There is a slight difference between the local version and the cloud version in order to fit on the desktop.
Greatly reduces flickering with smaller textures.
Recommended to use video inputs at or under 4K.
Outputs up to 4K.
Model access via the cloud for all subscription models.
Cloud Export is supported on both Mac and Windows.
Expect long render times for this diffusion model. Appears to have faster processing compared to Mini and HQ.
image
image
3826×1370 737 KB
Technical things to know:
Nvidia Only
Available for all licenses to run locally! Yes, you read this right!
Requires a minimum of 12GB VRAM though 16GB VRAM is recommended.
Targeted for GPU compute compatibility of 8.9. This is also the recommendation.
This model is included locally, using neuroserver.
image
image
3849×1384 1000 KB
This model is less ideal for content with:
Shots with heavy motion blur may be rendered as texture.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Not ideal for content below 480p
image
image
1920×656 1 MB
image
image
3129×1371 815 KB
Additional Updates to Note!
Neuroserver
Neuroserver installation happens through the installer with a checkbox. An in-app update is recommended.
image
image
761×564 17.4 KB
All diffusion models (Starlight Family) now require Neuroserver: Starlight Mini, Sharp, Fast 2, and HQ.
image
image
1120×105 20.2 KB
Quality for the Starlight Models is expected to be the same.
If for some reason the Neuroserver is not working as expected, reinstall the app to run the “Repair” mode.
Topaz Video 1.3.1
Hi Everyone!
We have an update that addresses a few things from the 1.3.0 release.
Improvements and Fixes
Features
Starlight Fast 2 model -Founders access update
Starlight HQ model -Founders access update
Bug fixes:
AMD/Mac Starlight Mini issue fix
Fixed “Update Installation” pop up issues
Fixed compatibility check for Starlight Fast 2
Starlight Fast 2 Local
image
image
1920×1326 1.23 MB
Updated for Pro and Founders.
Recap about the model:
Introducing the new Starlight Fast 2, for local desktop rendering! Starlight Fast 2 is the third variant to the desktop Starlight Model Family. Note that Starlight Fast exists on Astra. This model is great to get good diffusion results at a much faster processing time. Note that in order to make this model both available locally and still be speedy (for a diffusion model) there is a slight difference in quality for local renders.
image
image
769×1228 96.6 KB
Things to know:
Nvidia Only
Requires a Pro or Founders license to run locally.
Best results come from 2x and above upscaling.
Great option for faster local diffusion based rendering.
Recommended to use video inputs at or under 4K.
Outputs between HD - 4K.
This model works for Medium to High Quality content whether this is generative or live action plates/videos
speed can be 3-5x faster than Starlight Mini pending on your GPU and available VRAM.
Provides better treatment for flickering/stuttering frames (especially when camera moves)
Requires a minimum of 16GB VRAM though 32GB VRAM is recommended.
There is a slight quality difference between the renders with 16GB vs 32GB VRAM. The latter being a tad more detailed.
Model access via the cloud for all subscription models.
Cloud Export supported
image
image
1920×1594 2 MB
Starlight HQ Local
image
image
1920×595 804 KB
Updated for Pro and Founders.
Recap about the model:
Introducing Starlight HQ, for local desktop rendering! Starlight HQ is the fourth variant to the desktop Starlight Model Family. This diffusion model has a great balance of detail and sharpness while keeping the intent of the original video source. Provides higher quality outputs for a wide range of videos, though not necessarily intended for archival content it can still provide good results. Though addressing similar intentions as Starlight Fast2, outputs have generally been higher quality and consistent with this model’s longer processing time.
image
image
630×1012 110 KB
Things to know:
Nvidia Only
Requires a Pro or Founders license to run locally.
Processing speed is very similar to Starlight Mini.
Supports 480p to 4k inputs.
Outputs between HD - 4K.
This model works for Medium to High Quality content whether this is generative or live action plates/videos
Requires a minimum of 10GB VRAM though 12GB VRAM and up is recommended.
Model access via the cloud for all subscription models.
Cloud Export supported
image
image
3197×1093 1.2 MB
This model is less ideal for content with:
haloing issues common with vintage digital videos
hair quality can vary in detail when content is on the lower medium spectrum.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Small repeated patterns can flicker
Bug fixes
AMD/Mac Starlight Mini issue with notification of “export disabled …requiring model download before local export.”
Fixed “Install now” update installation from app closing.
Fixed the cancel, close or install buttons (from the “Update Installation” floating window) from being unresponsive when there is active exports.
Fixed compatibility check for GPU minimum values to prevent failures for machines that can not run Starlight Fast 2.
Topaz Video 1.3.0
Hi Everyone,
We have a pretty sizable release!
The team’s been busy and we now have several exciting features for you!
Behind the scenes, the Starlight Mini model (being the first diffusion model for the local desktop experience) has a unique architecture that’s unlike any of our other models. We essentially brought the best diffusion/restoration model to fit in the local app. Diffusion models are computational beasts and are very slow when processing. Since it’s original announcement, we have been trying to find way to make Starlight Mini faster while keeping the quality the same. However, over the months, this has been challenging. Ultimately, we do not want to sacrifice quality nor the type of treatment you get from starlight mini in order to get faster outputs. Though, we will continue to work and refine the Starlight Mini experience, we have brought two new exciting Starlight family models locally.
For a more speedy diffusion render experience we have Starlight Fast 2.
Those interested in a model that offers higher quality renders for inputs up to 4K, we now have Starlight HQ.
Additionally, I’m pleased to announce Gaia 2, a new animation model that I’m sure you’ll enjoy!
Features
[PRO ONLY] Starlight Fast 2 model
[PRO ONLY] Starlight HQ model
Gaia 2, the new Animation model
New project settings dialog
Telecine migration
New Transforms + Crop area
Horizontal + Vertical flip
Bug fixes:
Non-Square PAR Issue Fix
MXF Encoder Restrictions Fix
Image Sequence Timeline Refinement
and more!
Pro: Starlight Fast 2 Local
image
image
1920×1326 1.23 MB
Introducing the new Starlight Fast 2, for local desktop rendering! Starlight Fast 2 is the third variant to the desktop Starlight Model Family. Note that Starlight Fast exists on Astra. This model is great to get good diffusion results at a much faster processing time. Note that in order to make this model both available locally and still be speedy (for a diffusion model) there is a slight difference in quality for local renders.
image
image
769×1228 96.6 KB
Things to know:
Nvidia Only
Requires a Pro license to run locally.
Best results come from 2x and above upscaling.
Great option for faster local diffusion based rendering.
Recommended to use video inputs at or under 4K.
Outputs between HD - 4K.
This model works for Medium to High Quality content whether this is generative or live action plates/videos
speed can be 3-5x faster than Starlight Mini pending on your GPU and available VRAM.
Provides better treatment for flickering/stuttering frames (especially when camera moves)
Requires a minimum of 16GB VRAM though 32GB VRAM is recommended.
There is a slight quality difference between the renders with 16GB vs 32GB VRAM. The latter being a tad more detailed.
Model access via the cloud for all subscription models.
Cloud Export supported
image
image
1920×1594 2 MB
This model is less ideal for content with:
haloing issues common with vintage digital videos
hair quality can vary in detail when content is on the lower medium spectrum.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Small repeated patterns can flicker
Pro: Starlight HQ Local
image
image
1920×595 804 KB
Introducing Starlight HQ, for local desktop rendering! Starlight HQ is the fourth variant to the desktop Starlight Model Family. This diffusion model has a great balance of detail and sharpness while keeping the intent of the original video source. Provides higher quality outputs for a wide range of videos, though not necessarily intended for archival content it can still provide good results. Though addressing similar intentions as Starlight Fast2, outputs have generally been higher quality and consistent with this model’s longer processing time.
image
image
630×1012 110 KB
Things to know:
Nvidia Only
Requires a Pro license to run locally.
Processing speed is very similar to Starlight Mini.
Supports 480p to 4k inputs.
Outputs between HD - 4K.
This model works for Medium to High Quality content whether this is generative or live action plates/videos
Requires a minimum of 10GB VRAM though 12GB VRAM and up is recommended.
Model access via the cloud for all subscription models.
Cloud Export supported
image
image
3197×1093 1.2 MB
This model is less ideal for content with:
haloing issues common with vintage digital videos
hair quality can vary in detail when content is on the lower medium spectrum.
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
Small repeated patterns can flicker
Gaia 2
image
image
1920×758 853 KB
It’s about darn time that we got a new animation model, Gaia 2! This model works on a frame by frame basis, similar to an image model, which is perfect for traditional types of animation. The unique method is why this model is in it’s own category. Continuing with the Gaia initiative this model has a great handle on animation. This model addresses aliased edges, compression and sharpens without “over doing it”. This is not a restoration model, so this will not remove film hairs and scratches or generate details. As a conservative model, it aims to preserve the original source intent.
Here are the types of content that this model best performs with:
Traditional Animation
Motion Graphics
Stop Motion Animation
Note that the model is not limited to the content above, so experiment and have fun with this one!
image
image
580×588 37.4 KB
Things to know:
Available for both PC and Mac
Processing speed can be faster or equivalent to Proteus!
Supports several input resolutions.
Support a wide range of output resolutions.
Best results come from 2x and above upscaling.
No modifiable settings.
Stack this model by adding a second enhancement with another Gaia2 to get sharper results.
Cloud Export supported
image
image
1920×642 276 KB
This model is less ideal for content with:
This model works for Medium to High Quality content and it is worth noting that it is not a restoration model.
Haloing issues common with vintage digital videos
Not intended for very low/poor condition quality media.
Interlaced content needs to be de-interlaced first.
If there’s heavy grain you may want to denoise first. Some grain particulates may be seen as part of the texture. Pending personal preference.
Project Settings Dialog
image
image
1446×1320 120 KB
The Project Settings Dialog can be found from the File menu item and replaces old input settings panel. Some of the options for settings include project-wide input colorspace, black + white, timecode options, and image sequence FPS.
image
image
2311×973 171 KB
Telecine Migration
image
image
513×1048 53.9 KB
Telecine has been moved to the right panel, under “Interlaced” enhancement for more visibility. Moving this under a category that makes a little more sense will help one find this option faster rather than remembering to look into the settings menu.
Transforms + Crop area
image
image
639×321 14.2 KB
We now have a new area that includes video Transform options and Crop in the same location. This area is identified with the same crop symbol with rotations.
Horizontal + Vertical flip
image
image
1375×763 74.4 KB
Sometimes capturing footage can get all flipped around and you may want a quick way to fix this. You now have the ability to flip your footage vertically or horizontally.
image
image
1075×1876 1.16 MB
Additional Updates to Note!
Improved Performance
Improved some VFR and dropped framed issues.
Bug fixes
Fixed issues regarding previewing, cropping, and rotating videos with non-square pixel aspect ratios. Please let us know if you continue to run into issues with specific formats.
Fixed MXF render failures with DNxHR codec.
Fixed timeline related issues with image sequence formats.
Added re-download option for cloud exports.
Fixed links on the “About Topaz Video” pop up menu
NeuroStream
As some of you have heard the exciting news that we are implementing NeuroStream with Topaz Photo. This is something we are looking into for Video as well. However, while this is being worked on and tested, more details will be provided at a later and more appropriate time. Thank you!
Topaz Video 1.2.1
Changes to authentication
We are excited to announce that we have reworked authentication to make using the app a more pleasant experience. We have relaxed restrictions on offline app usage to match our other apps, allowing up to 30 days (or until the end of your subscription) of offline usage after authentication.
Other fixes:
fixed incorrect “Invalid username or password” issues.
prevented seeing unexpected paywalls popping up while the app is open.
reduced delay for seeing credits show up for cloud exports.
Added support for the denoise filter in the cloud.
Added more error messaging for out of memory issues.
Resolved odd timeline behavior for image sequences.
loosened restrictions for encoding to Material Exchange Format (.mxf).
Topaz Video 1.2.0
Hi Everyone,
Today we have a new release with a good collection of updates!
See more details below:
Proteus Natural, is the next installment to our popular Proteus ai model. Like Proteus, this is intended as a general up-scaler for most types of content and can render locally and on the cloud.
Great for most video conditions and types
Addresses compression and general digital artifacts
Reduces halos common with older content or dark sharpening halos from digital zooming
Respects the original source material -doesn’t overly enhance/modify source inputs
Clean upscale that does not distort or warp original source details
Addresses aliased edges
Great model to mix and match with a second enhancement
image
image
1920×1569 1.45 MB
What’s the Difference between Proteus and Proteus Natural?
image
image
1920×1836 1.39 MB
This model takes the best out of Proteus and improves it with a more natural output that doesn’t over sharpen or “over enhance” everything.
Though the original Proteus is quite robust, at times the background texture can appear extra sharp or crunchy. Natural takes this into account and minimizes these types of issues.
Additionally, smaller faces (background people -small in scale) can also have distorted faces as a result of Proteus trying to understand a face from pixels of low quality video data. Proteus Natural, reduces this issue and allows for a more subtle touch up rather than an over estimation of the face.
image
image
1920×939 707 KB
Things to Note
Unlike Proteus, this does not have modifiable settings. It’s a “click and play” type of model.
Due to the more natural intention with this model, it is worth noting that it is not the sharpest model available and this is by design. If additional sharpening is desired, it is recommended to couple this model with a second enhancement for the desired look.
Though speed can be similar to Proteus, it can take a little longer.
Designed as a 2x model similar to Nyx. Like Nyx and Rhea (4x), these are not limited to the upscale multiplier. You can still output the model for any resolution that your codec can support.
Also available as a second enhancement.
Available for Windows (Nvidia & AMD) and Mac
Additional Updates to Note!
Nyx XL Fix
We have resolved the issue that some have experienced with dropped frames associated with this model.
Aion Mac Fix
Those on Mac, may have had issues with the interpolation Aion model causing a color shift and burn on the video. This has been addressed.
UI Refinements
image
image
1150×588 82.4 KB
Drive-root paths (e.g., F:) are displayed in workspace/export/model directories settings.
Export queue scrolling won’t snap to the bottom when selecting “Newest First.”
image
image
1552×980 819 KB
Welcome/Project page clips the projects grid and no longer overlaps the reference videos if you have multiple projects and need to scroll down.
Processing Update
Videos with non square pixels can be cropped and exported without error or change to the aspect ratio.
image
image
1845×667 85.2 KB
Previews with telecine turned on will display in the timeline. Previously doing a preview while Telecine enabled did not display the green visual notification of the processed preview on the timeline.
Batch Exporting Improvements
When batch exporting with Quick export, this now targets the default directory instead of the last used path.
image
image
1512×1129 159 KB
Clicking “Start Export” button (while batch exporting) now closes the Export Settings pop up menu. Previously this would stay open and require the “X” button on the upper screen right corner to be pressed.
Cloud Export Fix
Cloud downloads keep their progress regardless if you change the sort order.
Cloud output resolutions are honored. Few have received a 4K output rather than the expected minimum resolution for Starlight.
Additional Known Issues
Starlight Mini: Mac
For those with Mac that have updated to MacOS 26, there is an issue with blank or overly smeared color outputs. We are currently working with our friends at Apple to find a solution with this recent OS update. They have assured us that we will have a resolution to this soon.
Announcements
CLI Update
We want to share an update regarding Topaz Video: we’ll be gradually phasing out the Command-Line Interface (CLI) and its support.
This was a thoughtful and deliberate decision, driven by our goal of delivering a more reliable, flexible, and powerful experience over the long term. Moving away from the CLI allows us to:
Enable more customizable and modern workflows
Remove technical constraints that limit advanced features
Improve error detection, validation, and recovery
Reduce turnaround time when addressing processing issues
Provide greater flexibility for performance optimization
We recognize that this change may require some adjustment, and we truly appreciate your patience as we make this transition. Our intention is to focus on an application experience that is more consistent, dependable, and capable as the product continues to evolve.
To clarify, this change applies only to the CLI and support for CLI. It does not affect API access or custom solutions for enterprise customers. Note, that existing perpetual licenses will retain CLI.
Thank you for your understanding and continued support.
Verification Checks Update
In the coming weeks there will be a patch or official update that will reduce the amount of verification checks for subscribers of the Topaz Video App. This will align with our Photo and Gigapixel apps. For Topaz Video, the more frequent checks are no longer necessary and should allow for a more enjoyable experience with less interruptions while rendering. In addition, this should better support those dealing with areas having less than ideal internet service. Further details to come with the update.
Topaz Video 1.1.1
Fixes:
Fixed Starlight mini exports not adhering to resolution selection on Mac.
Updated Starlight Sharp cloud export resolution options to take into account maximum resolution.
Fixed second enhancement failing with Nyx XL as the first enhancement.
Fixed exports failing when preserving alpha layer and applying crop.
Resolve issues with “Export as” option and batch exporting on Mac.
Prevent overwriting original and existing files when using “Export as”.
Fix bit-depth related issues in Adobe After Effects plugin when using 16-bit color.
More precise reasons given for disabled previews.
More accurate metadata when exporting with Rhea/Rhea XL.
Topaz Video 1.1.0
Improvements and Fixes
[PRO ONLY] Starlight Mini: You asked, we listened. The Starlight Mini name is back!
[PRO ONLY] Starlight Mini: AMD and Mac improvements
Grain: greatly improved!
New Denoise Filter!
Plugin Updates: AE & Davinci Resolve
Fixed Starlight cloud resolution mismatch on the cloud for all platforms
Engine Refinements
Codec Update
And More!
Pro: Starlight Mini - Boomerang of a name!
You asked, we listened. The Starlight Mini name is back.
image
image
1075×331 64.9 KB
Some Historical Context:
Starlight Mini had dropped the “Mini” after the original Project Starlight was renamed in ASTRA (our cloud app) as Precision Quality. Precision Quality, could only render on the cloud due to the complexity and size of the model.
Starlight “Mini” was originally named for it’s more efficient model size that still keeps a high quality output.
With Project Starlight renamed, it was thought best to promote the Starlight Mini model’s name to just “Starlight” for the desktop experience. This model produces outputs that rivals the original Project Starlight while adding some unique properties that is more natural to the original content. So, the “mini” part of the name was removed.
Unfortunately, this only caused confusion. We’ve listened and learned. It gives us great pleasure to return the “Mini” back to this model’s name. By a large request, the Starlight Mini name is back to stay!
Moving forward, Starlight, represents a series of AI models. Models such as Starlight Mini, Starlight Sharp, etc are variations of this series with specified properties to enhance and improve your footage.
Pro: Starlight Mini: AMD & Mac
We have improved the quality and performance of Starlight Mini for AMD and Mac by addressing the following issues that some have encountered:
RGB channel shifts (color smears such as chromatic aberration)
Export fails even if when matching the minimum requirements.
A few reminders for Starlight Mini
You will need at least a Pro account to render locally.
A requirement of an NVIDIA GPU with at least 10GB of video memory, but we strongly recommend a GPU with 16GB or more memory for optimal performance and quality.
One of the minimum NVIDIA GPU requirements, is that the GPU needs the compute capability of at least 7.0, which means all the legacy GPUs with compute capability smaller than this version is not supported, including GTX 10 series and Quadro P series.
In addition to these GPU requirements, Starlight also requires a minimum of 32GB of system RAM.
It is recommended for AMD, to have 20GB of VRAM.
The recommendation for Mac is to have a minimum 36GB of RAM.
This is an AI model that is both complex and resource heavy. Meaning, that it could take hours to process.
It is not recommended to run Preview or Frame Preview with this model. For Nvidia users it will not work. AMD and Mac users this may also not work. It is recommended to export a few seconds for Preview purposes.
Per Cloud Render, there is a 5 minute maximum for Starlight.
Grain: Cinematic and Natural!
image
image
1920×870 877 KB
I am absolutely happy to announce the very very long awaited update for our Grain is finally here! Though our previous grain was good in it’s own right, it always felt a bit like an overlay and not scientifically accurate nor natural when compared to actual film stock or digital captures.
When light travels through a camera’s lens, it hits either the film that causes a chemical reaction or the sensor with ISO that has to be pushed further to interpret light, especially in low light situations. It is at this point for either the film or sensor interaction with light, where the grain or noise gets generated and affects your output.
What’s New?
We have overhauled the grain to take the following considerations into account:
image
image
507×262 12.8 KB
Luminosity levels (happens in the background)
RGB Channel wavelengths causes different strengths for grain (happens in the background)
3 Grain types (select-able)
Size (modifiable) - overall dimension of the grain.
Density (modifiable) - is the strength of the grain.
Amount (modifiable) - The quantity of the grain.
Grain Types
Gaussian: This is grain with color.
Grey: Similar to Gaussian, however there is no saturation.
Silver Rich: This is a more organic grain, something a little more akin to older film stock. The parameters don’t need to be as aggressive. More conservative adjustments is recommended from the default settings.
Denoise: Pure denoiser for High Quality content!
image
image
1920×767 1.06 MB
Nyx has been our leading denoise option for Topaz. However, Nyx and it’s variations are also enhancement models, meaning that although they reduce noise, the models intend to sharpen and modify the original content.
Denoise (powered by Nyx High-Fidelity) was made with a singular purpose, remove grain from high quality content without adding any additional enhancements. This is perfect for VFX teams that have film plates that need grain removal without any additional changes to their source.
image
image
499×207 12.2 KB
Things to Note
Please note, that there are no modifiable properties to this model.
Option to extract grain – You must initiate this option by clicking on the check box for “Extract noise on export.” This option will take the denoised output and input and create a noise map tiff sequence in the same folder of the export.
There is no option to access the noise map from the export directly.
Additional Updates to Note!
After Effects and Davinci Resolve Plugins
AE – fixed freezing/hang issues experienced on some machines
Overall speed for Davinci Resolve and After Effects are faster for enhancements
Quality of the enhancements are better than previous plugin iteration
Starlight Cloud Rendering
Fixed the Starlight cloud resolution mismatch on all platforms.
Additionally, add info in cloud dialog in edge cases.
Engine Refinements
Models should run, export and do frame previews for Windows AMD and Nvidia RTX 20 series cards with greatly reduced crashes. Overall giving a performance improvement for Windows AMD and Nvidia.
This also includes better general performance for 2x & 4x upscales.
Codec Update
image
image
613×724 52.9 KB
We have an additional Codec, DNxHR which includes the following Profiles
LB
SQ
HQ
HQX
444
Additional Updates
Fixed issue that requires double clicking for play
Fixed Starlight Mini and Starlight Sharp failures experienced by some users with some encoders.
Fixed issue of the application not fully closing at times.
Note:
If you experience any issues, please ensure that your drivers are up to date.
Topaz Video 1.0.4 - Patch
Fixes:
Fixed incorrect error message during network connection issues when logging in.
Fixed issue where messages not showing properly during EXR file importing when there’s less than 10 frames.
Fixed preset menu button for showing custom presets.
Skipped showing redundant model version dropdown menu.
Fixed missing icons in dialogs.
Topaz Video 1.0.3 - Patch
Hello everyone,
Today we have a sizable patch addressing the following updates below.
Be on the look out for more coming to Topaz Video soon!
Fixes:
Resolved label issue for trial button.
Rotation if functioning properly.
Login getting stuck when there are no seats.1
Raised minimum number of frames to 10.
During installation when a resource path is set, this is no longer getting reset.
Removed “crop to fill” for specific models that do not support it.
UI elements remains the same size, even when switching to bitrate mode.
Starlight Sharp will now fit within the download indicator.
Frame interpolation models no longer displays redundant model if there are no other iterations.
Topaz Video will no longer crash when cancelling an upload to the cloud
Repaired the issue with the occasional encoder mix-ups when switching back and forth between image sequence and video codec settings.
Always display the trimmer popup for starlight models.
Updated the grid/list selection preference to not be title case.
Removed the crop from being “off” on “letterbox”
Starlight resolution options have been addressed. Includes: minimum (original), 2x, 3x and 4x as long as outputs are between 720p and 4K.
Fixed a few source pane related bugs including “Close all sources” from failing.
Resolved frame number calculation error.
Refined the occasional issue of displaying source in the cloud “preview window” instead of the output within the cloud export page.
Fixed the issue of of copy/paste the interlaced/interlaced progressive model settings from not switching over properly.
There is now a warning tool tip for audio codec AC3 if an issue occurs.
Enjoy the patch updates!
Topaz Video 1.0.2 - Patch with Plugins
Hello everyone,
We’ve just released v1.0.2 with a couple of important updates based on your feedback:
Fixes:
Plugins Are Back
The plugins that didn’t make it into Topaz Video 1.0.0 have been restored to their previous state. They’re fully functional again.
We’re also working on a rewrite for these plugins behind the scenes to improve both performance and quality, so stay tuned for those improvements in a future update, coming soon.
Topaz Video 1.0.1 - Patch
Hello everyone,
Today we have a small patch addressing the following updates below.
Windows
Fixes:
Addressed Legacy toggles for EXR/DPX.
Starlight (Mini) cloud error refinements.
Fixed Starlight (Mini) download loop that some users experienced.
Repaired Starlight (Mini) RLM model issues.
Starlight family can now support 4K ProRes.
Seat fixes for Studio Pro users.
Removed frame preview pixelation.
Known Issues
Nyx XL runs slow on Mac for inputs above 1080p.
Topaz Video 1.0.0 (New Studio Release)
We’re excited to announce Topaz Video with an incredible new set of AI models and capabilities as a part of our larger Topaz Studio bundle of products and some of our most exciting features to date.
Here’s a quick preview of what’s in this release:
Brand new Nyx XL denoising AI model for all platforms. Denoise while preserving significantly more detail than ever before.
New AI Model Variant: Starlight Sharp: One of our most powerful diffusion-based video restoration models gets even more detailed. Starlight Sharp can restore with an extra level of crispness not present in any other offering. Super demanding, currently available on Windows NVIDIA platforms for local rendering, or on the cloud.
Live Frame Preview is a game-changer that allows you to instantly few updates as you adjust model parameters and settings in the app. This makes your workflow so much better. Note this feature is in Beta and doesn’t currently jive with some models like Starlight.
New General Enhancement Preset + Tuning: Not just a new enhancement preset that works in most scenarios, but also a single slider that dynamically adjusts multiple model settings as once to get the best results possible. A huge time-saver! You’re going to love this one.
Right Panel Updates: We’re continuing to refine the right panel to make it easier to do what you need to do as fast as possible. Presets are now easier to access than ever, and we’ve simplified the way we present models and settings. Clean!
Get the latest version now!
v1.0.0
Mac: Download DMG | Download PKG
Windows: Download
Released September 16, 2025
There’s so much more coming, and I’ll talk a bit more about that at the end of the post. But I also know that you probably have questions about Topaz Video and our Studio release!
What is Studio?
Topaz Studio is our collection of the latest edition of our desktop apps (Topaz Video, Topaz Gigapixel, Topaz Photo) as well as our web applications (Bloom, Astra, Express), mobile iOS app, and more! In short - it’s everything that Topaz has to offer to enhance your images and videos.
How does Topaz Video fit into Studio?
Topaz Video is available as a part of our new Studio Launch. With this launch, Topaz Video will only be available under a subscription license. Topaz Video will install into a new directory so you can try it out, and of course you’ll still have access to Video AI forever under the legacy perpetual license. Note that new features and new AI models will be coming only to Topaz Video, along with the other products in the Studio suite.
What if I already own Video AI or other Topaz products?
Purchased a Topaz product within the last year? Then you get to try Studio for free for the length of your remaining license. You’ll also get some extra benefits.
If you are a Topaz customer with an active license and you maintain your new studio subscription, you’ll be considered a Topaz Founding Customer which will give you access to some of our more powerful Pro models such as Starlight and other Pro features at no extra cost as well with your Individual license of Topaz video.
Can I buy the individual products or a perpetual license?
Topaz Studio is an incredible value - all for one low monthly price. You can still purchase subscriptions to the individual products as you wish, but you’ll get the best value with the Studio or Desktop collection bundles.
We will no longer sell perpetual licensed products moving forward, but we will continue to update them to maintain compatibility with major OS updates for as long as possible.
When will I get my free studio upgrade and access?
Please note that the transition plan is underway and may take a few days - but if you’ve already purchased and hold an active license for Video AI, or if you own more than 2 Topaz Products on an active license, you’ll have automatic access to Topaz Video at no extra cost once the transition is complete. We’ll keep you updated via email, and you can check the status of your account on the My Account page.
More questions?
If you have questions about the transition to Studio - please post them in this forum thread where Eric talks about the intention behind the release and our future plans:
Topaz Studio, the new subscription collection of all Topaz apps, begins rolling out today. Topaz Studio includes 7 products, over 100 AI image/video models, unlimited cloud image processing, and unlimited local processing. [studio-header] In 2025, we’ve released dozens of new AI models like Astra, Bloom, Recover, and Starlight. Previously, you had to individually purchase the ones you wanted. Now, you get access to everything at the cost of just two individual products. In that spirit, the S…
We know many of you are adjusting to the new subscription model, and our goal with this update is to make Topaz Video more valuable and accessible than ever — so you can spend less time worrying about limitations and more time creating.
Now - on to the latest and greatest updates available in Topaz Video!
New Starlight Model Variant: Starlight Sharp
A new addition to the Starlight family of models, Starlight Sharp excels at processing low-resolution footage, delivering superior results for even the most challenging content.
image
image
1920×680 76.7 KB
image
image
1920×911 148 KB
The model provides improved handling of smaller faces and facial details for enhanced clarity, while producing noticeably sharper and more detailed enhancement results compared to previous offerings.
Dual Deployment Options: Available on both cloud and local platforms for maximum flexibility
Multi-GPU Support: Leverages multiple NVIDIA GPUs in parallel for enhanced processing capabilities
To access Starlight Sharp - click the “Starlight” model family and select the Starlight Sharp model variant.
image (83)
image (83)
814×1273 130 KB
Starlight Sharp, can generate native 1080p outputs out of the box, or scaled 4K outputs using a special second pass. 4k outputs are currently using a Nyx variant under the hood for the second stage upscaling process.
Please note that Starlight Sharp is currently only available for local rendering to Windows users with NVIDIA GPUs, and is quite demanding in terms of VRAM, requiring 16GB of VRAM in order to run the model effectively.
Important Notes
Local rendering is only available to Founder and Pro users
Cloud Rendering options are available for all users
16GB VRAM required to run this model
Currently available only on Windows + NVIDIA GPUs
Longer render times expected due to increased model complexity
1080p or 4k output options available
Standalone enhancement and cannot be combined with other filters
Use square-pixel video footage at or under 1920x1080 to avoid aspect ratio shifts
If working with interlaced videos, we recommend de-interlacing first before enhancement
Text that is too small, blurry, or inconsistent may not be processed correctly
New Nyx XL Model
Yet another new AI model, but this time added to the Nyx family of models for incredible denoising.
a20eae3771a7bebc2d0f5cbf9fda91e6467a7c13_2_1275x523
a20eae3771a7bebc2d0f5cbf9fda91e6467a7c13_2_1275x523
1275×523 78.7 KB
The original Nyx is primarily a denoise model tuned for preserving detail in high-resolution input footage. Nyx XL takes this to the next level by learning the frames and outputting a clean and sharp output.
Nyx XL is optimized for sharpening and denoising a wide variety of footage in many lighting conditions.
This model is geared for high quality and high resolution media.
Nyx XL lends itself as a great chaining model, when merging with a different enhancement. A great results occurred when coupling Proteus with Nyx XL as the second enhancement. What kind of chaining combo can you make with Nyx XL?!
Notes
This is a more complex and heavier variant of the Nyx Model series. Please note, that this will currently produce slow render times. This is something that we are looking to make more efficient.
Currently available on both Mac and Windows.
Please note that previews can take quite a while to render.
It’s not recommended to use with Seek Render due to long processing.
This is a one size fits all type of model, so there are no modifiable settings.
Though Nyx XL is not limited by a resolution size, this variation of the model is not an up-scaler. Essentially, it will function at the resolution you input and render out the same. Though you can output at a higher resolution, the details and quality will be based on the original input resolution.
Please keep the long processing in mind with your machine’s capabilities/resources when calculating larger resolution sizes.
Live Frame Preview
3c529bd19334b4a971de408abead3ff80fd53e34_2_1275x721
3c529bd19334b4a971de408abead3ff80fd53e34_2_1275x721
1275×721 207 KB
You can now generate Live Frame Previews inside of Topaz Video, just by moving the playhead around with your current model settings.
Essentially, when you stop on any frame with an enhancement, the frame you are on will automatically load with an instant preview of said enhancement. Note, that this will not pre-render a range, just the single frame. The purpose for this is to get immediate feedback as you modify settings to find the results you are looking for.
image (84)
image (84)
466×235 39.9 KB
You have the option to disable this feature by switching the toggle “on” or “off” above the screen right side of the timeline. Currently, this is disabled by default.
Notes
This will still function if you have live render in preview enabled. If enabled, the live render will update if the video is played.
Seek render supports one or two enhancements, motion deblur and frame interpolation at this time.
Starlight AMD & Mac Support
image_720
image_720
719×319 110 KB
At long last - Starlight - one of our most powerful video restoration models ever, now runs on Mac and Windows systems with AMD GPUs in addition to the current Windows NVIDIA support. We will continue to optimize performance and quality of this model, but we’re excited to give you early access.
The model is also now a part of the Starlight “family” model selector, and loses the “Mini” designation, as it actually outperforms the base model in most scenarios. We will be building future versions of Starlight with new architectures over time - Starlight Sharp being the first in the series of next gen variants.
Please note that these Mac and AMD models are in Beta, being converted specifically for these platforms. As a result of platform architecture and the way that floating point operations are handled, there will be a small quality difference as well as a difference in system requirements (particularly VRAM) for these versions of the model. More in the notes below.
Notes
Local rendering is only available to Founder and Pro users
Currently listed as “Starlight” on the AI Model list, no longer “Starlight Mini”
This iteration of the AI model is both complex and resource heavy. Depending on the hardware you are using (AMD, MAC or NVIDIA) the performance can vary and take hours to process pending the media input. We will work to improve performance over time.
It is recommended for AMD, to have 20GB of VRAM.
The recommendation for Mac is to have 36GB of RAM
12GB of system RAM required to run this model
New General Enhancement Preset with Tuning
image (79)
image (79)
436×409 18.1 KB
Try the all new Tunable Preset: General Enhancement! The first of it’s kind, this specialized preset is built on three setting into one. As you move along the slider, this will adjust the preset parameters automatically so that you can determine how light or aggressive you want to modify the preset. At default, the modifier is centered between the two extremes.
General Enhancement has the following properties that it’s addressing: sharpness, compression and detail. This preset can also be utilized as an up scaler.
Right Panel Refresh
a43ed643f280a200712a4d50f770bd5433bda139_2_1275x1104
a43ed643f280a200712a4d50f770bd5433bda139_2_1275x1104
1275×1104 77.4 KB
We’ve made a number of updates to the right panel in Video AI to make it more accessible, especially for new users.
image
image
298×394 25 KB
In user interviews, we found that many users had difficulty finding essential tools like Starlight Mini, de-interlacing models, or presets. Others mentioned that the UI felt inconsistent and hard to navigate, especially when trying to understand how different enhancement settings worked together.
We’ve taken time to do a refresh of the right panel to make powerful tools more accessible, and ensure a better first-time experience that increases success with the application.
Clearer organization of enhancement options, with more intuitive grouping and consistent naming.
Improved discoverability for models like Starlight Mini and de-interlacing.
Presets are now easier to find and manage, making it faster to get back to your favorite workflows.
Refined parameter labels and controls, helping reduce confusion on usage and prevent mistakes.
Advanced Tuning is where you can chose your mode for the enhancement and modify accordingly.
We believe these changes will reduce friction for new users, streamline the editing experience for existing users, and ultimately help everyone get better results faster.
Topaz Video AI 7.1.4
Added Topaz Video studio download path
Topaz Video AI 7.1.3
Fixes:
Fix for the In-app updater causing a crash on some Mac systems
Fix for floating license enterprise export validations
Fixed issue where projects with missing input files did not open properly
Topaz Video AI 7.1.1
We have some updates to share in our latest updates for Video AI 7.1.1.
Bug Fixes:
Fixed issue where ffmpeg processes crash when watermark is applied.
Fixed issue of incorrect audio concatenation when scene detection is applied.
Fixed some auth-related issues.
Topaz Video AI 7.1.0
Hi everyone,
We are excited to share some of our latest updates for Topaz Video AI 7.1!
See more details below:
Video AI 7.1
Released July 17, 2025
Windows Download
Mac Download (DMG)
Mac Download (PKG)
Major Features
Cross-Scene Artifact Suppression for Frame Interpolation
image
image
1920×896 315 KB
With Scene Detection enabled, the Slow Motion function for videos, is better equipped to handle videos with cuts. This enabled feature will treat each shot/clip between cuts as independent entities. This will prevent blending/smearing of frames from the previous or next shot from your full video.
image
image
502×874 72.9 KB
As noted when you enable this feature, scene detection won’t live render.
Things to consider
It is recommended to have at least 8 frames per cut within your video to prevent potential stalled frames.
While rendering, this will split the scenes as individual parts. Ref Image: (Green) Original Source, (Red Arrow) Scene parts, (Yellow Arrow) processing scene.
image
image
979×463 77 KB
Within the Render Queue, if a process errors and the “Retry Failed Scenes” message populates under the Status section of the render Queue. Simply Click the highlighted “Retry Failed Scenes” button to continue the render.
image
image
611×171 23.1 KB
image
image
image
368×103 11.7 KB
If a crash does occur, the individual parts will remain so that you can continue from where you left off rather than having to re-render the entire video.
Once the entirety of the render completes, the individual scenes will be stitched back into a single video. Ref Image: (Green) Original Source, (Blue Arrow) Rendered Video.
image
image
958×393 52.2 KB
If any stalled frames occur, it is recommended to “Enable externally previewable exports” via the Crash Recovery section within Preferences
image
image
1161×735 89.4 KB
At this time, you will need to export locally to take advantage of this feature.
image
image
522×429 53.3 KB
Presets for Film Stock
The following Film Stock Presets will require a second pass enhancement. At default this feature is not enabled. Please refer to the Second Pass Enhancement web doc for enabling and more information.
Film Stock 4K Light
image
image
1920×742 164 KB
This setup is for film or progressive older footage that need a light touch to bring out some details.
4K preset
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Much of this setting can be used as a base and further modified to your liking.
Film Stock 4K Medium
image
image
1920×783 112 KB
This setup is for film or progressive older footage that need an extra push from the light touch version. This uses a different set of filters for the overall treatment.
4K preset
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Works best with high contrast black & white films like the example above. Please ensure that “Black & White” is checked (within the Input Settings -Gear Button next to Presets), if your film is grayscale.
Much of this setting can be used as a base and can be partially modified to your liking.
Film Stock 4K Strong
image
image
1920×779 103 KB
This setup is for film or progressive older footage with lots of grain/digital noise.
4K preset
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Great results for 8mm/Super 8mm films digitally captured with products such as Wolverine Data MovieMaker PRO, Kodak Reels Digitizer and Eyesen Film Scanner Digitizer.
Also good results with grainy 16mm telecine digital files.
Much of this setting can be used as a base and can be partially modified to your liking.
Preset for MiniDV
This setup is for MiniDV or digitally captured videos that made use of Interlaced footage (horizontal fields) for the time. MiniDV saw a large range of quality variations and conditions throughout it’s reign beginning around 1995 and ending around the late 2000’s. There was a huge use case scenarios with home movies to actual television shows and feature films.
MiniDV Int HD Basic
image
image
1920×788 122 KB
Due to the large quality and use case scenarios, the settings is a basic starting point for interlaced MiniDV footage. The intention of this preset is to get an average result that looks good in many cases. This setting is therefore intended to be modified.
HD Preset. Can be adjusted to 4k.
To make use of the Interlaced footage the settings are prepared with the Interlaced Video Type category.
The resolution is set to custom in order to keep the aspect ratio and pixel type as Square Pixel. Change accordingly.
This setting is intentionally and fully modifiable as a solid base for further refinement.
Disable Live Export
image
image
2653×1339 231 KB
You can now disable or enable the Live Export viewer.
You can find this option within the Preferences menu under the Advanced Features section.
Please note that this is enabled by default.
Disable Live Render
image
image
1920×1628 149 KB
You can now disable or enable the Live Render function.
You can find this option within the Preferences menu under the Advanced Features section.
When disabled, areas that do not have a rendered preview will automatically play the original source.
Please note that this is enabled by default.
EXR Alpha Support
image
image
1920×585 88.3 KB
EXRs with an Alpha channel, can now be supported by giving you the option to enable Alpha channel outputs via the codec settings.
image
image
658×340 45.2 KB
*Note, that at this time EXRs with Alpha may show transparency of the alpha as a checkerboard pattern in the preview. Outputs do not perform any matte operations, so the full of the RGB image will remain intact. Refer to the example below:
image
image
1657×1102 128 KB
DPX Non-Conversion
image
image
1920×631 83.7 KB
No longer will DPX files need to be converted to Tiffs when working with Video AI. Now you can use DPX files like any other image sequence.
Starlight Mini: 10x Faster in the Cloud!
image
Starlight Mini, the second diffusion-based upscaling model Topaz Labs has released, is now up to 10x faster in the Cloud when compared to its original iteration. Starlight Mini brings what makes the Original Starlight so special, into a more friendly and cost effective alternative on the Cloud.
Some Starlight Mini Reminders below
Starlight Mini is still a large resource intensive AI Model, so long renders are expected.
Recommended to use square-pixel video footage at or under 1920x1080 to avoid any aspect ratio shifts
Starlight Mini renders up to 4K resolution. Note: Starlight Mini output resolution options are limited.
Starlight Mini is a standalone enhancement and can’t be combined with other filters.
Please note that if you have text that is too small, blurry or inconsistent, the result may not be “sp3ll3d” correctly. If you can barely make it out, so will the AI model. It may assume that the text is a texture or pattern.
If you have interlaced videos, it is recommended that you deinterlace first before enhancing with Starlight Mini.
Learn more about Starlight Mini: Here
Pro GPU: Selection
Those with Pro accounts that have multiple GPUs, will be able to select the specific GPU for processing.
VP9 16K Renders
Revisited the VP9 codec to fully support 16K outputs.
Known Issues
For Cross-Scene Artifact Suppression the preview may display a frame of overlap. However, it will produce successful renders.
Enjoy the updates with 7.1!
Topaz Video AI 7.0.2
Hi everyone,
Today we have a new release with some great updates that we have been working on!
See more details below:
Improvements and Fixes
Film Stock Presets: Light, Medium & Strong
Requires 2nd enhancement enabled
MiniDV Preset
A great starting point for interlaced MiniDV videos
Added option to disable live export in preferences
Added option to disable live render in preferences
EXR Alpha Support
DPX no longer converts to TIFF
Fixed cloud renders from duplicating frames when using frame interpolation.
Presets for Film Stock
The following Film Stock Presets will require a second pass enhancement. At default this feature is not enabled. Please refer to the Second Pass Enhancement web doc for enabling and more information.
Film Stock 4K Light
image
image
1920×742 164 KB
This setup is for film or progressive older footage that need a light touch to bring out some details.
4K preset
This preset requires the Second Pass Enhancement to be enabled.
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Much of this setting can be used as a base and further modified to your liking.
Film Stock 4K Medium
image
image
1920×783 112 KB
This setup is for film or progressive older footage that need an extra push from the light touch version. This uses a different set of filters for the overall treatment.
4K preset
This preset requires the Second Pass Enhancement to be enabled.
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Works best with high contrast black & white films like the example above. Please ensure that “Black & White” is checked (within the Input Settings -Gear Button next to Presets), if your film is grayscale.
Much of this setting can be used as a base and can be partially modified to your liking.
Film Stock 4K Strong
image
image
1920×779 103 KB
This setup is for film or progressive older footage with lots of grain/digital noise.
4K preset
This preset requires the Second Pass Enhancement to be enabled.
This preset has stabilization enabled as many older family films did not have a physical stabilizer for the time. You can disable this task for faster results.
Great results for 8mm/Super 8mm films digitally captured with products such as Wolverine Data MovieMaker PRO, Kodak Reels Digitizer and Eyesen Film Scanner Digitizer.
Also good results with grainy 16mm telecine digital files.
Much of this setting can be used as a base and can be partially modified to your liking.
Preset for MiniDV
This setup is for MiniDV or digitally captured videos that made use of Interlaced footage (horizontal fields) for the time. MiniDV saw a large range of quality variations and conditions throughout it’s reign beginning around 1995 and ending around the late 2000’s. There was a huge use case scenarios with home movies to actual television shows and feature films.
MiniDV Int HD Basic
image
image
1920×788 122 KB
Due to the large quality and use case scenarios, the settings is a basic starting point for interlaced MiniDV footage. The intention of this preset is to get an average result that looks good in many cases. This setting is therefore intended to be modified.
HD Preset. Can be adjusted to 4k.
To make use of the Interlaced footage the settings are prepared with the Interlaced Video Type category.
The resolution is set to custom in order to keep the aspect ratio and pixel type as Square Pixel. Change accordingly.
This setting is intentionally and fully modifiable as a solid base for further refinement.
Disable Live Export
image
image
2653×1339 231 KB
You can now disable or enable the Live Export viewer.
You can find this option within the Preferences menu under the Advanced Features section.
Please note that this is enabled by default.
Disable Live Render
image
image
1920×1628 149 KB
You can now disable or enable the Live Render function.
You can find this option within the Preferences menu under the Advanced Features section.
When disabled, areas that do not have a rendered preview will automatically play the original source.
Please note that this is enabled by default.
EXR Alpha Support
image
image
1920×585 88.3 KB
EXRs with an Alpha channel, can now be supported by giving you the option to enable Alpha channel outputs via the codec settings.
image
image
658×340 45.2 KB
*Note, that at this time EXRs with Alpha may show transparency of the alpha as a checkerboard pattern in the preview. Outputs do not perform any matte operations, so the full of the RGB image will remain intact. Refer to the example below:
image
image
1657×1102 128 KB
DPX Non-Conversion
image
image
1920×631 83.7 KB
No longer will DPX files need to be converted to Tiffs when working with Video AI. Now you can use DPX files like any other image sequence.
Have fun with the 7.0.2 release and give these new and updated features a try!
Topaz Video AI 7.0.1
Features
Added option to disable live export in preferences
Improvements and Fixes
Applying rotation no longer highlights the export queue
For Mac, fixed navigation with arrow keys, such as using the keys to select a file
Fixed disconnected drives from being picked up on app relaunch
Improved starlight-mini downloader reliability
Updated Starlight-mini!
Resolves loading model hanging
Improved performance on higher end hardware
Fixed certain videos producing black bars
Video AI 7.0 - NEW Starlight Mini (Local) AI Model
Introducing Video AI 7
We’re excited to announce the biggest update for Video AI in a long time with version 7.0 releasing today, giving you access to a version of Starlight which can run locally (!!!) on Windows machines with NVIDIA GPUs.
That’s right. Starlight can now run on your local machine with Starlight Mini!
Please note we’ll be working on rapidly expanding support to other Windows GPUs and Mac platforms in the future. If your system doesn’t match the specs to run Starlight Mini, you’ll have to use cloud rendering for now.
Starlight MINI, you say? Skeptical about the results? Trust us - Starlight Mini is no slouch. It’s one of the best video restoration models we’ve ever built. You’re going to be blown away by what’s possible. Here’s an example:
Keep reading below to learn more about Starlight Mini … and everything else in this release of Video AI!
Video AI 7.0
Released May 13, 2025
Windows Download
Mac Download (DMG)
Mac Download (PKG)
Please review the system specs below to run Starlight Mini:
Operating System: Windows
GPU: NVIDIA GPU with at least 8GB of video memory, but we strongly recommend a GPU with 16GB or more memory for optimal performance and quality
RAM: Minimum of 12GB of system RAM
Starlight Mini local rendering will not currently work on AMD GPU/Intel GPU/macOS systems, though cloud rendering is fully accessible at this time. We’re working on making this possible! :slight_smile:
Major Features
image
image
1681×520 41.7 KB
The First-Ever Video Enhancement Diffusion Model that Renders Locally On Your Machine
Starlight Mini is the second diffusion-based upscaling model Topaz Labs has released, and it is the first to be offered as a standard in-app update for all existing and new users. Starlight mini brings what makes the Original Starlight so special, into a more friendly size for local processing.
The results of Starlight Mini can be just as astounding as the Original Starlight. Take a look:
Like its predecessor, it is capable of enhancing footage that may otherwise be a challenge with our more traditional AI Models. This can be anything from tricky 8/16mm film to mini-DV footage.
It even deals with interlaced frames, right out of the box :exploding_head:
System requirements for this model are more strict than the overall application.
System Specs to run Starlight Mini locally:
Operating System: Windows
GPU: NVIDIA GPU with at least 8GB of video memory, but we strongly recommend a GPU with 16GB or more memory for optimal performance and quality
RAM: Minimum of 12GB of system RAM
We’ll run a little check when you load up Starlight Mini locally for the first time:
Some notes on Starlight Mini below:
Starlight Mini is roughly 3x cheaper and will be up to 4x faster than the original version of Starlight to render in the cloud! However, the cloud version is not yet optimized in this release.
Recommended to use square-pixel video footage at or under 1920x1080 to avoid any aspect ratio shifts
Starlight Mini renders locally at up to 4K resolution. Note: Starlight Mini output resolution options are currently limited, and we’ll be improving these options in an upcoming release.
Starlight Mini is a standalone enhancement and currently can’t be combined with other filters. You’ll want to use the “reimport” function to add additional filters such as frame interpolation.
tpz-video-7-4
tpz-video-7-4
1840×2300 180 KB
If you’d like to review the Starlight Mini workflow in more detail, please take a look at our Docs page here which contains examples of exactly how to access the model and best practices for using it.
Starlight Mini local shows different options for resolution output from the model, based on the size of the input video.
image
image
1054×853 61.3 KB
Starlight Mini will also respect the max GPU memory settings you have in preferences as a percentage of total available VRAM:
image (51)
image (51)
1554×998 117 KB
Live Export Previews
Live previews are now visible in the export queue! This is a game-changer.
Before, there was no easy way to compare exports or cloud renders, or even to preview results inside the application.
Now, you can confirm that your outputs are matching your intended look without having to wait for a complete render.
You can also compare rendering/completed output against source video:
image
image
1920×1016 136 KB
Additional Improvements
EXR Support
Video professionals have been asking for an improvement to EXR support, and it is finally here!
image (55)
image (55)
1920×899 132 KB
TIFF conversions for EXRs are no longer necessary
32 bit image sequences are now fully supported
Video AI provides greater color support for ranges outside of the standard 0-1 range
Hear the Difference with Volume Control!
Take full advantage of controlling your video’s volume. Though you may need to hear and see an enhancement with sound, you no longer need to do this with blasting sound waves!
image
image
471×197 9.99 KB
Sometimes it’s the little things that matter.
Pro Account Panel Updates
Account-level Generative AI Controls: Video AI Pro admins can now toggle generative AI features on and off for all sub-accounts. If generative AI is turned off, the Starlight capabilities will be hidden from the interface.
Account-level Cloud Render Controls: Similar to the controls above, admins can also toggle Cloud Rendering features on and off for all sub-accounts. If the toggle is off for cloud rendering, you won’t see those options in the UI, including Starlight cloud options.
Enjoy the unique qualities from this “first of it’s kind” Starlight Mini diffusion model on the local desktop and all the updated features of 7.0!
Known Issues
Live Export player may display over-exposed results for HDR model outputs in some scenarios
https://community.topazlabs.com/t/video-ai-7-0-new-starlight-mini-local-ai-model/90631
6.2.2 released on May 8, 2025
6.2.2 update was released to fix crashes on Mac
6.2.2 update also included fixes to login flows on all platforms
Fixed Topaz Labs plugins crashing Davinci Resolve
The two plugins are now combined a single plugin (Enhance)
To chain different models, chain multiple instances of the plugin
Fixed Topaz Labs plugins crashing After Effects
Fixed ffmpeg AV1 encoding for RTX 5000s
Added volume slider to the preview controls
6.2.1 released on May 6, 2025
Fixed Topaz Labs plugins crashing Davinci Resolve
The two plugins are now combined a single plugin (Enhance)
To chain different models, chain multiple instances of the plugin
Fixed Topaz Labs plugins crashing After Effects
Fixed ffmpeg AV1 encoding for RTX 5000s
Added volume slider to the preview controls
Topaz Video AI 6.2.0
Login flow updated.
TensorRT 10 update for 5000 series NVIDIA GPUs.
Improved export error messaging.
Disabled app actions on Welcome screen.
Fixed EXR to TIFF conversion sometimes rendering invalid TIFFs.
Fixed FPS counter in Cloud export view.
Pass crop to fit properly in cloud.
Fixed opening area for preset dialog.
Scale CPU usage utilization reduced for 2x/4x outputs.
Topaz Video AI 6.1.3
UPDATE: 6.1.3 patch now available with fixes for image sequence exports, thanks to all who reported this quickly!
Topaz Video AI 6.1.2
Added 4K Starlight renders.
Added Cloud export history for renders <7 days.
Added detail to metadata when HDR model is used.
Added unique ID for each OFX plugin.
Added trackpad pinch zoom for preview and timeline.
Added dynamic compression settings for AV1 (NVIDIA).
Fixed trim intermittently not being applied for Starlight.
Fixed cases where image sequence naming contains no less than 10 digits.
Fixed warped thumbnails for non-square PAR input videos.
Fixed reset timeline zoom on input video info changed.
Fixed image sequence frame count when duration changes.
Fixed focus reset when becoming disabled via exports panel.
Fixed wrong frame/black frame on switching back to input with previews.
Fixed single view audio error.
Fixed hidden installer window on app opening.
Fixed scaling for alpha layer with aspect ratio change for pillarbox/letterbox and crop to fill.
Fixed video thumbnail being included with image sequence exports.
Fixed trim hanging for cloud exports.
Fixed image sequence start number setting if input has its own start number.
Fixed Pro user reclaiming seats when they have expired personal license.
Fixed auth token for foreign characters.
Topaz Video AI 6.1
Today’s release is a very important moment in the history of Topaz Video AI. We are releasing Project Starlight 41, the first-ever diffusion model for video enhancement.
This is a significant evolution over our existing models, and it’s the single largest increase in model capability since Video AI launched.
Even the most challenging footage can be enhanced with Project Starlight. So please, try out the “bad” footage—footage that may have previously produced artifacts with other models.
Because of its size, this model is significantly slower and more expensive to run than previous models, and requires cloud processing on server-grade graphics hardware. But this is just the beginning. Our mission as a product and research organization is to advance visual quality to the furthest extent possible—it just takes a little more processing power to get there right now.
We see a clear path towards expanding access to Starlight and other models of this generation to very high end desktop GPUs, and Project Starlight is just the start of a new series of models.
Project Starlight Research Preview
Active Video AI users can currently process three 10-second weekly previews of Starlight for free. Results will be viewable through unlisted, shareable links with the new “Compare” button in Video AI.
These previews take about 20 minutes to render if servers are available immediately, and all exports are delivered at 1080p.
Starlight Early Access
Video AI users can also render up to 9000 frames of video using Starlight (5 minutes at 30 fps).
The current introductory under-cost pricing of 90 credits/minute will change over time as our team expands access and server capacity.
Our research team is eager to receive feedback and discuss this step-change in AI video upscaling. We appreciate your involvement on this journey to Video AI’s next frontier.
Project Starlight announcement on X 8
Topaz Labs CEO Eric Yang has also written a new blog post in our Roadmap section to talk more about Project Starlight. 18
Want to learn more about Enterprise deployments for Starlight and other models? Our new page details different options for bringing Video AI to your team. 5
Project Starlight is also available in the Video AI web app.
https://app.topazlabs.com/starlight
Topaz Video AI 6.0.4
Preview queue added! Monitor all active previews in one menu.
Several improvements to OpenFX performance/stability.
Use matching parameter values from Enhancement filter for cloud exports.
Cloud exports now use the Codec settings selected in-app.
Removed “Force CFR” control, replaced with fixed VFR exports.
Fix batch estimates for certain queues.
Fix cloud download paths.
Fix a crash on closing app.
Fix frame interpolation toggling on when telecine is enabled in input settings.
Allow cloud uploads and downloads with utf-8 characters in path to files.
Preserve zoom setting on both players when player dimension changes.
Make scrollbar visible on the Exports panel.
Fixed snooze update not applied.
In-app video player now always uses input tab’s audio track
Cloud exports list now view auto scrolls to most recent item.
Fixed cross-input queued preview injection.
Prevent lone export item from disappearing.
Only show PAR type + crop setting options under Enhancement in batch if batch items require these controls.
image
Topaz Video AI 6.0.3
Enabled Enhancement in UI when combo box for output resolution is changed from original.
Enabled HLG transfer function for SDR HDR.
Renamed focus fix tooltip to match buttons.
Updated warning text for forcing CFR.
Speed improvements to Frame Interpolation plugin for OpenFX.
Fixed Davinci Resolve plugin failing to load.
Fixed Davinci Resolve Frame Interpolation plugin sometimes not installing.
Fixed Davinci Resolve crashes when running Auto mode.
Fixed naming of duplicate tabs in tab options.
Fixed default preset behavior – will now create a new tab with the preset applied.
Fixed zoom resetting on play/pause.
Prevented preset tab from setting audio mode to none.
Keep manual parameters when viewing codec settings.
Always show second enhancement if active.
Allow network path imports and exports with Video AI Cloud.
Fixed Export priority handling.
Allow setting audio settings in batch exports.
Add zoom in/out keyboard shortcuts (Ctrl/Cmd +/-).
Accept more naming schemes for image sequences.
Improved accuracy of Cloud Credit balance.
Reset enhancement params to 0 if set to auto.
Keep bitrate export settings when switching between video + image sequence export options.
Keep file name during EXR conversion.
Prevent incorrect values for scaling with non-square PAR inputs.
Topaz Video AI 6.0.1
Included a “force CFR” (Constant Frame Rate) setting in encoder options.
Added a new feature to compare tabs in single view mode via a tab option.
Fixed stabilization filter bug where enabling jittery motion does not update filter correctly.
Prevented model parameters from resetting when navigating to the export page.
Added a display for elapsed time in export items.
Brought back tooltips for view mode buttons.
Moved the status label and added an estimated time for preview.
Fixed crop display bounds during zooming by ensuring the display area always aligns with the video view.
Known issues:
Enabling “force CFR” option may cause issues in the exported videos such as duplicate frames, missing frames, or frames in wrong order.
As always, thanks for being part of the Topaz Video AI Community.
6.0
This release represents a major revamp of preview comparisons, with a tab-based interface that allows for workspace customization. Using tabs makes comparing different model settings much simpler, and brings more screen real estate back to the playback windows.
Here’s a video overview of what’s new in v6!
Introducing Topaz Video AI 6: Smarter, Smoother, Stronger!
Introducing Topaz Video AI 6: Smarter, Smoother, Stronger!
Highlights
Tabs
We’ve heard your feedback and we want to make comparison, easier, better, and more intuitive. Tabs are self contained groups of settings. You can create them from scratch or use presets. They replace for our previous preview queue and A/B comparison mode. Now rendered previews are contained in each tab, and it’s easy to compare two sets of previews, or the original. Just drag the tabs to create the view you want.
Video player
Our upgraded video player has been available in Windows and is now coming to Mac. The video player supports HDR, audio, and smooth 8K playback. Getting high quality previews in app, has never been better.
Note: HDR support on Mac is still in progress
Exports
We’ve revised the export queue to take up less space, offer more information about your exports, and make it easier to track exactly what you’re rendering.
Sources
We’ve moved the sources panel to create more work area. You’ll find it right below the timeline on the right.
Presets
The preset system has been refreshed to make it easier to find your favorite settings and quickly apply them across projects. You can select a preset directly when starting a new project. Presets are tightly integrated with tabs, you can create a tab from a preset, or save a tab to a preset.
Timeline
We gave the timeline a lot of love in this update. The upgraded timeline is faster, more responsive and has thumbnails to help you find what you’re looking for.
Live render
We’re making huge improvements to the live render system. You can render specific length chunks using the render button, and now it’s easier than ever to compare two sets of settings. Just switch into Side by Side or Split View, and press the play button to start rendering.
Other updates:
A new Preferences toggle is now available to enable/disable Cloud processing in Video AI:
CleanShot 2024-12-03 at 19.13.39@2x
CleanShot 2024-12-03 at 19.13.39@2x
1920×1232 63.6 KB
Notes:
Stabilization is currently not supported as a Live Render filter in previews. Stabilization previews can be rendered using the Render button.
5.5.1
Using Topaz Cloud Credits 277, Video AI users can now offload heavy processing tasks to our servers for faster processing.
Use cases:
Processing large batches of videos in parallel: There’s no limit to how many simultaneous jobs you can queue on the Cloud Rendering service, which can greatly improve your overall output of clips.
Processing long videos on low-power machines: Cloud Rendering excels when working with videos that are 30 minutes or longer, which allows us to parallelize across more GPU instances. For users on laptops with less-capable GPUs, you can sleep your machine without worrying about renders being paused or cancelled.
Known issues:
H.265 and AV1 are the only enabled codecs for now. We plan to expand the list of supported codecs as we determine system stability. Cloud renders larger than 8K are also disabled for this release.
Video inputs under 10MB are not currently supported for Cloud Rendering.
Hyperion is not currently enabled for Cloud Rendering.
We’re excited to work together on furthering the efficiency of Cloud Rendering, and as always we appreciate your involvement and feedback!
New in 5.5.1:
Adds support for Intel ARC ‘Battlemage’ B580 and B570 GPUs
5.5
Using Topaz Cloud Credits 7, Video AI users can now offload heavy processing tasks to our servers for faster processing.
Use cases:
Processing large batches of videos in parallel: There’s no limit to how many simultaneous jobs you can queue on the Cloud Rendering service, which can greatly improve your overall output of clips.
Processing long videos on low-power machines: Cloud Rendering excels when working with videos that are 30 minutes or longer, which allows us to parallelize across more GPU instances. For users on laptops with less-capable GPUs, you can sleep your machine without worrying about renders being paused or cancelled.
Known issues:
H.265 and AV1 are the only enabled codecs for now. We plan to expand the list of supported codecs as we determine system stability. Cloud renders larger than 8K are also disabled for this release.
Video inputs under 10MB are not currently supported for Cloud Rendering.
Hyperion is not currently enabled for Cloud Rendering.
5.4
1. All-new, GPU-accelerated video player
Video AI 5.4 uses an mpv 10-based video player that unlocks incredible improvements to playback performance. Scrub through highly-compressed input videos, or massive uncompressed media with near-instant responsiveness. This new player also supports HDR displays and audio playback! The old live preview behavior with the “Render” button is also restored. Right now, the new player is available on Windows only. We are working on bringing this player to macOS very soon.
Playback of 8K video with old player:
Playback of 8K video with new player:
2. Hyperion SDR-to-HDR model
This model is designed to perform inverse tonemapping by increasing the bit depth, expanding the color gamut, and brightening the highlights of input videos. It improves contrast and recovers detail in highlight and shadow areas of standard dynamic range inputs, and produces HDR10 (BT.2020/PQ) output.
Controls:
Exposure adjustment slider: This slider allows the user to increase the exposure of the scene by up to 1 stop. This adjustment primarily affects the brightness of the shadows and midtones.
Highlight threshold slider: This slider explicitly tells the model the lowest intensity level to be considered a highlight. This slider should be set to a value higher than the brightest skin-tone (or any other color where accuracy is vital) present in the video. In practical terms, we recommend starting with the default setting and raising the value if undesirable color shifts appear.
Saturation boost slider: The HDR color space has a larger color gamut and supports more saturated colors. Turning this slider up results in more saturated outputs in HDR without loss of detail. This slider will, however, lead to a color shift.
3. Rhea XL (RXL) Enhancement model
RXL is an even larger variant of Rhea, and is designed to handle complex textures in a more refined way.
The model requires significantly more VRAM and GPU power than previous models, so expect some longer processing times compared to Proteus or Iris.
Here’s a link to an interactive player window with a sample of Rhea and RXL’s output with the same settings applied. 51
RXL is currently listed as a “model variant” under Rhea in the Enhancement filter menu. This may change in future updates based on feedback and model fine-tuning.
Known issues:
Hyperion SDR-to-HDR does not currently support Intel-based macOS systems.
5.3.6
Fixed playback stalling from corrupt projects.
Fixed copy/paste settings not working.
5.3.5
Fixed preset name being lost on input switch and view switch
Fixed interlaced + interlaced progressive models not saving to presets
Fixed copy/paste settings from export to input
Fixed default preset causing project to not load + authentication to reset
Fixed crop+focus fix causing wrong crop area on preview and exports
Fixed combo boxes being scrollable while open
5.3.4
Fixed exports not closing in certain conditions
Fixed export jumping to top of list despite sort order
Fixed playhead cursor resetting to middle of timeline on pause
Fixed Live Render sometimes not caching (being able to seek into live rendered areas and see the preview)
Fixed Input filenames falling out of box and not eliding
Fixed audio codec disappearing in preferences when PCM is selected
Fixed copy-paste overriding fps settings
Fixed empty directories in main drive
Fixed CPE failing from missing directories
Fixed batch processing ignoring codec settings, depending on preset
Fixed bug where using batch would reset your codec settings
5.3.3
Updated batch processing
Updated flow
If you dont change settings with multiple sources selected, batch processing will use each videos unique settings.
If you do change settings with multiple sources selected, batch processing will propagate the changes you made to every input video.
Fixed multiple resolutions and model settings applying across wrong videos.
Fixed batch processing resetting AI model.
Fixed timeline playhead centering issue when quickly playing and pausing.
Fixed bug where timeline is missing major tick.
Fixed timeline padding issue.
Fixed several issues with the hovertip of input items:
FPS in hovertip now will update with the change in input settings for image sequences.
Fixed precision for FPS field.
Added feature to recommend models based on input video height.
Enabled selection of recommended models automatically on turning on Enhancement filter.
Fixed - Last major tick now only visible if minor ticks follow.
Fixed issue where playhead will disappear when going out of bound of timeline view: now reappears at the beginning of the timeline or at the smallest timestamp it can reach without causing the timeline to go out of bounds.
Fixed AV1 appearing for some unsupported GPUs.
Fixed incorrect end trim marker when placed on the frame before last.
Removed autozoom of timeline on rendering a preview.
Modified “Reset timeline zoom” button functionality to match with the older versions.
Fixed bug where image sequence start frame number text field not working.
Updated End User License Agreement
5.3.2
Added zooming into timeline with mouse wheel.
Updated timeline to center around cursor head on pause.
Updated End User License Agreement.
Fixed issue where video views lose rotation on project load.
Fixed bug where changing image sequence fps values would not update the playback fps.
Fixed bug where changing image sequence fps values would not update timeline.
Quick export now correctly identifies how many jobs it will be exporting within batch processing.
Export as now correctly disabled for batch processing.
5.3.1
Fixed a bug where crop settings no longer displayed under resolutions that changed aspect ratio
5.3.0
Control panel updates
This update includes a new Model Discovery system in the right panel, which shows examples of when to use each Enhancement model:
In addition to the new carousel, users can also switch to a list view for more compact controls:
This new model selection tool will help users navigate our growing collection of AI models and find the right fit for any project. We’ve also added a star icon for models that have very broad applications.
Timeline updates
Video AI’s timeline now offers much more functionality, with frame-accurate timeline markers and a new system for timeline zoom:
This change will make working with very long videos much more manageable, and offers precise seeking for trimming.
5.2.3
Fixed network share locations causing an app crash on export on some systems.
Fixed certain Radeon cards showing the AV1 export option when not supported in hardware.
Fixed issue where batch processing overwrites output resolution of all exports to the current opened video’s export resolution.
Fixed issue where focus-fix option enabled will lock output resolution options while Enhancement is off.
Updated minimum window size to avoid UI elements overlapping.
Improved login flow for Standard+Pro license owners.
5.2.2
Fixed AI Processor reverting to Auto
Fixed Low Power Mode not appearing on macOS systems
Improved error messaging – fewer errors will be incorrectly reported as “Colorspace error”
Improved alpha layer preservation, image sequence input image sequence output now works across TIFF & PNG
Fixed constant bitrate exports for NVENC - now uses -rc cbr for setting rate control in constant mode
Fixed bug where out trim marker was not moving for short inputs
Fixed bug where filling intervals less than 5 frames would error out
Renamed Instant Render to Live Render
Fixed preview intervals merging - will not re-render processed areas but will also not merge the two intervals
Preview render region will always start where specified + fill in gaps until end of preview region
Update 07/11: Video AI 5.2.1 is now released with a fix for videos stored on network shares sometimes not importing correctly.
Today we’re excited to release Video AI 5.2!
This release includes the new Rhea Enhancement model, Pro seat management, Frame Interpolation for After Effects, Alpha layer support, and many UI and backend fixes.
Rhea Enhancement Model
Rhea represents a combination of Proteus and Iris. The model is intended to be more accurate in preserving fine details across a wider range of subjects, while also handling text in a less destructive way.
This model and Rhea internally processes inputs at 4x scale and then downscales to your selected output resolution.
We’re very excited to see your results with Rhea, here are some example frames from this model:
Frame Interpolation for After Effects
Today we are also launching Frame Interpolation in Adobe After Effects. This plugin addition includes access to all Frame Interpolation models for up to 16x slow motion conversion.
For more information on using the After Effects Frame Interpolation plugin, see this post in the Video AI Plug-ins forum section. 2
Alpha (transparency) layer support
Video AI can now copy alpha layers from inputs and merge them back with AI model output videos. Alpha layer copying is supported in the following codecs:
QuickTime Animation
TIFF
PNG
image
Pro seat management
After a pre-launch period with several Pro teams around the world, Video AI Pro will be officially launching later this week. This new license tier offers seat management, commercial usage rights, and multi-GPU support for teams working with Video AI.
As part of this transition in licensing, we will be discontinuing the previous implementation of multi-GPU rendering for the standard license. Our team found that less than 2% of users made use of the multi-GPU setting, and further development will be focused on multi-GPU optimization for Pro teams.
If you are affected by this change and would like to discuss further options, please contact enterprise@topazlabs.com
Bug fixes & other improvements:
Preview set saving functionality restored.
UI lag greatly reduced, specifically for long operations.
Ensures colorspace correctness for input videos with a wider list of colorspace related flags.
– Uses the setparams filter instead of -colorspace, ... flags to explicitly override the input metadata to ensure only the metadata tags are updated without changing the video data itself.
Timeline width now adjusts correctly between different video inputs.
Export preferences now save consistently.
Added pause/resume controls to right-click menu.
Fixed crop settings auto-populating after first digit.
Time remaining/elapsed toggle now available for both Previews and Exports.
Fixed “Open in Explorer/Finder” button for previews and exports.
Disabled automatic generation of TIFF sequences for Previews. These temporary files are no longer in use and the app now reads preview files in the selected Export format.
5.1.4
Stabilization no longer has potential to hang app on swapping between ‘Previews’ and ‘Exports’ tab
In-to-out no longer disables the render button if your cursor is near end of timeline
Fixed Certain inputs making the timeline UI bugged
Fixed certain files not responding to ‘Show in explorer’
Fixed default export area not remembering your choice
Fixed default profile export setting not remembering
Fixed clicking render preview sometimes not putting you in processed area
Fixed many scrollable areas not hiding scroll bar when not needed
Fixed zoom percentage not updating when using shortcuts
Fixed buggy UI for image sequence Video Settings option dialog
Fixed broken shortcuts for all player controls in Welcome page
Fixed crop snapping to min/max before you finish typing
Fixed clicking cancel on preset creation still creating the preset
Fixed bug where you could edit video input settings in batch mode
Fixed ‘Include Preview & Export Settings’ in preset dialog closing the dialog
Added time elapsed/remaining to previews section
Added ‘Close Project’ button under ‘Save project as’
Added “Remove Noise” preset to defaults.
5.1.3
Fixed render in-to-out range.
Fixed player looping over same beginning region when saving with a trim in projects.
Fixed various UI icon related problems
Fixed 'None' option showing twice in 'default audio settings' area in preferences
Fixed Instant Render rotation showing stretched output
Fixed EXR failing to convert / import
Fixed image sequence exports being lost intermittently
Fixed some Instant Render related crashing
Fixed a bug where toggling Instant Render only applies to the currently selected view
Previews no longer overlap when generating and will fill in gaps instead for optimal performance
Added full app version to the header
5.1.2
Fixed exports not appearing.
Fixed image sequence related crashing.
5.1.1
Fixed slow performance of exports.
Fixed Low Power Mode missing in preferences.
5.1.0
1. Instant Rendering (Experimental)
Video AI now renders previews as soon as you click ‘Play’.
Instant renders make the timeline more interactive and allow you to directly compare models with shorter wait times.
2. Frame Interpolation for DaVinci Resolve OFX (macOS + Windows)
Access Apollo, Aion, and Chronos directly from DaVinci Resolve. Convert to slow motion at up to 16x interpolation and make use of Resolve’s retiming controls for smooth speed-ramping.
To use Frame Interpolation in Resolve, be sure to follow the guide posted here. 7
3. New multi-GPU options (Video AI Pro)
Video AI now includes two multi-GPU modes. The existing “All GPUs” option now supports NVIDIA cards running TensorRT models, leading to major utilization increases for systems with 2+ NVIDIA cards. This mode is now named “Single video” under “GPU Settings”.
The second mode, “Multiple videos”, optimizes for large export queues when running on systems with multiple GPUs. We’ll have more to share about this Video AI Pro feature when it launches in June.
4. New Welcome Screen
Quickly reopen projects and resume renders using the new Welcome Page. This is a home for recent projects and includes the ability to select favorites for easy access.
5. Preferences UI Refresh
Navigate settings using the new preferences sidebar with expanded categories and new tooltips.
6. Colorspace settings added to Video Input Options
Set custom colorspace, color primaries, color trc, and color range for video inputs with metadata issues.
5.0.4
Changelog from 5.0.3
Fixed a bug where large file names take up all the space in export panel
Updated wording for Output Resolution area in Video Settings
Added project name to Windows title bar
Updated app icon for Windows (higher resolution)
Fixed empty time display mode in export panel
Added the 'hide filters' button to the Windows title bar
Fixed failing import of EXR/DPX conversions
Updated presets for old models (Proteus v3 -> v4, Chronos Fast -> Apollo)
Upgraded to FFmpeg 7 + OpenVINO 2024.1
Known Issues
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
5.0.3
Changelog from 5.0.2
Added back reset button for timeline zoom.
Added .SER as an importable file type.
Fixed certain input tiff sequences failing to load with projects.
Fixed 0s elapsed bug when finishing export.
Fixed stabilization not reloading in app (loading with error).
Fixed benchmarking failing to load.
Fixed trim + crop + rotation persistence across projects.
Fixed EXR+DPX failing to reconvert.
Known Issues
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
5.0.2
Changelog from 5.0.1:
Fixed variable frame rate exports.
Fixed rotation previews in Workspaces.
Fixed overriding FPS values in Workspaces exports.
Fixed some Workspaces related corruption of project.tvai.
Fixed main selected video sometimes not being remembered.
Fixed rotation/crop/trim related loss of thumbnails.
5.0.1
Fixed trimmed exports with audio not functioning as expected.
Fixed intermittent crashes caused by previews.
5.0
1. Workspaces
Video AI now keeps a library of active previews and exports, allowing for project-style organization.
Your app workspace will be saved and reloaded to pick up sessions exactly where you left off using the same preview files.
2. Pause/Resume for Exports
You can now pause and resume exports in Video AI.
Prioritize different queue items on-the-fly, start up a different project, or even switch to a GPU-intensive game without losing render progress.
Even after system reboots, power loss, or unexpected crashes, your export sessions can be restored.
Paused exports are saved to disk, freeing up all CPU and GPU utilization.
3. Wonder UI Refresh
New colors & icons across all Topaz Labs apps for a unified, streamlined look.
4. Nyx v3
Nyx 3 improves the consistency of denoising across a wider variety of videos.
The accuracy of the auto-parameter estimation has been improved, and the sharpen control has been restored.
The model no longer sharpens by default, significantly reducing the unnatural grainy texture ( orange peel artifacts 6). The figure below illustrates this issue in Nyx 2 and the result from Nyx 3.
fewer Moire pattern artifacts
image
image
1920×1080 169 KB
Input (Left) Nyx v2 (Center) Nyx v3 (Right) [Gamma increased for comparison]
5. DaVinci Resolve/OFX plugin updates
Dione Dehalo & Dione Robust added to Enhancement models.
Motion Deblur moved to a separate section of the plugin control panel.
Caveats / Notes:
The following settings must be enabled for pause/resume:
CleanShot 2024-04-02 at 17.09.25@2x
CleanShot 2024-04-02 at 17.09.25@2x
1002×1200 80.7 KB
Pause/resume is not currently available for exports using Stabilization.
4.2.2
Changelog from 4.2.1
Added 2x & 4x Upscaling to DaVinci Resolve OpenFX plugin.
Fixed preset issue with deinterlace 2x being overwritten to 1x.
Fixed problem where batch processing will only export one video.
Updated minimum trim length from 1sec to 6 frames of time.
Fixed inconsistent model labels in enhancement and second enhancement.
Fixed issue with ProRes reporting two Codec labels.
Fixed bug where opening preferences windows would stack infinitely.
Removed incompatible settings with mkv container.
Fixed interlace model resetting to Prob4
Fixed jittery output with Apollo Fast
Updated OpenVINO: compatible Intel devices will perform more reliably.
How to use 2x and 4x upscaling in DaVinci Resolve OFX Plugin:
Create a new timeline that matches the desired output resolution for your video.
Set the timeline to "center crop without resizing in the Timeline Settings panel.
Apply the Enhance effect to any clips that require upscaling to the Edit panel in Resolve. (The plugin will not show scale modes in a Color node.)
Select 2x or 4x scaling, an AI model, and a parameter mode.
If your desired output resolution is not exactly 200% or 400% the size of the input video, use Resolve’s transform tools to crop/fit the video to the timeline.
Known Issues
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
4.2.1
Changelog from 4.2.0
Added plugin installation selection to Windows .msi package.
Fixed batch processing issues with interlaced/mixed FPS inputs.
Fixed previews overwriting exports when using "Export As".
Fixed preview sets / views losing their selection across inputs.
Fixed MXF trim errors.
Fixed non-square PAR videos not fitting to preview window.
Fixed input-specific playback errors.
Fixed some cases of content scrolling beyond edge of containing window.
Added Export Error Resolver to macOS.
Filter states per view + selected view are now preserved across swapping inputs.
Resolve Plugin Updates:
Faster processing, reduced initialization time, resolved some memory access crashes.
Updated to Proteus v4
"Estimate" button fixed (macOS)
After Effects Plugin:
Plugin can now be installed to beta releases of After Effects.
Known Issues
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
4.2.0
Changelog from 4.1.2
First public beta plugin release for Davinci Resolve
Video AI is now available within Davinci Resolve as an OpenFX plugin. This first release includes all Enhancement models + Motion Deblur for both Windows and macOS.
The OpenFX plugin can be installed either during app installation (automatic on Windows and an optional checkbox on macOS) OR after app installation from the “Plugins” button in the app’s top menu bar.
image
The OpenFX plugin can be used in both the Edit workspace and as a Node within the Color workspace
image
image
1920×489 47.5 KB
Aion Frame Interpolation model
More accurate motion estimation for high-resolution inputs.
Reduced artifacting and ghosting when compared with Apollo and Chronos.
True 16x slow motion without duplicate frames.
Improved temporal consistency.
Can reduce tiling artifacts depending on available VRAM.
No TensorRT support in this beta release.
3D LUT option in Export panel
Process log footage directly in Video AI by selecting a .cube file to apply during exports
image
Low Disk Space Warning System
When queuing items for export, Video AI will now warn if the selected output disk is low on space. This threshold is fully customizable from the Preferences Advanced menu.
image
Other Changes & Fixes
Changed Previews/Exports sections to use tab buttons to toggle display.
Fixed bug where renamed FHD presets still have the previous preset as a duplicate.
Fixed bug where some codec names caused the export button to overlap.
Fixed bug where file names containing '#' were truncated in "Export As" dialog.
Fixed bug where FPS is doubled when reading interlaced video files and caused errors for Frame Interpolation.
Added options to append original filename and model names to image sequence export settings.
Fixed various rotation issues.
Fixed bug where cropping a rotated image sequence did not display the full frame.
Fixed bug where Auto Letterbox/Pillarbox crop was done incorrectly for rotated videos.
Fixed bug where switching inputs would reset rotation in preview views.
Fixed bug where input video thumbnails did not correspond to changed rotation.
Fixed bug where rotation metadata was not read when files were imported.
Known Issues
Resolve-specific
On first launch, you may need to manually enable the Video AI OpenFX plugin from Resolve's preferences menu:
image
image
1413×631 34.9 KB
TensorRT is not currently supported on NVIDIA cards.
It is recommended to enable "Show All Video Frames" in the Edit workspace when using the Enhance plugin.
image
Nyx performance is significantly slower compared to the in-app model.
Video AI app
Batch processing selection can overwrite FPS output.
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Looping previews has a stutter.
Live preview can have frame de-sync.
v4.1.2
Released February 13th, 2024
Properly closes ongoing preview/export and makes sure background processes are killed.
Added toggle button for displaying remaining time/elapsed time for ongoing preview/export processes.
Added sharpen value to metadata on Nyx Fast exports.
Renamed HD presets to FHD (1080p)
Added three new QT codecs: QuickTime v210, r210, and QTRLE (Animation).
AV1 codec now allowed in .webm container.
Enabled a preference option in 'Advanced' for disabling the conversion from DPX->16bit TIFF..
Fixed some inaccurate labeling in UI (SAR-> PAR and Tooltip updates).
Fixed a bug where the remaining time of ongoing processes could show negative values.
Fixed bug with time remaining text for finished processes showing 0s.
Known Issues
Batch processing selection can overwrite FPS output.
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will intermittently be over-rotated.
Looping previews has a stutter.
Live preview can have frame de-sync.
Auto-parameter estimation moved to GPU for non-AVX2 CPUs.
v4.1.1
Released January 30th, 2024
Changelog from 4.1.0
Fix crashing on certain specific inputs.
Fix crashing on previewing for certain inputs.
Fix certain EXR/DPX inputs not converting to 16bit Tiff correctly.
Updates to OpenVINO to improve Intel ARC stability and performance
Applying frame interpolation related batch processing commands will correctly apply the FPS relative to the original.
Fix top part of window being hidden for some users.
Fix bug where auto params being lost when switching to Theia.
Fix bug with generating temp files for Exports.
Fix many preset related problems/bugs.
Add toggle functionality to switch between displaying the remaining time and the elapsed time for export processes.
v4.1.0
Released January 16th, 2024
Changelog from 4.0.9
Model Updates
Proteus v4
Better handles noise and compression artifacts.
Works on interlaced and interlaced progressive videos as well.
Improves faces better than Proteus v3.
Reduces ghosting of the previous frame at scene changes.
Better results on Auto.
More temporal consistency.
Faster than Proteus v3 in most cases.
Iris LQ v2
Improvements in detail, especially, for non-face inputs.
Improves temporal consistency (reduces 'marching ant' artifacts).
Reduces ghosting of the previous frame at scene changes (same as Proteus v4).
Reduces color distortion.
Adds TensorRT models for faster inference on certain NVIDIA machines.
Nyx Fast v1
Improvements to processing speed -- twice as fast as Nyx v2.
Reduced oversharpening and texture artifacts compared to Nyx v2.
Allows tweaking Sharpen slider value (Nyx Fast only).
Full changelog
AVX2 requirement removed and replaced with a warning about possible errors.
In a future update, auto-parameter estimation will be moved to the GPU to handle cases where Video AI is running on an incompatible CPU.
image
Rebuilt EXR + DPX support in-app
EXR and DPX inputs will now be correctly downsampled to 16-bit TIFF for processing, and exported at 32-bit linear color.
This function requires a large amount of available disk space on the drive selected for exports.
Support for ACES and other color formats is still underway, and our team is actively seeking different use cases for EXR and DPX to further improve compatibility.
Added focus fix feature to enhancement.
image
Improve missed focus in videos with the new "focus fix" mode for the Enhancement filter.
This feature changes the way models affect input videos by downscaling inputs before upscaling.
Two different modes can be used to control the strength of this effect:
Standard mode first downscales input media to 50% and then upscales to 2x.
Strong mode first downscales input media to 25% and then upscales to 4x.
image
image
1920×748 65.9 KB
Added Full Disk Access prompt to resolve disk access errors on macOS.
Improved some UI elements in filters panel.
Fixed incorrect frame interpolation related bugs when applying presets.
Fixed bug where SAR information was not retained across inputs.
Fixed bug where switching between Relative to Auto and Auto modes would not reset sliders.
Fixed bug where Nyx is resetting "Recover Original Detail" to 0.
Fixed window display related errors.
Fixed fullscreen UI issues.
Fixed bug where the app would hold a filesystem lock on input videos after app close.
Fixed bug where audio codec getting reset during trim change.
Fixed bug with metadata loss on MKV exports.
Fixed bug where Export Error Resolver tries to resolve preview related errors.
Known Issues
Batch processing selection can overwrite FPS output.
EXR/DPX Exports may lose input frame number.
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will intermittently be over-rotated.
Looping previews has a stutter.
Live preview can have frame de-sync.
v4.0.9
Released January 2nd, 2024
Updated EULA and added it to help menu for viewing.
These changes only affect Enterprise users working with Video AI in certain venues such as movie theaters and premium streaming services.
Adobe After Effects plug-in now available for macOS.
Disabled audio automatically when there are no audio tracks in input.
Batch processing now preserves more from video input settings on export.
Setting priority on outputs now respects your priority decisions.
Fix rotation bug where it won't rotate around 180 degree axis of current selected rotation.
Batch processing crop will only apply to cropped input videos, not all the selected ones.
Fixed incorrect version labels in UI.
Added "Recover Original Detail" slider details to labeling.
Known Issues
Batch processing selection will overwrite individual video filter settings.
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will intermittently be over-rotated.
Looping previews has a stutter.
Live preview can have frame de-sync.
v4.0.8
Released December 19th, 2023
Added an Export Error Resolver that helps people automatically resolve export related errors.
When there is an export error, a dialog will pop up to assist in automatically resolving the error.
This will be specifically helpful for large batch processing related problems.
Added further support for batch processing.
Fix problems with non-square PAR related Image Sequence inputs.
Fix rotation not being respected for non-preview related views.
Fix many thumbnail related mismatches.
Fix timeline related bugginess surrounding Image Sequences.
Video Player now supports frame by frame seeking inside frame-interpolated previews.
Removed few-pixel-wide black overlay when in Comparison Mode.
Model Manager will now attempt download multiple times before reporting connection failure.
Known Issues
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will intermittently be over-rotated.
Slow-motion playback intermittently inaccurate across two slow-motion previews.
Looping previews has a stutter.
Live preview has some frame de-sync.
v4.0.7
Released December 12th, 2023
Changelog from 4.0.6
Model Manager GUI has now been added. Access with File -> Model Manager.
Reintroduced live preview to the new 4.x system.
Added a way to toggle Comparison Mode UI. You can now lock the left view to Original.
Improved frame desync related issues in player.
Improved 'Export' tab UI to show more details.
NLE Plugin Update
Added multi-frame rendering for After Effects.
Known Issues
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will not be rotated.
Slow-motion playback intermittently inaccurate across two previews with different slow-motion factors playing at the same time.
Looping previews and Play button has a stutter.
Live preview has some frame de-sync.
v4.0.6
Released December 5th, 2023
Changelog from 4.0.5
Updated the video player.
Improved speed/responsiveness of player in general workflow.
Video player no longer stutters through previews playback.
Video player no longer stutters when exiting previews.
Video player can handle larger preview load in timeline.
Known Issues
Inconsistent “Preview X frames” enable status.
Rotation w/ previews will not be rotated.
Slow-motion playback intermittently inaccurate across two previews with different slow-motion factors playing at the same time.
Looping previews has a stutter.
Traded some speed on play-button-click-to-playing to improve speed on preview playback.
v4.0.5
Released November 28th, 2023
Changelog from 4.0.4
Fixes for default constant bitrate values in preferences.
Fix for “open in external player” failing.
Fix issue where entering custom values into text boxes doesn’t hold.
Simplify presets UI.
Improvements to frame de-duplication in frame interpolation.
Fixes frame count for interpolated output.
Fixes initial frames for enhancements.
AE Plugin Changelog
Added ‘Auto’ & ‘Relative to Auto’ support.
Known Issues
Inconsistent “Preview X frames” enable status.
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Stuttering on entrance-to/leaving previews / looping.
v4.0.4
Released November 14th, 2023
Crop is visually applied to the input and previews.
Adds description to preferences for Import/Export explaining that they will be applied the next time a video is opened.
Hides telecine option for image sequences (previously did nothing).
Removes confusing “Processing” label on exports that was appearing while they are waiting in queue.
Fixes zoom and upscaling issues with non-square SAR videos.
Fixes inconsitencies in crop tool.
Fixes issue in split view causing video to scale down in size while dragging pane handle.
Fixes issues editing previously applied crop.
Known Issues
Inconsistent “Preview X frames” enable status.
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Stuttering on entrance-to/leaving previews / looping.
v4.0.3
Released November 7th, 2023
Improved accuracy and consistency of frame interpolation.
Fixed issue that sometimes caused Nyx to use 2x upscaling instead of 1x.
Known Issues
Inconsistent “Preview X frames” enable status.
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Stuttering on entrance-to/leaving previews / looping.
v4.0.2
Released October 31st, 2023
Adds related preview ranges underneath preview group, including processing status and stats for each preview.
Adds looping of preview ranges by clicking the “Loop” button in the preview list.
Adds: jump directly to a preview in the timeline by clicking on it in the list.
Adds: mouse scroll wheel zooms timeline when hovered.
Adds: choose under “preview” button if you would like to process both views or just current (default current).
Fix: only show message about Nyx scaling when using old nyx-1 model.
Fix: do not reprocess previews starting at the same starting timeframe as an existing preview.
Fix: looping on in-out range.
Fix: “Auto”, “Relative” or “Manual” showing in preview info for enhancement models that don’t use this feature.
Fixes: various minor UI issues.
Known Issues
Overlapping preview ranges may have strange playback effects. This will be addressed in the next few releases.
Inconsistent “Preview X frames” range selection (example: selecting 10 frame preview may generate 9 frames)
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Some users may see a slight “stutter” during playback of rendered previews. This is file-dependent and does not affect output videos.
Cropped videos will show the full input dimensions during Preview, but the applied Crop setting will be applied to Exports.
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration.
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v4.0.1
Released October 24th, 2023
Restored ability to enter custom slow-motion values (like 2.5x).
Better UI/UX for Direct Comparison view – in-progress previews are now dimmed to show their processing status.
Fixed timeline zoom precision related issues.
Fixed UX in tutorial slideshow.
Fixed blank GPU options in preferences.
Known Issues
Inconsistent “Preview X frames” range selection (example: selecting 10 frame preview may generate 9 frames)
Preview frames syncing may be off by roughly 1/2 a frame.
Rotation w/ previews will not be rotated.
Reprocessing overlapping preview areas.
Some users may see a slight “stutter” during playback of rendered previews. This is file-dependent and does not affect output videos.
Cropped videos will show the full input dimensions during Preview, but the applied Crop setting will be applied to Exports.
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration.
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v4.0.0
Released October 17th, 2023
Changelog from 3.5.4:
Comparison views: apply different filters to Views 1 & 2 to compare them. Freely switch between reviewing input clips and previews to find the best settings possible for your video.
Preview history: previews are tracked based on filter settings and shown grouped together on the video timeline.
Zoomable timeline: take advantage of timeline zoom controls to hone in on smaller areas of a large video.
In & Out Trim: replaced the “trim” edit mode with an In & Out option comparable to other video editing applications.
Refreshed UI and filter icons: tighter more informational display.
Nyx v2: improve results and 2x scaling mode.
Significant performance improvements for the in-app video player.
Crop controls moved to video input panel.
Known Issues:
Some users may see a slight “stutter” during playback of rendered previews. This is file-dependent and does not affect output videos.
Cropped videos will show the full input dimensions during Preview, but the applied Crop setting will be applied to Exports.
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration.
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v3.5.4
Released October 12th, 2023
Changelog from 3.5.3:
Fixes issue making it impossible to set custom resolution based on height or scale when locking aspect ratio.
Fixes issue with custom resolution presets based on height not being applied.
Known Issues:
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v3.5.3
Released October 10th, 2023
Changelog from 3.5.2:
Adds “Black and White” toggle to video input options.
This option will desaturate output to prevent models from adding color to Black & White videos.
Adds “Telecined” toggle to video input options.
This option will reverse the telecine process and deinterlace your video to 23.976 fps before applying AI models.
Adds detailed tooltips to Output Settings.
Enables ProRes, H264/5 and AV1 encoders for Linux.
Fixes many color-banding issues seen in the in-app player.
Fixes update issue related to frame interpolation / ffmpeg that some users experienced after updating to 3.5.1 and was not fully addressed by 3.5.2.
Fixes tooltips so they never go off-screen.
Fixes UI issues in video settings when information displayed is multiple lines.
Changes filter tooltips so they are now also seen when you hover on the filter name and not just the icon.
Changes “os” field in video-encoders.json to a regular expression (regex) evaluation to give more flexibility.
Known Issues:
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Using Telecine mode can cause a mismatch in playback over time. This will be addressed in a future release.
B&W toggle is causing a color shift in image sequence output.
v3.5.2
Released October 3rd, 2023
Changelog from 3.5.1:
Adds Iris options to select between Medium/Low Quality. Medium will use Iris-v2, Low will use Iris-v1.
Adds 1x & 2x deinterlace options for Interlaced videos when keeping original fps.
Adds detailed tooltips for all preferences.
Adds info about Iris & Nyx models to Progressive models tooltip.
Adds detailed metadata for Iris & Nyx enhance models to exported video file.
Adds “Noise” value for enhancement models to video metadata when not zero.
Fixes ffmpeg issue causing exported video to sometimes get stuck in playback.
Fixes for rotation issues for files with rotation metadata. This was causing many of the issues like “file opened upside down.”
Fixes issues selecting “original” fps option in fps selector.
Fixes FPS display to drop extra "0"s when the fps is a whole number.
Fixes filter order when deinterlacing which should lead to higher quality visual results.
Fixes metadata for exported video when using Iris to only say “Deinterlaced” when the original video was actually Interlaced.
Iris LQ/MQ
After reviewing user feedback, our research team has decided to split Iris into two models depending on input quality. Iris v1 will now be called “Iris LQ” and is recommended for very low resolution videos and highly compressed content. Iris v2 is showing better performance on medium-resolution videos (more than 480, less than 1080) and will now be labelled “Iris MQ”.
We believe this is the best way to communicate the specific strengths of each model, and we appreciate all the feedback that came in after the initial release!
Known Issues:
Currently not all the models are supported on Intel ARC (model support will be added week to week).
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.5.1
Released September 26th, 2023
Fixes the frame interpolation models repeating frames when changing FPS.
Reduces model loading time for Nyx, Proteus, Iris Auto and Relative to Auto.
Fixes stabilization not running on lower memory machines.
Fixes crashes related to image sequence fps change.
Adds warning when rotating after setting crop dimensions.
Adds 10-bit for NVENC AV1.
Automatic parameter estimation now runs on 8 frames instead of 20.
Fixes for Nyx as second enhancement: will now automatically make the first enhancement perform upscaling.
Iris-v2 is supported on Intel ARC.
Known Issues:
Intel ARC is currently unsupported.
For Interlaced and Interlaced-Progressive videos, setting FPS back to Original may cause issues. To get around this, re-import the file. This only happens after manually changing the FPS.
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Some vertical videos will be rotated an additional 90 degrees when imported
v3.5.0
Released September 19th, 2023
Adds “Nyx” Enhancement model.
Adds Iris v2 model.
Adds support for rotating videos 90, 180, and 270 degrees.
Adds crash logs to “logs for support” and changes format to zip file.
Adds change moving “Enable Previous Model Versions” preference to advanced section of settings.
Adds change renaming “Second Pass Enhancement” to just “Second Enhancement” to reduce any potential confusion with two-pass encoding.
Fixes bug causing logs to be inaccessible on Windows.
Fixes special characters in file path causing export failure when using Export As.
Fixes crash on app open for some machines.
Fixes trim accuracy issues for ProRes.
Fixes bug causing inaccuracy with video duration after resetting trim range.
Nyx:
A new model for denoising high-quality video.
The focus of the model is on footage from high-quality cameras captured in less-than-ideal lighting conditions.
Processes videos at 1x scale to maintain the level of detail in the input video while removing noise.
Iris v2:
An improved version of Iris v1, offerring more details and sharpness especially to faces.
It uses TensorRT models for Nvidia GPUs and could result in a 50% speed-up.
It also reduces the “previous frame’s ghost” artifacts.
It reduces “marching ant” (diagonal movement of texture) artifacts, especially, for 1x and 2x.
Attention:
The format for support logs is being changed to include license information (auth.tpz). While this file is encrypted and only useful on your machine, please try to avoid posting it or your support logs here by mistake. This would not likely lead to a problem, but better to be safe than sorry. Please submit any logs directly to our support team or devs
Known Issues:
Intel ARC is currently unsupported.
If using Frame Interpolation to increase fps over 2x, output may contain duplicate frames.
For Interlaced and Interlaced-Progressive videos, setting FPS back to Original may cause issues. To get around this, re-import the file. This only happens after manually changing the FPS.
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.4.4
Released September 12th, 2023
Adds AV1 encoder for AMD 7000 series and 700 iGPUs.
Adds option to reset all preferences to defaults.
Adds new “Advanced” tab section to preferences. More items will be added to this in the future.
Adds change to copy all subtitle tracks for MKV exports. Support for more containers will be added in the future.
Adds improvements to file size for dynamic high setting without sacrificing visual quality (targets 95 VMAF).
Adds fix vastly improving responsiveness for live preview for MKV & WebM formats.
Changes minimum preview duration to 0.5 seconds, which fixes some bugs that could happen with <10 fps videos.
Fixes NVIDIA Tesla GPUs not showing encoder options.
Fixes videos encoded with H265 from AMD GPUs not playing on iOS and macOS.
Fixes output bitrate appearing on an output item when using a codec that does not use bitrate (ex. ProRes).
Fixes bug making it impossible to seek by typing frame number after trimming.
Fixes target bitrate in preferences displaying incorrect number.
Known Issues:
Intel ARC is currently unsupported due to Intel driver issues.
For Interlaced and Interlaced-Progressive videos, setting FPS back to Original may cause issues. To get around this, re-import the file. This only happens after manually changing the FPS.
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.4.3
Released September 5th, 2023
Adds MOV container option for FFV1 encoder.
Adds dynamic bitrate options for Apple Silicon computers.
Adds “Import” section to preferences and moves “Default Import File Type” to this section.
Adds preference for default FPS for image sequence import. Changing fps on an image sequence that has been imported will no longer be “sticky” - set your default here instead.
Fix for aspect ratio input to reset to Custom after applying reset or auto crop.
Known Issues:
Intel ARC is currently unsupported.
For Interlaced and Interlaced-Progressive videos, setting FPS back to Original may cause issues. To get around this, re-import the file. This only happens after manually changing the FPS.
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.4.2
Released August 29th, 2023
Fixes many issues related to in-app player seeking and trim related precision.
Fixes for deinterlaced video playback and processing feedback.
Adds FFV1 encoder options.
Adds option to preferences to default to searching for image sequences during import.
Adds “Advanced” section to application settings and moved “Show Second Enhance Control” here.
Removes “Show Intermediate Resolution Control” - this will always be shown when Second Enhance is enabled.
Fixes bug that could cause processing failures on first app launch after updating.
Fixes for Preview button - enabled for all valid trim ranges.
Fixes to Export As to respect the selected folder in preferences.
Fixes issue causing one additional frame to be processed during preview
v3.4.1
Released August 22nd, 2023
Added second enhancement
Added new user guidance
Added preference to turn off new user guidance
Video player playback and trim functionality is more reliable and accurate
Improved dynamic bitrate usability
Improved tooltips
v3.4.0
Released August 15th, 2023
Added Dynamic Compression Options (H264 and H265).
Added ability to create presets from anywhere.
Added new user experience tips (with preference to disable all).
Added “refresh login” button to update screen that will allow you to update your license by signing in again.
Added quality Improvement for first few frames while using enhancement models (except Gaia).
Added Iris to benchmark tool.
Added speed improvements for Iris on macOS Ventura (1x and 2x upscale).
Added capability to run Apollo on M1/M2 with HD or above input videos.
Added processing change: Minimum number of frames needed is now 4 instead of 5.
Fixes first few frames repeating in frame interpolation.
Fixes login to work with proxy settings.
Fixes processing on Intel CPU machines (12th, 13th gen). No longer slows down when app is minimized or out of focus.
Fixes image sequence issues when non ASCII characters are in the path.
Fixes ANE process using up all Memory on Mac.
Fixes for Tesla GPUs. Now show NVENC H264 and H265 encoders.
Fixes for log files. Can now be opened when app is open (Ctrl+G, Cmd+G).
Potential Fixes
Processing failures when exporting large number of frames.
Intel Arc/iGPU processing issues.
v3.3.10
Released August 8th, 2023
Changes from v3.3.9:
Fixed Color Correction Parameters
Fixed Preferences UI Bug
Added more Tooltips
Improved Logging
Improved Recover Original Details Tooltip
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Seeking/preview may not work well with DPX inputs
v3.3.9
Released August 1st, 2023
Reworked custom output resolution workflow to better support presets
Added tooltips for video output settings panel
Fixed preset migration for video output settings
Fixing Theia to Interlaced Enhancement UI bug
Fixing Proteus and other sliders text box bug
Added tooltip for automatic pillarbox/letterbox removal in crop tool
Fixed textbox focus for output resolution
Fixed Interlaced Iris naming
Fixed integer overflow issue for start frame number
Changing toggle box to checkbox in UI
Added version labels for Proteus and Theia
Exports / Previews now show filename when hovering over thumbnail
Fixed video setting handling of SAR values for custom resolution
Fixed “proteus” showing up in FFmpeg command and added additional preset migration safeguard
Updated the live preview checkbox in preferences
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Seeking/preview may not work well with DPX inputs
Topaz Video AI v3.3.8
Released July 26th, 2023
Changes from v3.3.7:
Fixed image sequence previews overwriting or deleting previously completed exports
Topaz Video AI v3.3.7
Released July 26th, 2023
Changes from 3.3.6:
Iris Interlaced now has Recover Original Detail slider
Topaz Video AI v3.3.6
Changelog from 3.3.5:
Reworked Preset creation, modification, delete, and apply user workflows.
Reworked custom output resolution flow.
Removed (Modified) text from presets.
Fixed video scan type changing bug when switching presets.
Fixed Preview switching Iris.
Fixed auto crop stabilization preset issue.
Fixed Proteus defaulting to V1 instead of V3 in some cases.
Output FPS combo box does not duplicate the selection.
Enhancement filter will now stay enabled for custom output resolution.
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.3.5
Released July 18th, 2023
Changes from v3.3.4
Fixed invalid state for start frame number when passed an empty string
Fixed Start Frame Number getting invalid number when deleting value to modify it on Windows
Fixed values close to max available for Start Frame Number getting random or negative values
Start Frame Number & Same as Input locations have been swapped
Fixed Start Frame Number not resetting to original when selecting “Same as Input” in Preferences
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Export settings from old presets might not be applied correctly. Try creating a new preset if this happens to you.
Deleting custom Start Frame Number value may place that setting into an invalid state on Windows. As a workaround, select the value and modify it
v3.3.4
Released July 11th, 2023
Changes from v3.3.3
Adds option to export settings allowing a custom start frame number to be set for image sequence exports.
Adds a new preference for default start frame number when exporting an image sequence and not matching the input (ex. if you want first frame to be 0000.png vs. 0001.png or 0042.png)
Adds “.3gp” as an available file type when importing videos through the file opener.
Fixes bug causing image sequences to be re-exported with start frame off by one when matching start frame from import.
Fixes color space issues when importing image sequences and exporting as video. Color space will be converted to BT709 when input color space is RGB & exporting to YUV pixel format.
Fixes invalid output when exporting PNG or DPX to VP9
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Export settings from old presets might not be applied correctly. Try creating a new preset if this happens to you.
Deleting custom Start Frame Number value may place that setting into an invalid state on Windows. As a workaround, select the value and modify it
v3.3.3
Released July 4th, 2023
Changes from v3.3.2
Adds DPX as import option for Image Sequences.
Adds - when export is pressed, an output item will always be added even if there is an error. This should improve batch processing workflows in case of errors.
Adds more clear options to export preferences for “live preview” behavior. This replaces the “video export” option. Behavior will not change from your previous setting for this option.
Adds detailed tooltips in-app to explain “live preview” options.
Adds toggle for “live preview” to export settings to easily turn on/off live preview for exports.
Adds error handling when trying to copy TrueHD audio to MP4/MOV container.
Fixes “Encoder Profile” option in export preferences not showing your saved option when opening preferences.
Fixes “Audio Codec” list in export preferences to be filtered by chosen “Extension” option.
Fixes “videoai” metadata tag missing on exported videos.
Fixes error copying AAC audio with ADTS headers from M2TS video source.
Fixes “back to input” - will always return you to the input video corresponding to the current output video.
Known Issues:
The output settings of old presets may not be applied. Please modify them if needed
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.3.2
Released June 27th, 2023
Changes from v3.3.1
Fixes frames missing from end of video when exporting with ProRes.
Fixes app sometimes trimming off one frame when exporting full video.
Improves playback speed for output
Adds ProRes Proxy profile for exports
Fixes Iris artifacts on M2
Disables audio for image sequence inputs
Syncs license with server more often if connected to internet
Misc updates
Known Issues:
The output settings of old presets may not be applied. Please modify them if needed
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.3.1
Released June 20th, 2023
Changes from v3.3.0
New export settings UI and options (details below).
Large list of “encoders” has been replaced with separate selectors for encoders – ProRes, H264, H265, VP9, AV1 – an associated profile/codec, and a GPU option if applicable (ex. you have both a NVIDIA and AMD GPU mounted).
Image Sequence options are given through a tab option now to separate UX flows for Video and Image Sequence exports. Similar to video encoders, we have added Bit Depth options for Image Sequence exports.
Removed “Auto” option for Audio Mode, default behavior is “Copy”. The UI should show an error if this is not possible for your video for compatibility reasons. (ex. PCM audio cannot fit in MP4).
Changed “Transcode” audio mode name to “Convert”.
When Converting audio, you have the option between AAC, AC3, PCM codecs. For MP4 container, you will only be presented with AAC & AC3. When exporting with VP9 + WebM container, the only codec option is Vorbis.
For AAC & AC3 you have the option to set a bitrate for the audio. AAC default is 320k, AC3 is 448k
Saved preferences from previous versions should be migrated to new preferences. ex. if you have a preference set to export as “TIFF lossless 8-bit”, your preferences in the new version should show “TIFF” for encoder and “8-bit” for Bit Depth and this should still be applied when opening a new video.
Added MKV container to extension options for VP9 exports.
Known Issues:
The output settings of old presets may not be applied. Please modify them if needed
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
v3.3.0
Released June 13th, 2023
Changes from v3.2.9
Adds a new enhancement model Iris for face enhancement and/or improving low-to-medium quality progressive/interlaced videos
Adds Recover Original Detail option for Proteus, Iris, and Artemis models
Adds DPX support
Maps all audio streams with metadata to output
Fixes issue with estimate not working for certain image sequences
Misc bug fixes and improvements
Known Issues:
Videos with mismatched metadata and streams will display incorrect duration
Frame number preview length may shorten on app restart
Submit files here: Dropbox 35
Please take a look at the Video Roadmap Update 734 for more context on what we’re focusing on right now. Thanks in advance for your feedback!
Enhancement model - Iris
Primary aim of the model is to improve human faces
This model is also designed to enhance low- to medium-quality video, particularly, with noise and compression artifacts. Note, the model tends to lose details on high-quality videos
This model is trained to work on all three categories - progressive, interlaced, interlaced progressive
Please look at some examples in the roadmap post 734
Recover Original Detail
Purpose of this slider is to retain details from original source if the model’s output is overly smoothed
It adds the source texture details to the model’s output to bring back the original details