Forum Archive Home -> Computer -> Audio Compression

Audio Compression

Nitro89 posted 2008 Jul 17 08:03

I feel that WMA is the best compression for me, but I am still using 9.2, I noticed there is also 10 professional and 9 lossless. I need a reasonable filesize as I had with the 128kbps - CBR - 2 Channel - 41khz, but now I want better quality too, it is also important that I need it to be compatible with my MP3 (supports wma and mp3). I have tested them out, I noticed 9.2 lossless gives a huge filesize yet the 10 professional gives a reasonable filesize, on 1/2mb higher than my CBR track. Is there much difference/worth changing from 9.2 CBR to a lower compression method?

Midzuki posted 2008 Jul 17 09:26

http://en.wikipedia.org/wiki/Windows_Media_Audio : Microsoft claims that audio encoded with WMA sounds better than MP3 at the same bit rate; Microsoft also claims that audio encoded with WMA at lower bit rates sound better than MP3 at higher bit rates.[44] Double blind listening tests with other lossy audio codecs have shown varying results, from failure to support Microsoft's claims about its superior quality to supremacy over other codecs. One independent test conducted in May 2004 at 128 kbit/s showed that WMA was roughly equivalent to LAME MP3; inferior to AAC and Vorbis; and superior to ATRAC3 (software version).[45] Another test performed by ExtremeTech showed different results, however, placing WMA at the top of the list in terms of quality.[37] Some conclusions made by recent studies: * At 32 kbit/s, WMA Standard was noticeably better than LAME MP3, but not better than other modern codecs in a collective, independent test in July 2004. * At 48 kbit/s, WMA 10 Pro was ranked second after Nero HE-AAC and better than WMA 9.2 in an independent listening test organized and supported by Sebastian Mares and Hydrogenaudio Forums in December 2006. This test, however, used CBR for WMA 10 Pro and VBR for the other codecs. * At 64 kbit/s, WMA Pro outperformed Nero HE-AAC in a commissioned, independent listening test performed by the National Software Testing Labs in 2005. Out of 300 participants, "71% of all listeners indicated that WMA Pro was equal to or better than HE AAC." * At 80 kbit/s and 96 kbit/s, WMA had lower quality than HE-AAC, AAC-LC, and Vorbis; near-equivalent quality to MP3, and better quality than MPC in individual tests done in 2005. * At 128 kbit/s, there was a four-way tie between aoTuV Vorbis, LAME MP3, WMA 9 Pro and AAC in a large scale test in January 2006, with each codec sounding close to the uncompressed music file for most listeners. * At 768 kbit/s, WMA 9 Pro delivered full-spectrum response at half the bit rate required for DTS in a comparative test done by EDN in October 2003. The test sample was a 48 kHz, 5.1 channel surround audio track. : WMA has been subjected to a number of complaints. "Some audiophiles challenge Microsoft's claims regarding WMA's quality," according to a published article from EDN.[12] Another article from MP3 Developments wrote that Microsoft's claim about CD-quality audio at 64 kbit/s with WMA was "very far from the truth."[54] At the early stages of WMA's development, a representative from RealNetworks claimed that WMA was a "clear and futile effort by Microsoft to catch up with RealAudio 8"[55] Microsoft has sometimes claimed that the sound quality of WMA at 64 kbit/s equals or exceeds that of MP3 at 128 kbit/s (commonly considered to be near-transparent). In a 1999 study funded by Microsoft, NSTL found that listeners preferred WMA at 64 kbit/s to MP3 at 128 kbit/s (as encoded by MusicMatch Jukebox).[56] However, a September 2003 public listening test conducted by Roberto Amorim found that listeners preferred 128 kbit/s MP3 to 64 kbit/s WMA audio with greater than 99% confidence. This conclusion applied equally to other codecs at the same bitrate, leading him to conclude that: “ No codec delivers the marketing plot of same quality as MP3 at half the bitrates.[57] ” It is important to note that both MP3 and WMA encoders have undergone active development and improvement for many years, so their relative quality may change over time. A July 2007 public listening test by Sebastian Mares found that 64 kbit/s HE-AAC audio (encoded by Nero Digital) was statistically tied with 64 kbit/s WMA Pro audio, in terms of listener preference.[58] Believe it, or not. Choice is yours, don't be late. :D

Nitro89 posted 2008 Jul 17 14:43

What is the best audio compression for a file size around 5mb at 128kbps? I always thought WMA 9.2 CBR was good but clearly not

Nitro89 posted 2008 Jul 19 12:52

I have done some reading on various websites containing the matter, AAC I read was probably best quality however its not supported by most MP3 players, only Sony and the Apple Ipod. WMA I read was good but not as good as AAC, but this article is out of date (2004) and there may be new results, is there a way I can compare audio quality using my computer or anything, I have compared using my ear, I can only detect small differences in quality.

poisondeathray posted 2008 Jul 19 12:55

I personally think AAC is slightly better (again, just opinion) This is totally subjective. You probably can't tell the difference anyway at 128kbps on a cheap MP3 player with $5 headphones. Try a $5000 audiophile reciever and $800 headphones and you might be able to discern a difference...

Nitro89 posted 2008 Jul 20 12:06

I ran my own little test today and compared statistics. For this, I used 'Winamp 5.54', 'Media Info' and my Aerosmith Album, 'Get a Grip' ('Livin' on the Edge'). I inserted my OFFICIAL Aerosmith 'Get a Grip' album, opened up Winamp and selected 'Audio CD (D:)' tab on the bottom left of the screen. I then selected 'Options', 'Preferences', 'Media Library', 'CD Ripping' and 'Encoder' and selected my encoders, for this experiment, I chose: - 'WMA Encoder v1.22a' - Encoder Format: 'Windows Media Audio 9.2' - Sample Format: '16 bits, stereo, 44100 Hz' - CBR - Bitrate (Bits/Sec): '128016' - 'aacPlus (HE-AAC) Encoder v.1.28a' - Bitrate: '128' - Channel Mode: 'Stereo' - Advanced Options: Ticked 'Tune for speech' After I ripped the track in both settings, I used 'Media Info' to compare them: - 'WMA Encoder v1.22a' - File Size: 5.64mb Codec: WMA2 Codec/Info: Windows Media Audio 2 Codec description: Windows Media Audio 9.2 - 128 kbps, 44 kHz, stereo 1-pass CBR PlayTime: 6mn 7s Bit rate mode: CBR Bit rate: 128 Kbps Channel(s): 2 channels Sampling rate: 44.1 KHz DeviceConformanceTemplate: L1 - 'aacPlus (HE-AAC) Encoder v.1.28a' - File Size: 5.60mb Codec: AAC LC Codec/Family: AAC Codec/Info: AAC Low Complexity Bit rate mode: CBR Bit rate: 113 Kbps Channel(s): 2 channels Sampling rate: 22.05 KHz Resolution: 16 bits Using statistics, it does look like that WMA offers more in the statistical region, shows a better 'quality to size ratio'. WMA offers higher bitrate and sampling rate, I want to know why AAC is giving me 113 Kbps and a 22.05 KHz sampling rate rather than 128 Kbps and 44100 KHz sampling rate. Also, I have read that AAC only encodes in VBR, not CBR.

poisondeathray posted 2008 Jul 20 12:25

You might want to use equivalent settings if you are going to do a comparison. 44100Khz sampling rate vs 22.05KHz might make a difference :) I think you can do AAC CBR. There's a setting for it in MeGUI (based on behappy) and Belight You can also change the sampling rate in Belight. In MeGUI, it uses the original rate.

Nitro89 posted 2008 Jul 20 12:40

I also converted the file in iTunes, the file is 5.66mb, VBR 44100KHz, but I noticed its a High Quality conversion, the High Quality AAC codec is only available with Winamp Pro. Both Winamp and iTunes won't allow me to change sampling rates on the AAC. iTunes File Size: 5.66mb Codec: AAC LC Codec/Family: AAC Codec/Info: AAC Low Complexity PlayTime: 6mn 7s Bit rate mode: VBR Bit rate: 128 Kbps Maximum bit rate: 157 Kbps Channel(s): 2 channels Channel positions: L R Sampling rate: 44.1 KHz Resolution: 16 bits StreamSize/String: 5.60 MiB

Nitro89 posted 2008 Oct 02 09:00

Whats the best VBR audio codec in terms of sound quality?

Login/Register to our forum to be able to post here.