AfterDawn: News

The results of 128kbps extension public listening test

Recently RareWares organized a public lossy audio compression listening test. The formats and encoders in this test were: AAC(Apple QuickTime by Dolby), LAME MP3,MusePack, OGG Vorbis, WMA Pro, Blade MP3.

How to interpret the plots: Each plot is drawn with the five codecs on the x axis and the ratings given (1.0 through 5.0) on the y axis. N is the number of listeners used to compute the means (average ratings) and 95% confidence intervals. The mean rating given to each codec is indicted by the middle point of each vertical line segment, and the value is printed next to it. Each vertical line segment represents the 95% confidence interval (using ANOVA analysis) for each codec.

The results lead us to some interesting conclusions. First of all, MusePack is considered to be optimized for higher bitrates but it’s performance seems to be top notch even at the 128kbps range. Dolby’s AAC is close and so is OGG Vorbis, but the performance of WMA Pro by Microsoft was a slight surprise perhaps. LAME MP3 encoder is competitive, but loses to the newer and more advanced formats. BladeEnc demonstrates a bad example of audio compression and hopefully these test results help to wipe out the myths of the quality of BladeEnc permanently.



Read the full results and commentary at RareWares.

Written by: Lasse Penttinen @ 4 Aug 2003 11:50

• Previous article

• The reasons behind declining CD sales

• Next article

• Alcohol 120% driver update

• 6 comments

• rjamorim

  Just for your information: I plan to start a 64kbps test in about a month, comparing Vorbis, HE AAC, WMA pro, MP3pro and Real Audio Cook.
  
  Best regards;
  
  Roberto.

  4.8.2003 12:00 #1

• cd-rw.org

  Even though I am aware of the difficulties in executing a high bitrate test, I would love to see a ~192 kbps range test.
  
  It would be interesting to see the difference between LAME and newer formats. On the other hand, people would need yet another reminder of the defects of 192kbps CBR encoding.
  

  http://CD-RW.ORG
  Serving The Burners
  Burning The Servers

  4.8.2003 22:59 #2

• rjamorim

  Indeed. But the problem of high bitrates is that all formats (even Lame MP3) easily reach transparency on most samples.
  
  Therefore, this test could only be taken by real golden ears (the likes of Guruboolez) and only using problem samples.
  
  That raises another issue: problem samples are representative of a very small percentage of the various musical styles. Therefore, the test wouldn't be representative - I.E, wouldn't be useful to anyone but people that like to listen to those kinds of samples.
  
  If I conduced such test, I would expect results with all codecs ranking 5 most of the time pouring in my mailbox. :-/
  
  Regards;
  
  Roberto.

  4.8.2003 23:36 #3

• Shegax

  yeah
  

  Shega
  (adrenaline Is the closest Total freedom)

  5.8.2003 00:10 #4

• seanbyrne

  I prefer a codec that keeps good average quality for a typical bitrate, e.g. the MPC codec here looks interesting. WMA9 Pro VBR is definitely an improvement over the previous version (from Windows Media Encoder 9, not the one in Windows Media Player 9), but seems to be a disaster for me for music with stereo voices, e.g. M People & Billy Joel.

  5.8.2003 02:53 #5

• rjamorim

  Well, keep in mind that WMApro is still in version 1. We can expect it to be improved on upcoming versions.

  5.8.2003 11:58 #6