Computer Music Audio Quality, Part 3

bramankp -- Fri, 01/06/2012 - 09:04

FLAC is not the problem. Period.
 
Period. Period. Period.
 
So, I read the article while dousing my head with water to keep from flaming in this post because I really wanted to put some thought behind what I wrote. I performed the following test ...
 
1) Downloaded a 24/96 WAV file from http://www.2l.no/hires/index.html.
2) The FLAC 1.2.1 release to perform a WAV-FLAC-WAV transormation at both compression levels 0 and 8.
3) Given, orig.wav (the original, downloaded file renamed), orig-flac-wav-0.wav (the transform at compression level 0), and orig-flac-wav-0.wav (same thing at level 8) ...
4) ALL FILES ARE PERFECTLY IDENTICAL. This includes both the WAV header and PCM data. (I used a Linux/UNIX command called "diff" to verify this as well as the command md5sum from GNU coreutils 7.6).
 
What, if possible, could I conclude from this?
 
1) If I played back the files and somehow all of the files in step 3 sounded differently I have no way to attribute the differences to the FLAC step. I could have very well copied the files and renamed them without ever using FLAC and it would have turned out the same results.
2) If somehow the files in the article *ARE* different and that's the reason they sound different then it must be the utilities creating the files themselves that are in error, not FLAC. FLAC is a perfectly lossless system given my tests. Perfect.
 
Maybe I sould author a letter to the editor but I just wanted to get my own test results out there for someone to comment on.
 
 
Paul Braman

bramankp -- Fri, 01/06/2012 - 09:06

 Look at all those spelling mistakes and grammatical errors. Got me so riled up. :)

Robert Bertrando FB -- Fri, 01/06/2012 - 17:25

 I'm actually beginning to wonder if an unstated purpose of this series is to undermine the whole field of computer audio??  There's certainly a poor understanding of some basic principles.
 
It's been posted in some other more informed forums that there could be differences in the playback of two identical files from a computer's hard drive, depending on their position on the hard drive and their degree of fragmentation, and the number of other computer processes running in the background during file playback.  Certainly a significant weak point in the playback chain is the "on-board sound card" whence the S/PDIF output is taken (as I pointed out in my letter to the editor a month or so ago).  Trying to say a file might contain "jitter" is non-sensical, as is saying that we have no way to quantify "jitter".
 
Others have also already posted a number of tests of WAV > FLAC > WAV conversion, some with many cycles, and have confirmed recovery of the original file in all respects, whether by checksums, null tests or simple listening.

bramankp -- Fri, 01/06/2012 - 21:02

 I couldn't really find but one post in one other forum talking about this issue. Care to link some?

Robert Bertrando FB -- Fri, 01/06/2012 - 22:52

Audio Asylum & What's Best Forum, for two.

Robert Bertrando FB -- Tue, 01/10/2012 - 17:02

 I noticed yet another rather glaring error:  the authors state that "professional editing programs like Izotope" only work with uncompressed (WAV) files, which is true as far as it goes.  What it doesn't say is that true professional editing workstations like Pyramix and Sonoma DO use FLAC!
 
And of course it's already been noted above that interconversion between FLAC and WAV is no different from Windows' standpoint than copying or renaming a WAV file.  I guess we can't copy, move or backup our audio files anymore?

All content, design, and layout are Copyright © 1999 - 2011 NextScreen. All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.