[SirDrexl]

Quote:

Originally Posted by jcx /img/forum/go_quote.gif extract the flac to .wav and compare the original and formerly compressed wav with: Audio DiffMaker should show if meaningful differences exist even if alignment problems makes bit comparison difficult

The people who claim there is a difference don't deny that if you extract the complete wav file from the flac, you will get the same data. What they are fixated on is that there could be a difference in playback, since the file is being decoded as it plays. I guess the way to test that would be to capture the SPDIF output when playing each file and compare the difference, if possible.

I don't really believe there would be a difference though, unless the CPU is seriously tied up with something else. Most players buffer the output anyway, so they should be capable of decoding properly during playback.

 

[mesasone]

Quote:

Originally Posted by Ham Sandwich /img/forum/go_quote.gif True. My reply was more focused on potential differences in a PC setup. MP3 players have the same and similar issues -- different code paths for different formats. If the different code path causes different processing and/or conversions to be done on the audio data then there is the possibility for audible and real differences. In MP3 players you can have some formats decoded in software on the CPU and other formats decoded in hardware on a special purpose DSP. That's two very different code paths designed by two (or more) very different engineering teams. There is no telling what kind of funny business and shortcuts are done on the DSP for decoding. What goes on inside the DSP is beyond the control of the software developers making the software for the MP3 player. Based on software development, and the mess that real world actual software actually is, there is reason to believe that there actually could be audible differences in a lossless compressed format vs. WAV. If the code and processing done on FLAC vs. WAV is equivalent there should be no audible or even measurable differences between the FLAC output and WAV output. The problem is that in a closed source system you cannot be sure that the code and processing actually is equivalent. I'm mainly playing devils advocate here. I'm not one who believes WAV sounds better than FLAC.

Not to be overly semantic, but wouldn't this mean that the two files/formats sound the same and the difference heard comes from the hardware/software and not the file? I know the difference is subtle and this is probably what you meant, but I think it's worth clarifying.

 

[CrackBerry9000]

Either they sound different or my cables are burnt in real nice, I will let you decide.

 

[Ham Sandwich]

Quote:

Originally Posted by mesasone /img/forum/go_quote.gif Not to be overly semantic, but wouldn't this mean that the two files/formats sound the same and the difference heard comes from the hardware/software and not the file? I know the difference is subtle and this is probably what you meant, but I think it's worth clarifying.

Yes. WAV and FLAC are the exact same audio data. They will sound the same with all else being equal.

Potential differences would be due to software treating them different. For example the WAV file might go straight through with no processing done while the FLAC ends up going through a code path that results in the audio being converted to 32 bit float then back to 16 bit integer.

 

[jcx]

Quote:

Originally Posted by SirDrexl /img/forum/go_quote.gif The people who claim there is a difference don't deny that if you extract the complete wav file from the flac, you will get the same data. What they are fixated on is that there could be a difference in playback, since the file is being decoded as it plays. I guess the way to test that would be to capture the SPDIF output when playing each file and compare the difference, if possible. I don't really believe there would be a difference though, unless the CPU is seriously tied up with something else. Most players buffer the output anyway, so they should be capable of decoding properly during playback.

I agree but you have to test to see that you've got the same data to start with - maybe the data was hit with some processing like replay gain before the flac compression and there is a difference in the data

 

[tosehee]

Non-sense, imo.

 

[deadhead12]

I can tell the difference between 128k and all the rest of my digital music library. The only 128k songs I listen too are off Pink Floyd's Piper at the Gates of Dawn, so if the track I'm listening to is off that album, its 128kbps, if not, its probably lossless or 320.

BTW- If anyone can spare a copy of the aforementioned album, pm me.