Head-Fi.org › Forums › Equipment Forums › Computer Audio › Why does wav sound better than FLAC?
New Posts  All Forums:Forum Nav:

Why does wav sound better than FLAC? - Page 3

post #31 of 43
4/30/07 at 3:21pm
Quote:
Originally Posted by Febs View Post
Neither test should be influenced by bias or expectations. Indeed, the whole purpose of an ABX test is to remove the potential influence of expectation or bias from the testing environment.
Shouldn't ... , not convinced ... , sorry ...
Reply
post #32 of 43
4/30/07 at 3:48pm
Your posts are very difficult to follow. Would it be too much to ask you to write complete sentences?
Reply
post #33 of 43
5/1/07 at 2:45am
Quote:
Originally Posted by Febs View Post
Your posts are very difficult to follow. Would it be too much to ask you to write complete sentences?
Okay lets do this an other way ... , see next post ...
Reply
post #34 of 43
5/1/07 at 2:58am
Quote:
Originally Posted by Febs View Post
Neither test should be influenced by bias or expectations. Indeed, the whole purpose of an ABX test is to remove the potential influence of expectation or bias from the testing environment.
Quote:
Above you say Neither test SHOULD be infleunced ....
I replied "Shouldn't" which is not convincing ....

DBT don't work, and for the same reason neither do ABX, as ABX is DBT with some added "features" ...

Law of mathamatic's states, you change something one side of equation same equal change must be applied to other side of equation ... which is probably not really relavant here but ...

No matter what you do to DBT or ABX you still have NOT changed anything on the other side of the equation i.e. the person's BIAS ATTITUDE EXPECTATIONS, and it is these that determine what the outcome of the so called "test" will be, not what is done on the other side ...

So "sorry" for both DBT or ABX, they are not conclusive nor consistent, tho ABX has the potential to be a bit better than DBT ....

Way that cookie crumbles here .... , end of story...

And I haven't even mentioned the ENVIROMENT that forged those BIAS'es EXPECTATION's and ATTITUDE's - and I am not prepared to go there, ... , so again sorry ...
Reply
post #35 of 43
5/1/07 at 3:08am
Something is messed up on the decoding end or the encoding end.


Honestly, FLAC compression is the same zipping the wav file then re-extracting it. There's NO change in the original file. That is unless some replaygain or normalization or something is taking hold.
Reply
post #36 of 43
5/1/07 at 3:21am
Quote:
Originally Posted by Redo View Post
Something is messed up on the decoding end or the encoding end.


Honestly, FLAC compression is the same zipping the wav file then re-extracting it. There's NO change in the original file. That is unless some replaygain or normalization or something is taking hold.
Agreed, and the only way to check if there is a difference it to do a bit for bit binary comparison, and if there is a difference, then to go looking for borked codec's or settings (replaygain/normalization/decompression ... , etc)
Reply
post #37 of 43
5/1/07 at 3:28am
Quote:
Originally Posted by Wodgy View Post
Does anyone have a reference to this alleged 2dB reduction? I just encoded and then decoded a WAV through Flac 1.1.3 defaults, and the files are bit-for-bit identical.
x2.

That's the whole point of lossless. Otherwise it might as well be called "almost, but not quite lossless".
Reply
post #38 of 43
5/1/07 at 3:32am
Quote:
Originally Posted by java View Post
I replied "Shouldn't" which is not convincing ....

DBT don't work, and for the same reason neither do ABX, as ABX is DBT with some added "features" ...

Law of mathamatic's states, you change something one side of equation same equal change must be applied to other side of equation ... which is probably not really relavant here but ...

No matter what you do to DBT or ABX you still have NOT changed anything on the other side of the equation i.e. the person's BIAS ATTITUDE EXPECTATIONS, and it is these that determine what the outcome of the so called "test" will be, not what is done on the other side ...

So "sorry" for both DBT or ABX, they are not conclusive nor consistent, tho ABX has the potential to be a bit better than DBT ....

Way that cookie crumbles here .... , end of story...

And I haven't even mentioned the ENVIROMENT that forged those BIAS'es EXPECTATION's and ATTITUDE's - and I am not prepared to go there, ... , so again sorry ...
The bias exists only when the user knows, consciously or not, which input they are biased towards. DBT removes that variable. How can you be biased to, say, WAV files sounding different than FLAC when you don't know which are WAV and which are FLAC? Please explain, because you're not making any sense. You can certainly be biased towards one or the other, but how on earth can it affect the results if you don't know which is which? The only way to tell is if they sound (or whatever for other DBTs) different, in which case the results will show this. If they sound identical there's no other way to tell them apart, so why would the results not show a statistically even distribution?

And on the topic, the short answer is that they can't, assuming you've eliminated all variables other than the compression. The long answer is that you probably have some kind of post processing going on in one of the files and not the other. As others have suggested, I'd be that the FLAC files have replaygain enabled.
Reply
post #39 of 43
5/2/07 at 8:41pm
Quote:
Originally Posted by error401 View Post
The bias exists only when the user knows, consciously or not, which input they are biased towards. DBT removes that variable. How can you be biased to, say, WAV files sounding different than FLAC when you don't know which are WAV and which are FLAC?
Because of a possible problem with his decoder and/or encoder. Or some other software related problem (or even the CPU being to slow). If that is the case, then he will hear a difference between them with an ABX or DBT test, but the difference will not come from there actually being a quality difference, but rather from the software problem that nobody else has.

If a FLAC file decodes to a WAV file bit identical to the original WAV, then they are exactly the same. Period. Factors that we don't yet know about or are able to test is a moot point here. Want an example? Open up a major windows system DLL in a binary editor, and change a single, random bit. Save it and reboot. Hmm, I wonder why you get that BSOD when you try and start the computer now?

Don't respond with I don't use windows or anything like that -- you know what I mean.
Reply
post #40 of 43
5/2/07 at 10:01pm
Quote:
Originally Posted by 003 View Post
Because of a possible problem with his decoder and/or encoder. Or some other software related problem (or even the CPU being to slow). If that is the case, then he will hear a difference between them with an ABX or DBT test, but the difference will not come from there actually being a quality difference, but rather from the software problem that nobody else has.
Well, yea - but that has nothing to do with bias.
Reply
post #41 of 43
5/2/07 at 10:05pm
True, but in that case I would work on correcting the problem rather than using all WAVs, because of how large they are.
Reply
post #42 of 43
5/2/07 at 10:57pm
Quote:
Originally Posted by error401 View Post
As others have suggested, I'd be that the FLAC files have replaygain enabled.
FLAC files cannot "enable" ReplayGain. ReplayGain data in a FLAC file is in a part of the file separate from the audio data; it's existence has nothing to do with how the audio data is played back unless the software player has been configured to use it (and I don't believe there is a software player out there that supports ReplayGain without letting you disable it). It's kind of like a post-it note on a CD that says to play it at volume level 8...if the person playing it does not want to follow those instructions, he/she is free to ignore them and enjoy the CD at whatever volume level the player is set to use without any loss of quality or volume compared to a CD without the post-it note.
Reply
post #43 of 43
5/2/07 at 11:26pm
Quote:
Originally Posted by Scrith View Post
FLAC files cannot "enable" ReplayGain. ReplayGain data in a FLAC file is in a part of the file separate from the audio data; it's existence has nothing to do with how the audio data is played back unless the software player has been configured to use it (and I don't believe there is a software player out there that supports ReplayGain without letting you disable it). It's kind of like a post-it note on a CD that says to play it at volume level 8...if the person playing it does not want to follow those instructions, he/she is free to ignore them and enjoy the CD at whatever volume level the player is set to use without any loss of quality or volume compared to a CD without the post-it note.
I think thats what he meant too. Since wav files don't support any tags, they cant have RG tags either, so only FLAC will play the music with RG "enabled" ie. when the player is setup to use RG values.
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Computer Audio
Head-Fi.org › Forums › Equipment Forums › Computer Audio › Why does wav sound better than FLAC?