Head-Fi.org › Forums › Equipment Forums › Computer Audio › FLAC is brighter than WAV
New Posts  All Forums:Forum Nav:

FLAC is brighter than WAV - Page 18

post #256 of 284
Quote:
Originally Posted by hypostasis View Post
I decompressed the FLAC and did an MD5sum on the resulting WAV. It mached the original WAV perfectly. I repeated this several times on different FLACs whenever the distortion appeared. Always the correct MD5sum.
Conspiracy theory:
I think I figured it out. The MD5 match does not say that the compared files are necessarily also a bit to bit match. It only says they are bit to bit match with very high probabilty. Nevermind how small the chance is for two files being different while having the same MD5 - it obviously did happen in your case - and even worse: not only once, byt multiple times on every single flac file examined, which leads me to an inevitable, almost crazy sounding (still I couldn't believe it myself at first) conclusion, that the FLAC encoder does use lossy compresion and to cover it up, it than changes some bits on the output so, that the resulting files give exactly same MD5's as the original ones and thus gaining imunity against MD5 attacks. I am really, really disapointed from FLAC after reading this thread, it is so obvious that this so called FLAC is nothing more than only cheap hype.

Yeah... you may say, that this is only a conspiracy theory and you can laugh at me, perhaps disagree with me. But you know waht ? There isn't much more you can do about it, coz it's definitely TRUE.

If I didn't mention it yet: big thanks for this thread, which really helped me to open my eyes and to get rid of some prevailing myths about FLAC (e.g. FLAC being lossless) and I am eager to see some more quality stuff like "FLAC vs. CDA", WAV vs. CDA" and if that's not enough, than I am pretty shure that crunchy delicacy "Wav vs. Wav" should blow mind of almost every head-fier.
post #257 of 284
Quote:
Originally Posted by appletree View Post
Conspiracy theory:
I think I figured it out. The MD5 match does not say that the compared files are necessarily also a bit to bit match. It only says they are bit to bit match with very high probabilty.
Yes it does. It's an exact bit for bit copy. If there's a difference in playback then there's a bug somewhere.
post #258 of 284
I'm sorry, but...

Quote:
Originally Posted by appletree View Post
Conspiracy theory:
I think I figured it out. The MD5 match does not say that the compared files are necessarily also a bit to bit match. It only says they are bit to bit match with very high probabilty. Nevermind how small the chance is for two files being different while having the same MD5 - it obviously did happen in your case - and even worse: not only once, byt multiple times on every single flac file examined, which leads me to an inevitable, almost crazy sounding (still I couldn't believe it myself at first) conclusion, that the FLAC encoder does use lossy compresion and to cover it up, it than changes some bits on the output so, that the resulting files give exactly same MD5's as the original ones and thus gaining imunity against MD5 attacks. I am really, really disapointed from FLAC after reading this thread, it is so obvious that this so called FLAC is nothing more than only cheap hype.

Yeah... you may say, that this is only a conspiracy theory and you can laugh at me, perhaps disagree with me. But you know waht ? There isn't much more you can do about it, coz it's definitely TRUE.

If I didn't mention it yet: big thanks for this thread, which really helped me to open my eyes and to get rid of some prevailing myths about FLAC (e.g. FLAC being lossless) and I am eager to see some more quality stuff like "FLAC vs. CDA", WAV vs. CDA" and if that's not enough, than I am pretty shure that crunchy delicacy "Wav vs. Wav" should blow mind of almost every head-fier.
ROFL!!!!!! Hmm... what do you think the motive behind this, er, conspiracy is? Also: DBT?
post #259 of 284
Quote:
Originally Posted by appletree View Post
Conspiracy theory:
which leads me to an inevitable, almost crazy sounding (still I couldn't believe it myself at first) conclusion, that the FLAC encoder does use lossy compresion and to cover it up, it than changes some bits on the output so, that the resulting files give exactly same MD5's as the original ones and thus gaining imunity against MD5 attacks.
Sweet! I wish I had thought of that!
post #260 of 284
Quote:
Originally Posted by goldenratiophi View Post
what do you think the motive behind this, er, conspiracy is?
I know it exactly, but can't speak publicly coz they are after me. Hope you understand. Ehm... sorry for the inconvenience.
post #261 of 284
Quote:
Originally Posted by maarek99 View Post
Yes it does. It's an exact bit for bit copy. If there's a difference in playback then there's a bug somewhere.
Stop silly joking !! This thing is real and huge ! It is even bigger than Da Vinci Code !


But seriously speaking, it is not 100% bit-to-bit, coz by definition MD5 is only a fingerprirnt (it does not contain complete information about the message) therefore two different files can result to exactly the same MD5 e.g: http://en.wikipedia.org/wiki/MD5
Quote:
However, now that it is easy to generate MD5 collisions, it is possible for the person who created the file to create a second file with the same checksum, so this technique cannot protect against some forms of malicious tampering.
But this seems to be quite an offtopic.
post #262 of 284
Some time ago I was messing around with the very old 8hz mp3 encoder. It had a utility included to add a .WAV header to any file and adjust it so it aligned on sample boundaries, thus making it look like a .WAV file. The reason for this was so you could run the file through a audio compressor, decompress it, and then view the results the lossy compression had on the file. Of course you could also run the converted files through ANY audio compressor.

I done just this with a few ZIP and large bitmap files. FLAC does quite bad compressing non-audio data, ZIP faired much better. This was not surprising considering FLAC like all lossless formats is designed specifically to operate on patterns within audio data. I did however get some interesting results with large bitmap files. They compressed quite well with FLAC. Nearly as good as actual audio data.

Nonetheless, once the FLAC compressed files where decompressed (using --decode in FLAC.exe) and then the .WAV modifications removed, the resulting file was 100% identical to the original. Had even a single bit been changed by the FLAC encoder, then the resulting decompressed file would have been corrupted. ZIP files don't handle corruption. A single bit change will throw off the CRC values and render the file useless. You would be unable to decompress any part of the resulting ZIP file. But it didn't in any of my tests which only means the FLAC decoder is 100% lossless.

So to re-cap: ZIP file -- > convert to WAV --> compress with FLAC --> decode back to WAV --> convert resulting WAV file back to ZIP. Resulting ZIP file 100% identical to original ZIP file, bit for bit. Same goes for every other file that I put through this process.

Off-topic, but somewhat relevant: WinRAR has a "lossless" format built in.... If you compress only .WAVs file in a RAR file, WinRAR will use a specialized lossless audio compression algorithm that achieves much higher compression than its standard algorithm does. It gets within a few MB when compared to FLAC. But it still can't beat it.
post #263 of 284
Quote:
Originally Posted by stevewm View Post
Off-topic, but somewhat relevant: WinRAR has a "lossless" format built in.... If you compress only .WAVs file in a RAR file, WinRAR will use a specialized lossless audio compression algorithm that achieves much higher compression than its standard algorithm does. It gets within a few MB when compared to FLAC. But it still can't beat it.

I've noticed that as well. Winrar compresses .wav files far beyond the other non audio compression formats out there. It still falls short of FLAC by a long shot though.
post #264 of 284
This is the most pointless debate ever. It boils down to one of three things.

1. It's all in the OP's head
Solution: ABX testing. Prove you can't tell.

2. PC fault
Solution: Run some diagnostics like MemTest86+ and SuperPi, see if there are any memory or disk errors.

3. En/decoder mis-configuration
Solution: Check all the settings, maybe you are decoding to 32 bit or with ReplayGain or dithering or something?
post #265 of 284
YOU GUYS ARE WRONG!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

There is a BIG, FRIGGIN HUGE difference in sound quality between wav and flack!!! The SQ difference is bigger between those two formats, than it is between a 128kbps mp3 and a .wav file!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
post #266 of 284
Quote:
Originally Posted by isamu View Post
YOU GUYS ARE WRONG!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

There is a BIG, FRIGGIN HUGE difference in sound quality between wav and flack!!! The SQ difference is bigger between those two formats, than it is between a 128kbps mp3 and a .wav file!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Try PCM audio stream in your WAV container, instead of the 32kbps (mono as well?) audio stream you currently use!
post #267 of 284
Quote:
Originally Posted by isamu View Post
YOU GUYS ARE WRONG!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

There is a BIG, FRIGGIN HUGE difference in sound quality between wav and flack!!! The SQ difference is bigger between those two formats, than it is between a 128kbps mp3 and a .wav file!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Wow. You really believe that? Oh well, because you yelled it at us maybe you're right.

Nah. Dude, if you don't like flack [sic] might I suggest that you just don't use it.

You can't just jump in a yell some unfounded opinion and expect to be taken seriously. These threads are fillled with discussions. If you have a point to make, then make it. But, I suggest that you need to do more to convince us of your views than simply state them as an unquestionable universal truth.

It might be worth remembering that many members here can't actually tell the differnce between wav and 128kbps MP3 in a blind test (can you?). Your starting assumption is flawed, and it got worse from there.

Do you mind telling us, coherently, why you think flac is worse than wav?
post #268 of 284
Quote:
Originally Posted by fordgtlover View Post
It might be worth remembering that many members here can't actually tell the differnce between wav and 128kbps MP3 in a blind test (can you?). Ywav?

I dunno about that, Even 128kbps and 192kbps is significant in the clarity.

The rest though, spot on.
post #269 of 284
Quote:
Originally Posted by fordgtlover View Post
Wow. You really believe that? Oh well, because you yelled it at us maybe you're right.

Nah. Dude, if you don't like flack [sic] might I suggest that you just don't use it.

You can't just jump in a yell some unfounded opinion and expect to be taken seriously. These threads are fillled with discussions. If you have a point to make, then make it. But, I suggest that you need to do more to convince us of your views than simply state them as an unquestionable universal truth.

It might be worth remembering that many members here can't actually tell the differnce between wav and 128kbps MP3 in a blind test (can you?). Your starting assumption is flawed, and it got worse from there.

Do you mind telling us, coherently, why you think flac is worse than wav?
erm, I believe this is what the english would call a "woosh"
post #270 of 284
..computers need to warm up?
lol.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Computer Audio
Head-Fi.org › Forums › Equipment Forums › Computer Audio › FLAC is brighter than WAV