Uncompressed Lossless (WAV) vs Compressed (FLAC / ALAC) - O/T discussion moved from main forum thread

Discussion in 'Sound Science' started by haiku, Apr 30, 2016.
First
 
Back
1 2 3 4 5 6 7 8 9
Page 10 of 11
11
Next
 
Last
  1. ev13wt
    Look at this clusterf3ck of a thread. Reminds me why I stopped looking at General. Blood pressure.

    At loss for words for what goes on in General Headphones. Its way beyond: "Hey look, cute, noobs discussion noob stuff."
     
    Last edited: Sep 8, 2017
    spruce music likes this.
  2. Kdubbs82
    I must admit. I can’t tell the difference between uncompressed formats. What are you listening for?

     
  3. gregorio
    It's not really a case of what to listen for, it's a case of either: 1. How to degrade the performance of your playback device to the point that it can't handle the computationally simple task of unpacking a flac or alac file, or 2. How to influence/bias your perception so that you perceive a difference where there is none.

    1. Achieving #1 is not as easy as it may appear. Firstly, the computation requirements of unpacking a flac or alac are tiny compared to the computational power available in modern (or even moderately old) playback devices. Secondly, for there to be a difference between flac/alac and say wav or aif, you have to degrade the performance of your device to the point that it struggles to perform the unpacking of alac/flac but not to the point that it struggles to convert wav of aif into raw PCM for the DAC chip. This balance can be quite difficult to achieve, it requires exactly the right amount of performance degradation.

    2. The very act of describing a potential problem can be enough by itself to bias perception and a significant amount of audiophile marketing is based on exactly this principle. If it doesn't occur by itself, automatically in response to marketing and/or anecdotal evidence, it can be difficult or impossible to consciously bias your perception to this degree though.

    By way of analogy, quantum mechanics tells us that a sub-atomic particle can spend a proportion of it's time anywhere in the universe and therefore, as we humans are made entirely of sub-atomic particles, we can spontaneously disappear and reappear anywhere else in the universe. So here we have a "potential problem" and maybe now that you're aware of it you might be a bit more worried about it? There's no/little anecdotal evidence describing this happening and therefore you're probably not worried at all and, science predicts the probability of this occurring to all the sub-atomic particles in your body at the same time is so tiny that you'd have to wait for much longer than the age of the universe to stand even a chance of experiencing it. And that's what we've got here:
    We have a "potential problem" with flac/alac unpacking, Wav or Aif CAN sound better and you CAN spontaneously disappear and re-appear somewhere else in the universe but there's not much chance of either! Although to be fair, there is more chance of the former than the latter. So, "Listen and you'll hear the difference" is the same as saying "watch someone and you'll see them disappear", both of which are true, you just need the right conditions; a device screwed to exactly the right degree and watching someone for a very, very long time. In both cases though, you'd have a lot more to worry about than the difference between alac/flac and wav/aif!

    G
     
    Kdubbs82 likes this.
  4. bigshot
    That was a whole lot of words there.
     
    upstateguy likes this.
  5. DJtheAudiophile
    I’m not going to speak on flac, because you have so many options to compress the flac file size in XLD but in iTunes to my ears Wav sounds better by a long mile than ALAC. Good enough that I only rip my cd’s in wav now. Just my PERSONAL opinion.
     
  6. bigshot
    What do you generally base your PERSONAL opinions on?
     
  7. upstateguy
    +1
     
  8. DJtheAudiophile
    Sound quality. The alac sounded muffled and empty compared to the Uncompressed Wav File.
     
  9. gregorio
    Unless you've got a seriously screwed playback device, screwed to exactly the right degree, then you should be questioning your ears rather than the codec, as there is no difference in sound!

    G
     
  10. DJtheAudiophile
    Funny. Anyway, I’m not the only one that feel this way.

    https://positive-feedback.com/Issue63/aiff.htm
     
    Last edited: Jan 9, 2018
  11. gregorio
    Yes, very funny. BTW, this isn't the "I feel this way and so does someone else" forum, this is the Sound Science forum!

    G
     
  12. bigshot
    Have you considered the possibility that they both sound exactly the same and you are just hearing them differently? There are ways to make sure your personal opinions aren't based on personal biases. Here at Sound Science, we really enjoy sharing how you can determine the difference between actual sound and subjective impressions. If you're interested, we'd be happy to fill you in on how to do that.

    I wonder which Admin is the one who occasionally thinks it might be a good idea to move stuff like this into our forum? Does he think it's like throwing meat to the lions or something?
     
    Last edited: Jan 9, 2018
  13. pinnahertz
    Silly article. No actual testing was done, just fully biased subjective opinion. So....
    Being curious, I started with a .wav file of pink noise, used XLD to convert to AIF, and FLAC, then used (gasp!) iTunes to convert the wav to ALAC, then null tested each conversion against the original WAV.

    All nulled better than 100dB. If there's a difference between lossless codecs, it's not showing up...at all...in file analysis.
     
  14. bigshot
    (You might want to explain the concept of a null test to him)
     
  15. pinnahertz
    A null test is where two signals or files are summed with one inverted so the result is only the difference between them. Essentially one is subtracted from the other. It's an indicator of how similar or different the two are. Identical data produces a perfect null reading the equivalent of no audio.

    The files tested were all 16/44.1, so the theoretical noise floor of a file would be 96dB. Since the resulting nulls were over 100dB, that indicates all files contained identical data, there was no difference. The summing engine uses 64 bit floating point math, hence the greater than theoretical noise floor null.
     
    bigshot likes this.
First
 
Back
1 2 3 4 5 6 7 8 9
Page 10 of 11
11
Next
 
Last

Share This Page