1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Testing audiophile claims and myths

Discussion in 'Sound Science' started by prog rock man, May 3, 2010.
782 783 784 785 786 787 788 789 790 791
793 794 795 796 797 798 799 800 801 802
  1. bigshot
    Keith has already done a listening test comparing lossless to lossy. He couldn't discern a difference.
  2. KeithEmo
    Actually, as I recall, even using content I wasn't familiar with, I DID score significantly above random.
    However, I'm not sure of the relevance either way.
    Even assuming that I couldn't tell the difference with a certain few files, that wouldn't prove that this would ALWAYS be the case.

    This is all moot, however, because, when applied to digital data, both "lossy" and "lossless" have very specific technical meanings.
    A file that has been losslessly compressed, then losslessly uncompressed, will be IDENTICAL to the original.
    By definition, a file that has been LOSSY compressed, then uncompressed, will NOT be identical to the original.
    Therefore, one is a known quantity, while the other is not...

    Now, from the point of practical usage and philosophy......
    At that point you are left with the question of whether the lossy copy will be close enough to be "indistinguishable from the original to you".
    Bear in mind that, with the lossless copy, there is no question - because we KNOW it is identical to the original.
    Personally, given the choice, I will always choose something that I KNOW is perfect over something that "might be good enough"...
    And, since the cost in terms of extra storage space is negligible, I see no reason not to do so...

  3. Sgt. Ear Ache
    the cost in extra storage space isn't really negligible.

    I've taken a couple tests and could reliably discern a difference up to 192kb mp3s. 256 and up sounded perfect to my 52 year old ears. Spotify's "extreme" setting is golden afaic. :)
    Last edited: Jan 8, 2019
  4. KeithEmo
    I guess "negligible" is relative.

    The entire collection of albums that I actually listen to, in lossless FLAC format, fits on a USB hard drive that cost me $129 .
    Reducing the size of the files, even significantly, would save me at most a few cents per album.
    Or, looking at it another way, my entire audio collection takes up as much space as a dozen high quality Blu-Ray movies, or one or two 4k movies.
    (And, for portable users, a 128 gB USB stick is now less than $30 )

  5. Phronesis
    Agreed. Money spent on storage is a pittance compared to what people spend on gear chasing sonic nirvana.
  6. Sgt. Ear Ache
    expense isn't really the issue. I just have limited space on my Iphone which is the source for a fair bit of my listening. If I were ripping a large collection onto my laptop or something I'd not worry about it...although frankly, since I can't tell the difference anyway I'd probably still go with 320kb mp3.

    I also do my best to not spend ridiculous amounts chasing sonic nirvana. I have pretty reasonable, real world expectations and I keep the expenses down. I won't spend crazy money on negligible sonic improvements...
    Last edited: Jan 8, 2019
    sonitus mirus likes this.
  7. bigshot
    Wasted space is wasted no matter how economical it is. If you can't hear it with human ears, it doesn't matter. I have a massive library that would be a royal PITA to manage lossless. AAC 256 VBR sounds identical to lossless and it works perfectly in everything, from my main speaker system to the limited space available on my iPhone. Why bother with bits you can't hear?
  8. Sgt. Ear Ache
    yeah I have a bit of a philosophical issue with the whole law of diminishing returns aspect of things. I don't like spending 4 times as much (whether it be money or data space or whatever) for a 3 or 4% improvement. It feels like spending $80000 on a car that - compared to the $20000 version - has 5% more horsepower and a set of designer floor mats.
    Last edited: Jan 8, 2019
  9. KeithEmo
    Obviously this is going to depend on exactly how you plan to use something....
    HOWEVER, I dispute your absolutely general statement that: "If you can't hear it with human ears, it doesn't matter".
    In fact, all you can say for certain is that: "Using your current equipment, and your ears, you cannot hear any difference."
    Therefore, at most, you may reasonably generalize your results to others who plan to listen to those files on similar equipment, in similar ways, using similar ears.

    Let's try a simple example.

    Dolby PLIIx is a surround sound synthesizer that is featured on virtually every recent surround sound processor produced in the past decade.
    (DSU, the Dolby Surround Upmixer, is the new version which replaces it on all Dolby Atmos enabled processors.)
    Both synthesize surround sound from stereo content by deconstructing the audio content and routing it to various channels based on things like relative amplitude and phase.
    (So, if you alter the signal in any way that's significant to the decoder, you change its output.)
    Likewise, there are several "headphone processors" that utilize phase relationships contained in the content to exaggerate separation and produce other "spatial effects".

    I know a few people who prefer surround sound, and who play most or all of their stereo music through their home theater processor, using PIIx to turn it into surround sound.
    Have you actually confirmed that, when you apply PLIIx to a lossless file and an AAC file, you get exactly the same result every time?
    Have you also confirmed that they produce exactly the same audible output using the new Dolby Surround Upmixer?
    And, have you confirmed that they produce the same result with various "headphone processors"?
    If not, then, while you may reasonably assert that "they sound exactly the same in stereo".....
    You really have no idea whether the lossy encoding has significantly altered the way those files will play through one or more of the commonly used surround sound decoders.
    (Yet we do know that, since lossless files actually deliver an exact reproduction of the original, they will produce the correct result.)

    This is NOT some sort of obscure scenario....
    MANY people listen to their stereo music using a surround sound decoder.

  10. KeithEmo
    In that case I would agree with BigShot.
    If that's the only place you're going to listen to the files - then all that matters is how they sound on your iPhone.
    (I would also say that applies to streaming, where all that matters is how the files sound today, because you can't save them anyway.)

    HOWEVER, if it was a choice between paying $12 for MP3 files and $15 for lossless, I would still:
    - pay the extra $3 for the lossless version - just in case it makes a difference somewhere down the road
    - save a backup copy of the lossless file somewhere
    - make a good quality lossy copy to put on my phone

    There's a sort of saying that is often repeated among photographers (and I consider it to apply to audio as well)....
    - if you save a copy at the best quality, and find you don't need it, you can ALWAYS make a more compressed copy at lower quality to use today
    - however, if you save a copy at lower quality, then find out you need a higher quality copy later, you're probably screwed

    That's why photographers always save the negatives... :beerchug:
    The reality is that, when purchasing music, the difference in cost between lossless and something like MP3 usually isn't all that much.

  11. bigshot
    With lossy and lossless the space overhead is significant and there's absolutely no sound quality advantage at all. You're only getting the audio equivalent of packing peanuts padding out the file sizes. "HD" audio is even worse for that.

    With KeithEmo's ears and KeithEmo's equipment, there is absolutely no audible difference between high data rate lossy and lossless, so for him it's pointless. Any extra cost involved with lossless files is wasted. It would be better used on getting more music.
    Last edited: Jan 8, 2019
  12. sonitus mirus
    That is only true when a CD is still widely available for sale everywhere. In fact, on Amazon, today, if a new CD is available for sale, that is typically cheaper than the mp3 version, and with a Prime membership, the delivery is free and overnight.

    The problem is when I run into something I want to hear, but is not available to stream on music services, or if something has only been released in a Hi-Res format.

    The Grateful Dead Studio Albums Collection, as an example, is not available in a CD format nor is it available at 16/44.1. I can get it in a lossy format for $94.99 (Amazon 256kb mp3) or I can splurge on a 24/96 lossless version for $199.98 on HD Tracks.
    Last edited: Jan 8, 2019
  13. KeithEmo
    I'm sorry, but your bit of hyperbole about "packing peanuts", while colorful, is also technically untrue.


    When you use a lossy encoder, the encoder deliberately discards information which it considers to be "perceptually unimportant".
    (It discards whatever information it decides can safely be discarded "without producing a noticeable difference" - based on the criteria chosen by whoever programmed it.)
    It may in fact be true that YOU have no use for that information... or even that many others have no need for it either...
    However, it is also true that information present in the original file has been discarded, and the contents have been altered.
    (Anyone who doubts this should open copies of both in their favorite audio editor and look for themselves.)

    I would also really appreciate it if you would stop misrepresenting the other facts as well.
    On a specific test, consisting of certain files provided by BigShot....
    - I was able to easily recognize the difference between the lossless version and some low-sample-rate lossy versions
    - and I was not able to recognize the difference between others beyond what would reasonably be expected by random chance

    HOWEVER, unless you are willing to accept that MY ability to recognize the difference between a few specific files, on a certain few pieces of equipment, on a certain afternoon, comprises "the limits of human hearing ability and experience".....
    Then that doesn't seem especially significant in the big picture.
    (In fact, it almost seems like a single, isolated, almost anecdotal result...... hmmmmm....... )

  14. KeithEmo
    Then you have a dilemma.
    When a CD is available I would always buy that... and make my own MP3 if I wanted one.
    I should also note that, if you purchase a commercial MP3, you are trusting whoever encoded it to have done an optimum job.
    MP3 encoders are not standard, and there are many different options, so they are far from equal.

    As for the Grateful Dead Studio Remasters...
    I can state that I purchased the 24/192k version of that set - and it sounds very good.
    (I would go as far as to say that I own several different remasters of some of the albums included in the set - and the versions in that set are by far my favorites.)
    It has also been heavily remastered and remixed - which may well account for all or most of the difference.
    I guess it's up to you to decide whether you trust the MP3 version to sound as good as the 24/192k version.

    Last edited: Jan 8, 2019
  15. bigshot
    Sound you can't hear doesn't matter. It can be irretrievably lost and it still doesn't make a lick of difference when you sit down to listen to music.
782 783 784 785 786 787 788 789 790 791
793 794 795 796 797 798 799 800 801 802

Share This Page