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

24bit vs 16bit, the myth exploded!

Discussion in 'Sound Science' started by gregorio, Mar 19, 2009.
344 345 346 347 348 349 350 351 352 353
355 356 357 358 359 360 361 362 363 364
  1. gregorio
    1. What exactly did I change?

    2. What reason, apart from a better production (or better master), could there be for it sounding better, baring in mind that the whole point of this thread is a simple explanation of the science/facts for why 24bit does not sound better than 16bit?
    2a. It would be a reasonable question, if there were a reasonable alternative (to being due to a better production/master). The only reasonable alternative is that two tracks (at say 24/96 and 16/44) are in fact the same but one is perceived by some people as better because of some expectation/cognitive bias but then, there can be no reasonable answer to your question, because you'd have to DBT a significant number of people with most of the available 24bit and 16bit versions (or all of them if you want a precise answer) and it's not practical/reasonable to do such a test/study.

    3. I can't help that, sorry. Maybe if your question were more reasonable or you reworded it so it made more sense, then I could provide you with an answer that would make more sense to you?

    Last edited: Dec 4, 2019
  2. castleofargh Contributor
    While I understand the attraction for a ready made test file, I can only suggest to convert files yourself.
    I see a few reasons for that:
    - you get to pick songs you know very well, and that could probably improve your chances to notice a change.
    - You get to test your own conversion method. So if you find no audible difference, you gain confidence that your way of doing it is at the very least good for your ears. I personally find that reassuring given how many albums I've converted to 16/44 or lossy formats to put on my DAP.
    - You make sure that nobody messed with the files to push their own agenda.

    If whatever tool you use offers to apply dither in some way when converting from 24 to 16bit, you probably should do it. Chances are that the file will still be audibly transparent without any dither, but why take a chance?
    Last edited: Dec 4, 2019
    Daiyama likes this.
  3. taffy2207
    Let's not forget the monetary aspect as well. A lot of Audiophiles have invested heavily in Hi-Res Music. It's a bitter bill to swallow for them to admit that they could be wrong about it and thus potentially wasted their money.

    It doesn't just apply to Hi-Res Music, unfortunately. It applies to gear as well, particularly people who bang on about 'endgame' set ups.
    Last edited: Dec 4, 2019
  4. board
    @gregorio might be able to answer your question if he sees your post, as he and nick_charles talked about music samples on the same page (page 3) of this thread. However, nick_charles erased his post, so I couldn't figure out if they were talking about the samples provided by Soundkeeper Recordings.
    So, I downloaded the samples from the first songs Soundkeeper provided (Minis Azaka) and looked at them briefly.
    I did the following:
    * Using SoX resampler in Foobar (quality: best; passband: 95 %; phase response: 50 % (linear)). I downsampled the hi-res files to 16/44.1 and then I also upsampled the 16/44.1 to 24/96.
    * Then using the dynamic range meter I analyzed the three original files, which all showed identical numbers:


    Then I used Voxengo's plugin CurveEQ in Audacity to compare two files, which had to be converted to the same sample rate, as they would otherwise show quite big differences (that's apparently how CurveEQ works), and then it became "interesting". CurveEQ shows the difference in EQ between two files (volume level doesn't matter). The orange line represents the difference between the two files, so if the orange line is completely flat, the EQ is identical (although volume levels can be different).

    24/96 downsampled to 16/44.1 vs. 24/192 downsampled to 16/44.1:


    24/96 vs. vs. 24/192 downsampled to 24/96 looks exactly the same.

    HOWEVER, when compared the original 16/44.1 to the other two files downsampled to 16/44.1 it looked like this:


    It's the original 16/44.1 file that has this incread high frequency energy.
    So, I'm wondering if the 16/44.1 file has been created with a bad sample-rate converter, which has a lot of ringing around 20 kHz. Or could it be something else?
    Also, it's not for me to say if this was done intentionally by the label, or they simply used bad conversion without knowing it and then concluded (if this is audible) that 16/44.1 is crap.
    Last edited: Dec 4, 2019
    Daiyama likes this.
  5. board
    I see your point, and I suppose I agree, but if you and I (and others) don't want to convince group 1, why are we trying to argue with them?
    Why are we not just saying "please back up your statements with an ABX test and until you do we won't respond to you or engage in any kind of argument about this" and then simply stop responding to these people?
    Certain scientists refuse to debate with creationists because they believe that engaging in a discussion gives some credibility to the creationists, meaning that by participating the scientists show that they take the creationists seriously.
    Bill Nye did debate Ken Ham on the issue, and although I've only watched very short clips, then I think it ended up being a fun debate for the spectators, so ...
    Last edited: Dec 4, 2019
  6. Daiyama
    Thank you all and especially board for his extensive research.
    I will take your advice and will look through my music library and see which 24/96 tracks I have (surely less than 1%) and will downsample this by myself.
    I have foobar2000 (no special converting plugin), is this sufficient? (and audacity to check).
    And also I do not want to be the hundredth person to whom you explain all that.
    Is there a link to a page where most of this stuff is easily explained?
  7. castleofargh Contributor
    Intuitively, we would expect that. But in practice it's a lot messier. The best argument in favor of getting the hires file is often the worst reason of all. That a sound engineer was asked to make a different version for the different formats. While that was completely justified for vinyls vs digital because of the physical limitations of vinyl, it makes no sense for various digital resolutions. And yet it's not uncommon. In fact I'd argue that pretending to have audibly improved resolution was such a fail that they have no other choice but to manufacture the audible difference.
    I've tried to get some understanding of this and watched/read as many interviews of sound engineers on this topic as I could. A few argue that the hires version will be purchased by an elite with better more neutral gear and a taste for more dynamic more blablablah elite audiophile flattery. While the rest of the pleb will listen to the album in their car or some cheap stereo system without extended frequency response. So the argument was to make CD versions that will still work fine with those poor uneducated people's circumstances. And this is ironically about the most reasonable explanation I've seen beside the guys who simply admit that they were asked to make different versions so they did their job. Most other explanations felt to me like the talk of crazy people or at least the talk very unscrupulous salespeople. My favorite was a guy who argued that the 16/44 conversion sounded nothing like the glorious hires master he had made, so he needed to remaster the CD version to bring back some subjective life into it.:clap::clap::clap: I'd love to see such a person try a blind test, just for the lolz.

    And that's just the tip of the iceberg where we assume that all resolution deliveries of an album will have been made at the same time by the same guy. Just think of the MQA ubermasters of death (or whatever they're called when they can be unpacked to ??/192*). We rarely know what master they took or why, we have no clue what they did with it. So does it sound different? Is it a different master? Maybe. If so, is it better or worst? I'm going to guess that the correct answer is a solid "it depends on the track". And that's been the feedback of most people over the years about the many attempts to sell us the same things in a different box for a boosted price. Some masters are certainly enjoyed by many in those hires formats. But just as well, you can find people who prefer their old CD version for some albums. Is it habit? Is it just a vague subjective matter of personal taste? Did someone murder that latest master that just happens to be released in hires? Probably a little bit of everything happens to all formats. And I suspect those who claim that a resolution or a format always sounds better, to be people who live high on placebo.

    *MQA is a good or very bad example as their super revolutionary products might have lower effective bit depth than even a CD. Maybe we should ask them to explain audibility and bit depth? ^_^
    PhonoPhi and miksu8 like this.
  8. board
    You can at least install the separate SoX resampler plugin in Foobar to convert sample rate.
    Otherwise, if you only want to change bit-depth, you can also use Foobar:
    Right-click on the track, choose "convert", choose "wave" and then in the drop-down menu there's the option of bit-depth.
    It should also be said that if you want to change both sample-rate and bit-depth you would have to go through both processes describred, if I'm not mistaken.

    Maybe some of you have already seen this, but Mark Waldrep from AIX records and Real HD Audio is doing round II of his "hi-res challenge", where he grants access to select songs of his AIX catalogue of 24/96 recordings, which he has then converted to 16/44.1. I trust him and his equipment, but the way he is doing his challenge, it is possible to cheat. But you can, however, get true hi-res files from him. As he has repeatedly pointed out, many files being sold as hi-res are actually not real hi-res, as they were recorded on analogue tape in days of yore, so they don't contain much ultrasonic content to begin with. All his AIX recordings are true 24/96 recordings.
    Daiyama likes this.
  9. board
    I completely agree. As Ethan Winer once said, people don't like to be told that their hyper expensive gear has been a waste of money. In extension of my post above about the two groups of people, the people who spend extraordinary amounts of money on their stereo system tend to build their entire lifestyle around this, so telling tham their equipment has been a waste of money is in their eyes equivalent to saying that they're wrong and that there's something wrong with them and that their entire adult life has been a joke. Just look at how almost all audiophiles react when they're being told these things about their beliefs. It's usually "the test is wrong", "some stupid test equipment can't measure better than what I heard", etc., and usually their falacious beliefs end becoming reinforced.
  10. bigshot
    I'm sorry. I misread your question too. I didn't see the "how often". I have lots of SACDs and lots of CDs. I haven't found that SACDs sound better more often than CDs do. It's pretty much a crap shoot. The best way to find out what the best sounding release is, is to speak with knowledgeable collectors who have heard all the available masterings and can point you to the best one. That is just as likely to be a CD as it is an SACD or blu-ray audio disc. In many cases, they all use the same master, so it doesn't matter which one you buy.
    miksu8 likes this.
  11. board
    Whenever possible I compare the different versions first and then I buy the version I prefer. I've sometimes seen people rave about certain releases, and when I listen for myself I find that I prefer the one that those people dispise, although there is often consensus between me and others.

    As for "24 bit versions sound better because they're better produced", if it wasn't clear by everything that has been written by various people here, then it's not the bit depth that makes it sound better, since it will sound the same when converted to 16 bit - it's simply that it's better produced to begin with, no matter the bit depth.
    I don't have so much experience with SACDs, but I agree that it's a craps shoot.
    miksu8 likes this.
  12. castleofargh Contributor
    I work in a very simple way. If I've heard one master a lot(usually discovered the album with it), anything else will feel wrong to me. No matter how clean or nice another version is, for me habit and nostalgia triumph over almost anything else. To the point that I've been looking to acquire all the old classical music from my youth(quite the battle to try and find a symphony from vague memories of what the cover looked like, and google image :dizzy_face:).So far I've very much enjoyed the ones I could find, and I honestly like them more than other better known and acclaimed versions. Even though I admit that several of those acclaimed version deserve every bit of praise.

    In the end I like my familiar stuff better. And while it's obvious with classical music where mastering is but a small piece of the puzzle, I also tend to feel that way for other genres. I discovered Iron Maiden with the "A real live one" CD, I now have pretty much all the albums(so long as Dickinson is the singer!!!!!!), and a bunch of DVDs of their concerts. so again I'm talking more than subtle remastering. But even after all these years, if a song was on "A real live one", that's the version I'll like the most. Pretty much all remasters of my favorite music tend to disappoint me.
    So I have to say, it's been a massive money saving bias. I usually hunt biases like I'm Highlander, but I'm starting to feel grateful for that one. I just hope someone will shoot me the day I drift from that to saying: "it was better before".
  13. bigshot
    Nowadays, I buy music I'm not familiar with, so I'm not stuck with the duckling patterning.
  14. old tech
    I understood that it is not the lack of ultrasonic content which means that analog recordings can never be hi res, but rather the resolution ie signal to noise and the associated dynamic range through the relevant (to young humans) 20 to 20khz frequency range - hence the 'res' in the high res... From this perspective first generation analog tape on studio equipment with noise reduction is at best equivalent to 14 bits digital. The ultrasonic aspect is irrelevant.
  15. board
    That would be the 24-part of the 24/96 equation (the bit depth). The ultrasonic part would be the 96 part.
344 345 346 347 348 349 350 351 352 353
355 356 357 358 359 360 361 362 363 364

Share This Page