Why do USB cables make such a difference?

Discussion in 'Sound Science' started by Cartma, Jul 17, 2017.
Tags:
  1. bigshot
    OCD, autistic obsession and paranoia drive a lot of hobbies. Not just audiophila. I'm always amazed when I see people focusing on making sure that every one of their blu-rays is lined up on the shelf in alphabetical order, or worrying about not having one or two songs out of Frank Sinatra's output, or hauling around big ass WAV files on their phone because they can't sleep at night worrying that smaller file sizes might be throwing out something they might want. None of that stuff makes sense. I understand that people with these sorts of problems can't help themselves. I try to calm them down with facts and logic, but if that doesn't work, I just let them spin their wheels. If they keep their obsessions to themselves and don't try to visit them on other people, it's fine.
     
  2. amirm
    And foobar ABX fully qualifies for that. Indeed it is the "gold standard" in online forums even from the most diehard objectivists. It is why I run it, not because it is my favorite way to do it.

    Since you have no familiarity with it, in foobar ABX, you are given the two files and a random version which you have to vote as A or B. You as the listener have no control whatsoever on what is being played. And the computer of course is completely unbiased. It will select A or B randomly as "X" and no way, no how can you game that aspect.

    And *every* test and outcome I have posted has been an online challenge created by others. None were created by me and I had nothing to do with selection of content.

    Again, spend five minutes running it and see if you can game it. You will be heavily frustrated. When I can't tell the difference, the results are really, really poor. You can't get lucky as you have to achieve 95% confidence and that can't be done except by remote chance.

    Here is an example of me failing to hear a difference in Entreq grounding box:

    foo_abx 1.3.4 report
    foobar2000 v1.3.2
    2016/02/14 08:50:25

    File A: C:\Users\Amir\Documents\Test Music\Entreq 2 digital\test_4_output_entreq.wav
    File B: C:\Users\Amir\Documents\Test Music\Entreq 2 digital\test_4_output_no_entreq.wav

    08:50:25 : Test started.
    08:52:22 : 01/01 50.0%
    08:52:30 : 01/02 75.0%
    08:52:43 : 02/03 50.0%
    08:52:51 : 02/04 68.8%
    08:53:03 : 02/05 81.3%
    08:53:32 : 02/06 89.1%
    08:53:58 : 03/07 77.3%
    08:54:12 : 03/08 85.5%
    08:54:27 : 03/09 91.0%
    08:54:31 : Test finished.

    ----------
    Total: 3/9 (91.0%)

    ====

    So 91% chance that I was guessing even though I got some right.

    Here is another online challenge that I failed at initially:

    foo_abx 1.3.4 report
    foobar2000 v1.3.2
    2014/07/19 07:27:54

    File A: C:\Users\Amir\Music\Arny's Generational Loss\sb20x_original.wav
    File B: C:\Users\Amir\Music\Arny's Generational Loss\sb20x_pass1f.wav

    07:27:54 : Test started.
    07:28:25 : 00/01 100.0%
    07:28:40 : 00/02 100.0%
    07:28:52 : 01/03 87.5%
    07:29:04 : 02/04 68.8%
    07:29:17 : 03/05 50.0%
    07:29:38 : 04/06 34.4%
    07:29:50 : 05/07 22.7%
    07:30:35 : 05/08 36.3%
    07:30:46 : 05/09 50.0%
    07:31:02 : 06/10 37.7%
    07:31:15 : 07/11 27.4%
    07:31:27 : 08/12 19.4%
    07:31:43 : 09/13 13.3%
    07:31:52 : 10/14 9.0%
    07:32:04 : 11/15 5.9%
    07:32:15 : 12/16 3.8%
    07:32:26 : 13/17 2.5%
    07:32:39 : 13/18 4.8%
    07:32:54 : 14/19 3.2%
    07:33:06 : 15/20 2.1%
    07:33:18 : 15/21 3.9%
    07:33:41 : 15/22 6.7%
    07:34:08 : 16/23 4.7%
    07:34:16 : 17/24 3.2%
    07:34:26 : 18/25 2.2%
    07:34:38 : 19/26 1.4%
    07:34:51 : 19/27 2.6%
    07:35:05 : 19/28 4.4%
    07:35:28 : 19/29 6.8%
    07:35:39 : 19/30 10.0%
    07:35:41 : Test finished.

    ----------
    Total: 19/30 (10.0%)

    ===

    That says there was 90% chance that I was NOT guessing. But that is not acceptable to me. If you look toward the end, I started to lose what I thought was a definite difference as failures started to pile up after getting 19 right.

    You would do well putting aside that negative energy and insulting remarks and actually run some double blind tests with a "D." Heck, run it single blind and see if you can pass with the tricks you mention. If it were easy, tons of people would do it.
     
    Last edited: Nov 7, 2017
  3. amirm
    All commercial content? Forget that. Tell me how you determined the effective bit depth of music in all the content in your library. How did you objectively analyze them? Or is this another "everyone knows" that when we challenge, has no answer?
     
  4. bigshot
    I'm sorry, I'm sure you are a very nice person in real life, but I don't trust what you say any more. No offense, it's just that you seem to be putting on an act for the forum and I don't think you're being completely honest with us. I'll just leave it at that.

    Gregorio is a professional sound engineer. He knows exactly how music is produced and engineered. Of that I have no doubt.
     
    Last edited: Nov 7, 2017
  5. JaeYoon
    I think a lot of audiophiles who cannot sleep without WAV never want to read up on hard work of LAME developers or prowess of QAAC today.

    In their mind it's throwing away data they worry they might want. But it may not be perfect, but a lot of work has been done trying to remove all inaudible information and keep pretty much as much as possible or everything that is audible depending on bitrate that a person can find transparent to them.
    Like some people might find 128 kbps Lame MP3 not always transparent, but 128 AAC VBR might even do it.

    I keep my FLAC at home, and portable devices use lossy. When we are outside and walking about I think those details are last thing we are thinking about. Usually it's other things on our mind, like where we are traveling to, what do we want to do outside, etc.
     
    Last edited: Nov 7, 2017
  6. bigshot
    I rip everything to AAC 256 VBR and never look back. The CDs get boxed up and stored in the garage.
     
    Darren G and JaeYoon like this.
  7. gregorio
    That's a bizarre question to ask, in light of what I've already said on the matter. Firstly and most importantly, it's an avoidance/deflection from the question/s I asked! And secondly, as I'm sure you are aware, there is no objective measure of dynamic range. The answer to your question depends on how we define dynamic range in the first place and if we're going to talk about fidelity, on the intention of the creators.

    You appear to think that I'm trying to discredit you and/or the science, I'm not. It's important to know the ultimate limits of human capabilities under all conditions and that's what science attempts to give us BUT, if you're going to bring real life into it, then the reference has to be all real life conditions rather than all test conditions which do not all occur simultaneously in real life, unless we manufacture them for testing and in real life, commercial content is designed for entertainment purposes, not for the purpose of testing the utmost limits of human capabilities. Indeed, if we define dynamic range as do musicians and if we actually used even say an 80dB range for that purpose, we'd have incompetently produced content! However, the fact that you are taking test figures and erroneously calling them real life figures is besides the main point because my question to you, the one which you apparently refuse to answer, is based of your test figure!

    Amirm, I am trying not to fight with you because most of what you say and do I not only entirely agree with but really appreciate! With such a paucity of reliable information, it's certainly a very worthwhile endeavour to highlight equipment performing poorly relative to it's marketing claims and/or price. However, that's not the only affliction which needs addressing; as technology not only improves but gets cheaper, we've also got the growing problem of over-performance sold as actual audible improvement and not just with some DACs but with the digital audio formats themselves. Therefore, it's also a worthwhile endeavour to highlight when this is occurring!

    G
     
    ev13wt and JaeYoon like this.
  8. amirm
    I hear you. This is one of those things that is hard to believe, i.e. that someone can hear distortions that we ourselves think is an impossibility. There is fortunately science behind this but one that is not well known outside of industry and research.

    The key here is understanding the concept of a trained listener. No, I am not talking about self-appointment "golden ears." I am talking about listeners that have gone through training and verification of their listening ability to be well above general public and that includes audiophiles. A great published example is that of work for Harman in their loudspeaker research through controlled listening tests. See this example: http://seanolive.blogspot.com/2008/12/part-2-differences-in-performances-of.html

    [​IMG]

    See how much better the discrimination and ability to identify artifacts are among trained listeners compared to other groups. Assuming that high-end audiophiles are represented by the "audio reviewers" the ability of trained listeners is about 5 times better!

    How did that magic come about? As I mentioned, it is through specialized training. Fortunately Harman gives away that software and you can try to a) test your own abilities and b) become trained: http://harmanhowtolisten.blogspot.com/

    The software is like a video game in that it starts easy and progressively gets more and more difficult. The concept is simple: you are presented with an original track and another that has been modified by a single filter/EQ changing the frequency response. When the "Q" (bandwidth) of the filter is large detection is easy as it impacts so many notes. But as it gets narrower, less of the music spectrum is impacted and the job gets hard.

    I remember running that test when it first came out. I got to level 2 or 3 and could go no further. Pretty disappointed. :) So I practiced some and then gave it up. Fast forward a year or two later and I was at Harman with Dr. Sean Olive presenting us the same test in one of their reference testing rooms. Here is a picture of Sean and you can faintly see the How to Listen software output on the projection screen:

    Harman Reference Room - small.jpg

    What you don't see is behind me which is a dozen high-end audio dealers some of whom are acousticians who had come for this training.

    Sean started the program at level 1 and everyone got the answer right more or less to level 2-3 as I had originally. But the moment it went above that, I and Sean where the only ones who could guess correctly. I think I got to level 6 before I failed. That bit of training had made a noticeable difference in my abilities. My happiness on that front disappeared when Sean continued with utmost ease giving the right answer to levels 7,8, 9, etc. I mean he was not even trying! I asked him what the minimum requirement was for someone to become a trained listener and he said level 12. :eek:

    After the event finished, I had all of the dealers come to me and ask me how I did what I did. They were amazed at it and to them it was like magic. From then on, and to this day when I run into them at shows, they consider me to have great ears even though I consider myself a failed attempt compared to trained listeners at Harman.

    Now, the tables would be turned if we were talking about hearing non-linear distortions. Again, another story. :)

    The year is 1998 and I am at Microsoft and was given the audio/video codec team to manage. I figure I better learn what they are doing and compress some files into 128 kbps MP3. After some 30 years of being an audiophile I thought it would be a walk in the park to hear artifacts in it. To my shock and horror, I found the sound to be as good as the CD! I could not believe the outcome. Here we had 92% of the file thrown away and the other 8% sounded like the original! So started a journey to learn not only to hear compression artifacts but also learning to be a "critical listener." What is a critical listener? I can look past the music and focus on it as an instrument would. I can isolate fidelity aspects much like someone lip reading can understand what you are saying without sound. Slightest changes are important and I know how to find and focus on them.

    After some 6 months of non-step trial and error I all of a sudden realized that I could artifacts that others could not. And I became part and parcel of our signal processing team in evaluating any major changes to our audio algorithms. Those skills have remained with me and as I mentioned translate to finding small differences that others cannot.

    Now you might say why we should care what a few trained listeners can hear. Well, unfortunately that is not the end of the story, pun intended. It turns out there are individuals in general public and audiophile circles that have the same critical listening skills. We were working with one of partner companies that was doing large encoding jobs for our end customers (electronic music distributors). One of their technical people was complaining about an artifact which we could not hear. He was local so we invited him to Microsoft. He comes over with his track and plays it for me and my codec team. We all look at each other and can't hear any issues. In amazement, he kept asking, "can't you hear this high-frequency artifact???" And our answer was no! :)

    Being an important customer meant that we had to investigate it anyway. The codec team went to work and actually found and fixed the problem, proving objectively that he was wrong. And sadly for me, his ears and brain were better than mine. We hired him right away to work in our codec quality verification team. :)

    Bringing us full circle, you are absolutely right that the types of distortions we talk about elude vast majority of listeners. Maybe all but a few. Having seen so many audiophiles, musicians, professional recording and mastering engineers fail these tests, I am well aware of difficulty of hearing such artifacts.

    That is not the topic though. The topic is what standard bar of fidelity do we set that when we talk to someone else, we are confident they can't hear any artifacts there. In this instance, you questioned me and as I have demonstrated over and over again during many years, I am able to hear some of these artifacts. So I sure as heck am not going to brush the problems under the rug. Seeing how producing quality products doesn't cost much, that is what we need to do as a group.

    What we don't want to do is go around and accuse people as corrupt when we haven't done our homework to understand whether they have the capability to hear things that we may not have.

    Finally, let me say that as I have gotten older, I have lost a lot of my high frequency hearing. So I am probably nowhere as good as I once was. And others may be able to hear much better than me if they still have their full spectrum hearing. So let's make sure there is a nice safety margin beyond what I can hear. Again, it doesn't cost much of anything to do that.
     
    ev13wt likes this.
  9. bigshot
    The reason I don't have confidence in your reports of your test is because I think you are being disingenuous with us. It has nothing to do with science. If you'd like suggestions on how you might avoid that in the future, I would be happy to provide some.
     
    Last edited: Nov 8, 2017
  10. theorist
    I claim no expertise here on audio science or trained audio listening. Unlike Gregorio or Amirm, indeed, I would probably have difficulty testing correctly above the Harmon band one, if even that, no matter that I have been an audiophile for 40 or so years. But even if I cannot consciously identify that there are distortion artefacts present in the recorded music or play-back system that I am listening to, yet alone what they might be, I suggest that unconsciously I perceive these distorts, at least at their higher levels, so that this distortion gives me a sense of conscious unease or dissatisfaction with what I am listening to. That is why I want to listen at least to Red Book standard recordings rather than MP3 and am willing to pay for a four figure DAC over a $29 one, although I never say no to a bargain!
     
  11. bigshot
    If you are subject to unconscious influences, it’s more likely to be bias than it is sound you can’t hear. A controlled blind test could remove that factor and let you know for sure where your subconscious influence is coming from. You might not want to know though
     
    Last edited: Nov 8, 2017
    Darren G and JaeYoon like this.
  12. JaeYoon
    Yeah! Testing from end user is important

    another important thing to state is, money does not guarantee distortion free products.
    without self-user testing, there is no proof the companies who sell you a four figure dac is outright lying, and also has distortion in their DACs. Let that sink in, IT MIGHT have distortion too. Just because it cost a lot of money, does not guarantee anything.
     
  13. castleofargh Contributor
    what I read is that your biases and preconceptions create insecurity which in turn make you uncomfortable until you move up into what you assume to be the more reassuring stuff. it's simple, known and felt by all of us at some point. and it really explains a lot. couldn't you satisfy yourself with that reason? because when I think of some far fetched hypothesis about the subconscious becoming conscious anytime preferences are involved, and conveniently subconscious most of the time a controlled test is involved, I don't see much rational thinking or any evidence telling me to explore that idea. instead I see that it would be pleasing if my preferences were objectively right. and that weird idea would nicely lead to that conclusion. which is IMO more about self gratification and confirmation bias than it is about a plausible hypothesis.
    I'm not a big fan of Ockham's razor, but in situations like that, it's hard not to think about it. I fear that something won't be enough, so I ruin myself with stress and doubt. and when I remove that cause of doubt in my head, I feel much better. the story or every human being. ^_^
     
    JaeYoon likes this.
  14. RRod
    @amirm What were the noise levels in that room?
     
  15. ev13wt

    Of course, with training and dedication to a certain slice of something, lets stick with "codecs", for 10 hours a day, for multiple years, you should be quite competent at listening.
    Being able to tell one codec from another is a piece of cake. (Edit: 1990s)

    In real world conditons however...

    Don't get me wrong, I basically agree with what you are saying. But to argue benefits to a consumer over those differences is a challenge. They hear more differences in monster cables vs. zip cord and beats by dre vs. included in the box.
     
    Last edited: Nov 8, 2017
    gregorio, Darren G and RRod like this.

Share This Page