Why do USB cables make such a difference?

Discussion in 'Sound Science' started by Cartma, Jul 17, 2017.
Tags:
  1. bigshot
    When people try to convince others that they hear "better" than other people, it instantly makes me think that they are looking for ego gratification. I admire people who know a lot about music who can speak about why music expresses what it does and how it is constructed. That is knowledge. Claims of being able to detect harmonic distortion at -90dB and being able to hear frequencies above the capability of redbook require independent objective verification.
     
    JaeYoon likes this.
  2. JaeYoon
    They also must have evolutionary ears beyond a normal human being too.
    :joy::joy::joy:
     
  3. Whazzzup
    apple losses cheap and plentiful for me, then get dropped in my flac alias network and stored on my external server ssd roon core as AAC 44.1 16 bit 257 kbps, and is most certainly the largest improvement in sq listening through USB to my TT. so while i may prefer high res some through this same streamer the 16 bit never sounded so good, no complaints.
     
    JaeYoon likes this.
  4. amirm
    I assume you mean the Harman room. If so, they have published an AES paper on it: A New Reference Listening Room For Consumer, Professional and Automotive Audio Research, Sean E. Olive

    In there there is this target goal:

    "A Balanced Noise Criterion (NCB) [7] of 10 or
    better was the desired target for background noise. In
    order to achieve this, a double-shell room design was
    employed as illustrated in Fig. 2."

    If you are not familiar with NCB ratings, here is a good graph that shows it:

    upload_2017-11-8_13-46-10.png

    So getting better than NCB-10 means at the threshold of audibility. Alas, final measurements were not provided at the time of that paper:

    "Background noise measurements of the room are still underway and
    will hopefully confirm our subjective impressions that
    the room is sufficiently quiet for controlled listening
    experiments."

    Even though I have been in that room multiple times, it has always been with other people there so I can't comment on quiet it was.

    The test in questions by the way don't rely on volume level as they are plenty loud and you are just listening for tonal changes.
     
    RRod likes this.
  5. RRod
    The room for the MM test was at 19dBA, and a difference was audible with a 111dBSPL peak level on this disc. In my listening room at 35dBA at best, that would be a 127dBSPL peak and yeah, not happening.
     
    gregorio likes this.
  6. bigshot
    A 19dB noise floor would probably be higher if you factor in the noise floor of the blood pumping through the veins in your ear canals.
     
  7. gregorio
    This raises a number of issues, largely covered by 1. What we are listening for (trying to identify) and 2. How we are accustomed to listening and making determinations:

    1. For example: What musicians tend to listen for is quite different to recording and mastering engineers. Musicians listen for the aspects of performance over which they have direct control/responsibility, the technical performance choices consciously or unavoidably made by the musician/s and the artistic merit of those choices. In this respect they have extremely acute listening abilities, especially as far as their own particular instrument is concerned. Indeed, the difference between what musicians and engineers are listening for is demonstrated by vastly different vocabularies. For example, musicians think in terms of pitch, engineers in terms of frequency and classically trained musicians have a very wide vocabulary (including mainly Italian terminology) to describe aspects of performance/note production/musicality which are of little direct concern to engineers. This is why historically a Producer was required, to effectively to bridge the gap between engineers and musicians. Having trained as a classical musician from my early teenage years, studied in a renown music conservatoire and then been a pro orchestral musician for a number of years, I assumed my listening/analysis abilities would be better than most pro engineers but when I switched over and became a pro engineer myself I found that although my assumption was correct in some respects, there were many other respects/aspects which I was unaware even existed and in those respects my listening/analysis abilities were therefore no better than an average, untrained member of the public.

    2. As an engineer, I am accustomed to working/analysing in a certain way. For example, focusing on some effect/artefact, locating the points in the audio file at which that effect/artefact is most obvious, repeating/looping those points, isolating (solo'ing) individual channel/s, making qualitative judgements and adjustments and then pre-rolling the full mix to judge those adjustments in context. That's not how most musicians tend to work/analyse, typically because they do not have the tools to do so.

    I downloaded the Harman test, just to see what it was testing and how difficult it was. I only had limited time, did it on my rather old laptop with an old pair of stock Apple ear buds, listened to the first trial of each test and picked 3 of the tests to go further and judge difficulty. Up to and including level 5 I found it easy, picking the correct answer without fail within a couple of seconds or so. At level 6 (with the "Peaks/Dips" test) I did a couple of trials and got the correct answers but I needed more than a few secs, so I stopped at that point. I gave up during level 5 of the "Reverberant" and "Noise" tests, again for time reasons due to the increased number of items to compare.

    My observations:
    A. Given more time and running the tests in my studio, I think I would probably be struggling at about level 8 and not progress further than level 9, although I'm obviously only guessing at the magnitude of differences between the higher test levels with the noise and reverb tests. Nevertheless, I'm fairly certain I would be unable to reach the minimum Harman requirement of level 12!
    B. I'm not at all surprised many musicians failed, going back to point 1 above, some/most of those tests are not really what musicians have trained their hearing to detect. I'm also not surprised that audio engineers failed, at least partially due to the format of the test being so different to how we are accustomed to working/analysing (as explained in point 2 above).
    C. I'm sure I would benefit from practising and improving my Harman level but I'm not sure how much. How much time/effort would it take and how much of my improvement would only be in terms of improving my Harman test score, greater familiarity with the test format and songs (where in those songs the various artefacts are most noticeable)? Looking at it another way, if it were possible to create an identical level 12 test with completely different songs, would a Harman "trained listener" still be able to pass or would their level be reduced by their lack of familiarity?
    D. Following on from the last point, I'd be interested in the results of some further tests. For example, how would Harman trained listeners' level be affected if they didn't know what they were listening for? IE. A test where the test level could simultaneously be altered for all of the 13 different band and attribute tests but the listener would not know which of those tests her/she was listening to. Without knowing what aspect of the song and which artefact to focus their concentration on, would that trained listener still be able to pass level 12? Another example: If instead of trying to identify which of 12 EQ bands one peak or dip belongs, I wonder how they would do in the same test but with say two or three simultaneous peaks and dips instead of one? Obviously that's a more difficult test but it's possible the gap between Harman trained listeners and audio engineers would be smaller as audio engineers are more accustomed to this sort of scenario.

    I'd really like to answer this fully but don't have the time, except to mention one point: Again, there are two sides to this coin. I take it you are aware of the McGurk Effect and that we can easily hear a difference where in fact there is none. Extrapolating this, the "bar of fidelity" could be (erroneously) argued to be infinite, a fact exploited by many audiophile manufacturers. The other side of the coin, there are those who can hear artefacts which exceeding few others can but it requires an entire set of specific conditions. It's also been demonstrated that some people can hear up to about 25kHz but again, only certain people and only under specific (and potentially dangerous) conditions. In real life, how likely are we to find these people AND encounter the required entire set of specific conditions?

    1. To be aware of something only subconsciously, you still need to be able to hear it! Much of my work is directly in this area, I'm deliberately trying to affect the audience's perception without them being consciously aware of it and the levels we're talking about are very significantly above the thresholds of audibility.
    2. As mentioned by others, there is the audiophile myth that more money gets you a better performing product. While that is typically true up to a point, in the audiophile world as the price increases the performance either effectively stays exactly the same or in a surprising number of cases actually gets worse once we get beyond that point and, that point is often significantly lower than audiophiles realise (or are willing to admit)! You yourself, earlier in this thread, posted an example of a $150 USB 2 audiophile cable which actually failed to meet USB 2 specifications.

    G
     
    Last edited: Nov 9, 2017
    JaeYoon likes this.
  8. theorist
    Gregorio, just for the record, I think you are mixing me up with someone else here, as I have made no post about a $150 USB cable that did not make specification, but I take your point.
     
  9. gregorio
    Quite possibly, if so, my apologies.

    G
     
  10. theorist
    No problem, I do enjoy your pragmatic contributions to the threat.
     
  11. Whazzzup
    What the heck is going on with ferrite cores. I mean you just spent huge doe on dave 14g usd and then 10 g usd on blue 2 and these guys are talking about using 2-18 ferrite cores on a usb cable to make it sound better? I find my server works amazing well and no need for a ferrite at all? I mean you just spent the cost of a medium sedan and need ferrites
     
    Last edited: Nov 9, 2017
  12. amirm
    When we come to these forums and dispense advice, we have no way of knowing who is reading it. There is by definition a spectrum of listeners. Therefore it makes sense for such advice to encompass the population at large. This is what I call achieving channel transparency.

    There is also a second reason for my recommendations: when a device fails to achieve the specs of the silicon they are using, and others can get there with little to no price premium, we need to rally around the companies that do good work. This is what I call engineering excellence.

    You can say you care about one, both or neither. What I don't have any patience for is to say no one should care for either as Bigtop did. That doesn't sit right with me. It is saying let's not give people the information. To what end I don't still know. (Actually I do know but it is a darn bad reason).
     
  13. bigshot
    My name is Bigshot. Why are you so passive aggressive to the people around you? That's no way to become liked and respected.

    As I said before. I only care about the quality of the sound human ears can hear for the purposed of listening to recorded music in the home. I totally understand that some people want stereos that far exceed human hearing ability, and that's fine for them. I'm sure there are people who own race cars and only drive the speed limit. But my system has a purpose- and that is to reproduce music. If it does that perfectly for human ears, I'm satisfied and I can move on to focusing on music, not worry about abstract numbers on a page. I reward manufacturers by supporting the ones that make equipment that does the job cleanly, efficiently and inexpensively. I don't need to reward companies for producing sound that no human on earth can hear.

    I'm simply stating facts to support a clearly stated logical point of view here. You may not share that point of view, but you can refrain from taking side swipes at me. It doesn't make you look good.
     
    Last edited: Nov 9, 2017
  14. castleofargh Contributor
    different people can have different interests. let's not argue about what people should want.
    the only real issue we have here is the amalgam that sometimes can be made between existing differences(as in measurable), against audible differences with the best conditions, best test signal while absolutely focusing on the known issue, against somebody noticing a problem while enjoying a song casually. those 3 are pretty much never similar. more often than not they are several magnitudes apart, so IMO it's important to always be clear about what is being discussed. I personally welcome measurements even if to demonstrate that we shouldn't have to care because of how low they are. information is information and we really can't brag about having too much in amateur audio. those who don't care just jump to the post after and all is good.

    with USB cables(sorry for going back on topic^_^), I'm afraid that some anecdotal defects and poor testing methods have given many the idea that the magnitudes to expect from changing cables are huge and clearly audible. when IMO if 2 USB cable are audibly different, then one of them is bad, and/or the DAC is bad and the cables just reveal the issue by chance. it's absolutely not right to think about "tuning" the sound with USB cables.
    it's certainly not that USB cables cannot create an audible impact, it's that they shouldn't in any normal environment with ok gears. for those sitting on a cellphone tower, maybe special measures are needed, but let's not mistake some extreme circumstances for the norm.
     
  15. bigshot
    If I was building a recording studio, I would worry about the extremes. But for a portable DAC or DAP or a stereo system in my living room, that sort of stuff is irrelevant. Even if I was building a recording studio, I'd use Monoprice USB cables. Anything above that is just a waste of money.
     

Share This Page