USB audio cracked... finally!

[jabbr]

1 hour ago, mmerrill99 said:

It doesn't mean that this wire should therefore be left in the cable - it actually means that the 5V wire is best taken out of the cable as some USB cables have done - some run the 5V wire as a separate outside the main cable.

 

Again, I can't see any advantage of working with this capacitive coupling & leakage between the USB signal wires & 5V wire?

 

That entirely depends on the cable. In an ideal world we could model the transmitter, cable and receiver network and derive ideal resistance, capacitance and inductance parameters a.k.a. termination. When the intrinsic capacitive coupling between D+/- and VCC/GND contributes to this circuit then this coupling is intended and desirable. Similarly for inductance.

[jabbr]

2 hours ago, marce said:

A cable is a cable and their are only limited things (very) that a cable can do to a signal, without the addition of components. I have not seen any hints or information that gives any indication of what this cable is doing (and yes I have read the whole thread .).

Crosstalk checking is also readily simulated, some EMC advisers will give some basic indication, but the best results come from the 3D field solvers. Of course then there is the USB spec...

 

Yes we are playing a guessing game here. A cable is just a cable except when it isn't, eh?

Our salient hints are that:

1) the "Clarixa" is @PeterSt's cable which is designed to precisely conform to the USB spec, so, by inference the "Lush" is not so precisely adherent

2) The "Lush" does not contain "active elements" nor internal resistors 

...

leaving inductance and capacitance -- its elementary 

[jabbr]

21 hours ago, Jud said:

 

21 hours ago, pkane2001 said:

~$5M. Cash or wire transfer will be fine.

 

"Properly" would likely require university facilities (people who are accustomed to setting up science experiments that are intended to result in academic publication, which is what I would consider "properly"), and possibly fMRI facilities to research audibility without running into the "Iowa gambling task" problem.

 

 

First what is the real question? The electronics are fairly worked out and standard. As we've said can be modeled simulated & measured using off the shelf equipment. Who is paying for this?

 

Or is the question whether these electrical differences are audible?

 

Thats different. What are the limits of human audibility of phase noise? That would be the start of a scientific question.

 

fMRI is not considered definitive by most folks I know i.e. those without a vested interest in fMRI. Too easy to "dial in" results

[jabbr]

5 hours ago, PeterSt said:

 

Maybe they exist, but I have never seen a thicker (or as thick) one. 

 

All of that. But mainly air (which is the dielectric).

Does the air need to be kept hot? ... oh that's the dialectic

 

[jabbr]

13 hours ago, PeterSt said:

I will be experiencing the for me rare and quite "impossible" situation to end up in a country without internet and email. This will last for a few weeks. Never worry about not receiving any answers, as I just won't be seeing any sort of email or forum messages, etc.

 

Where could this be?, I'd like to threaten to send my 15 year old there -- do they have boarding schools?

 

?

[jabbr]

This is a good reference which discusses digital interconnect theory extensively: 

https://www.amazon.com/High-Speed-Digital-System-Design-Interconnect/dp/0471360902

[jabbr]

1 hour ago, Ralf11 said:

Has anyone done a double-blinded listening test yet?

I had two very similar glasses of Bourbon on Friday. Does that count? ... oh wait

[jabbr]

3 hours ago, mav52 said:

I own two Pass Amps,  XA30.5 and a Int-60  Neither have anything written about amp breakin in the owners manual..  Ineresting

Two different issues. 

1) temp equilibration -- this is easily measurable and the manuals recommend a warm up time ~1 hr.

2) long term "break in" not discussed in manual. NP allows that all sorts of electronics seem to settle over time and that this might be due to the listener settling as well.

[jabbr]

54 minutes ago, marce said:

Sorry but I have never seen any proof of cable burn in, it is taking things a bit far, though if proof was presented then I would change my views...

 

Firstly I am unsure about how much cable "burn in" isn't actually listener burn in but  perhaps you've seen artifacts on a scope when the leads are shaken -- vibration / motion sensitivity than can occur for a variety of reasons but is seen on scopes. 

 

Here is is one of many unequivocal demonstrations: 

 

 

So unsure what the limits of both measurable or audible settling are, but there are some physical mechanisms which might play a role in at least some situations -- connector settling is another issue.

 

I don't spend much time being concerned with cable burn in, but I do vibration isolate (ball bearings) all my equipment

[jabbr]

24 minutes ago, pkane2001 said:

 

I like it! So, to get your system settled in after changing anything (including a cable), you need to play music or white noise at maximum volume for a period of a few days?

 

 

I don't know of a physical mechanism that requires music to be played to effect physical relaxations, perhaps voltages might effect connectors -- could try the trick of hooking up headphones to amp output (beware -- don't turn music on!!!) and see if hum decreases -- I honestly don't know what the limits of audibility are just that there are a host of mechanisms wrt electronic settling -- probably should become inaudible after less than a few days but who knows (and I don't lose any sleep over these sorts of things)

[jabbr]

6 minutes ago, Theobetley said:

What about art versus attempts at art. Isn't that the same thing. My daughter can play piano but can't play like Lang Lang. Differences are in the launch the dynamics, the sustain and the sudden-transients. The notes are the same (bits are bits) but not the appreciation of the differences in artistry. The pianos are different (power supplies analog) so are the the piano players (signal integrity). This is much what highend components provide. My attempt at explaining what good highend components do for me.

My high strung string instrument players detect differences in string vibrations that are entirely inaudible to me -- new strings need to be broken in -- some strings never "settle down" and occasionally the instrument needs to be brought to a master craftsman (perhaps out of state) if something just isn't quite right...

[jabbr]

1 hour ago, barrows said:

I am no physicist, but it does seem to me that bias across dielectrics (cables, wires, and PCBs) could cause them to take a charge which might then effect the signals moving through them, considering that a cable, or a PCB, has capacitances.

 

There are numerous physical effects which have been studied in great detail. The capacitances are all measurable (and real) -- but the charge times are hard to say days! Also the capacitances are measurable, oscillator frequencies etc... when we get a really good crystal oscillator, they don't insist on several weeks "burn in" before the phase error settles down (temperature can be stabilized with an oven). There are also non-linear "memory" effects which have been studied particularly at transistor junctions e.g.  http://www.eng.auburn.edu/~wilambm/pap/2011/K10147_C011.pdf ... just a quick example there are many many more, in specific look at 1/f -> 1/f^2 noise as a function of electronic position/clustering at the silicone lattice/junction.

 

1 hour ago, barrows said:

An, perhaps interesting, aside; at audio shows, such as RMAF, one can see that many exhibitors leave systems playing, music, or "break in noise" when they shut down.

Oh yeah, the Lush USB cable is "breaking in" here...

 

The most obvious reason would be thermal stability -- there are circuits which have temperature compensation for bias voltages (particularly to prevent thermal runaway) but in general easiest to leave the electronics at equilibrated temperature. 

 

@PeterSt does a great job.

[jabbr]

22 hours ago, Ralf11 said:

I'll just keep pointing out delusions

 

Careful. There is scant scientific evidence about most of consumer audio. Certainly not enough to be certain about many things. Lack of scientific proof does not equate to disproof. Measurements that fail to show something are hardly proof that they don't exist (i.e. There might be a better measurement) An individual's personal listening impression is not a delusion. 

 

Not to to say that there isn't loads of snake oil -- just not necessarily  in this thread -- @PeterSt has alluded that he has used a specific electronic characteristic to design the Lush and I believe him -- he wants to keep it proprietary (I don't know the details myself)

[jabbr]

16 minutes ago, Ralf11 said:

Understood, but that isn't what I meant.

 

I am talking mostly about the avoidance of real comparisons (say 15 or 20 of them, at least single blinded).

 

I don't see any real science even among those doing measurements. How successful would someone be as a scientist if all they ever tried to do was disprove everything? Heh. Ever been on a standards committee? You don't get asked back ?

 

In any case most people here aren't scientists nor have access to real measurement equipment nor want to spend their free time organizing real double blind studies. It's really easy to mess those up if not done correctly -- think how much trouble the FDA has at either not approving drugs that are dangerous or only approving drugs that actually work. Who would pay the $$ needed to do a real study?

[jabbr]

4 minutes ago, Ralf11 said:

 all scientists do is try to disprove things

 

(or formulate things that can be disproven)

 

 

Well disproving the null hyp is one thing but it's the second part that is key: the formulating part. BTW who was that guy who confirmed Einstein's GTOR? Google must know... but Einstein kept all the credit...

 

In any case care to formulate a study protocol for folks at home to DBT two USB cables? One that passes muster... Post for us... let's say the Lush and another cable to keep this going n topic...

[jabbr]

4 minutes ago, pkane2001 said:

That's why it's important that others, who have access to such equipment, publish their measurements and results for others to view. Most appropriately, it should be the job of the designer/manufacturer to provide measurements to back up their claims, especially when these seem to go against known science and require a real stretch of imagination to explain.

 

Firstly I have better things to do with my time than provide measurements on other people's products -- I don't even do detailed measurements on all the products I own -- and particularly if I developing something I avoid looking at their IP. 

 

Secondly even when I show you guys a video demonstrating the electrical effect of cable shake -- you (collectively) just make jokes -- it seems that there is a certain segment whose interest is in explaining why not rather than why --I'm saying there should be a balance .

 

Now if a claim goes against known science then I entirely agree with you but the claims being discussed here just flatly dont go against accepted science:

1) There's no accepted science that has determined that cables can't have a sound

2) same goes for other devices such as USB SI eg Uptone Regen, Intona etc (I'll lump peter's Phisolator in with his DAC where it belongs)

[jabbr]

2 minutes ago, esldude said:

Where did you show video demonstrating electrical effect of cable shake?  I missed it.  Saw the one about multi-layer caps in test gear.  Saw the part about probes being rapped on the table.  I have posted the sound of tribo-electric effect.  Missed the video demo about cable shake. 

 

You're correct that was a test probe ... of course the MLCs are in the end connected to the scope... tribo-electric effect with cables ... where did you post this (refresh our memory)? In any case not all cables exhibit this and I was just saying that so called "burn in" may actually be the tribo-electric effects, etc, "settling down"... and I would expect this to occur fairly quickly.

[jabbr]

17 minutes ago, Ralf11 said:

 

I am going to guess that you are not a scientist.  Anybody (ok, any young grad. student) can do the formulating, and some journals in my own field are rife with speculation (fancied up with math).  It is the test that is key.

 

Got 2 people?  I have better things to do with my time too....  but it's easy to have one person swap cables while the other is at work or wherever.

 

C'mon. DBT with one subject (the other human is doing the swapping)? Really?

 

It could easily cost $10K to do a real study ... ever done a budget for a funded human study? Maybe this could be crowdfunded to be objective. You are a scientist, so let's see a protocol? Or... yeah too much work, that's my point

[jabbr]

27 minutes ago, The Computer Audiophile said:

Things that go into the body may need a different standard. That said, nobody has to buy products that don't pass that standard. Such as, "not tested or approved by the FDA." 

 

 

 I think that, unfortunately, the severe expense of FDA testing greatly increases the cost of new drugs such that they are unaffordable. The placebo effect will often get you about 30% of the effect, and if its affordable then that is better than nothing if real drugs can't be purchased.

 

That's why I often give a pass on a product that isn't that expensive, because short of patent protection, the manuf may not want to spill be beans, or the required testing might require a higher price...

[jabbr]

12 minutes ago, Ralf11 said:

I've posted a methodology twice now...  certainly 1 person - the buyer;  I check what I buy during the return period

 

but it would be easy to include 20 people if they want to do a test and pool the results

Can you quote where you posted?

 

but point being that you weren't trying to be scientifically rigorous? Would you be willing to write up results and submit to a journal? that would be science as opposed to the casual testing we do almost daily... y'know the kind with statistics, measures of validity, pre-test probabilities post test etc, discussion of stats methods e.g. ? Bayesian etc... that's all a bit of work ... and then your friendly Editors would think you've been hitting the sauce too hard ...

[jabbr]

4 minutes ago, Ralf11 said:

yes, but not hard to do a simple t-test

 

Are you a scientist, jabbr?  scientifically rigorous is a fairly wide landscape...

Researchgate tells me my papers have 3000 citations -- 15 publications -- I've done science. Last year I helped a few members of our Symphony get funding to do a randomized controlled study that involved music perception and human effects. My research assistant wrote the protocol under my direction. I volunteered my time. I really don't like to argue by authority (and no need to get into more details) but I've got a great idea of how much work it really is, and prefer to drink bourbon on off hours.

[jabbr]

39 minutes ago, Ralf11 said:

ok, flipping 2 things from Popperisitic science and some other things had me wondering

 

for the record, I'm a biologist  (and not  chemist in drag type either)

 

What you might consider "flipped" may have been taken out of context -- and also understand that I'm pretty much always doing something else, and so make errors ... but let me explain:

 

There seems to be this idea that if someone does a single measurement (or set) using a single technique, and fails to show a difference, let's use ASR with the Regen as an example, that this is scientific evidence that the DUT isn't doing anything. Now in this case the so-called investigators seems to have a bias against finding a difference. The manuf has shown measurements (eye pattern) showing an effect. Now ASR has published another set of measurements that demonstrate exactly what I described months and months ago, that the DUT should reduce the linewidth (or tighten up the peak) but this isn't really discussed. I would just say that there isn't real science being done.

 

Certainly I agree that the manuf ought provide data to support claims, and if it were my product, I would insist on such data. I wouldn't be able to produce a product at such affordable prices -- I guess that's why I do consulting  ... but the point is that folks should realize that most manuf are working really hard and probably not earning $1000/hr ... so y'now if we are going to ask for more, expect higher prices...

 

So yeah, Popper, that's old hat and sort of debunked (but you know at a crazy level that is perhaps not even understandable), but not really relevent here (unless we are doing real science). I think Kuhn might be more apropos, and personally, I always look at who are among the first to either create an idea or popularize an idea -- @PeterSt and @Miska, very different but both with the computer based upsampling idea and both with fantastic actual implementations ... and this idea is so so important, critical, essential, for computer audio, that I have and enjoy a Phasure NOS1a, and have a Lush and am upgrading to the G3 -- it all sounds great and I won't bore everyone with my own subjective impressions, but also realize that I've built @Miska's DSC1 DAC (which is terrific) but have also designed a balanced discrete jFet version so if anything I should be personally biased against PCM and the NOS1a -- it does irritate me endlessly that there should be a dependence on OS optimization for sound -- I consider that a bug, not a feature -- and hope to eliminate this dependence.

 

In any case there is really alot to be learned here and just hope that folks can keep an open mind -- if you don't have an open mind, then you aren't a good scientist (whatever Popper may have written about that, if at all, but regardless)

 

Edit: uh oh ... that last part sounds like I'm channeling Peter ... its the bourbon talking  

 

[jabbr]

2 minutes ago, Ralf11 said:

 

don't take it the wrong way or applying to you, but today a co. - let's say a co. making cables filled with a toxic liquid metal - could get lots of sales from users posting on the internet (either by conf. bias or by shilling) without ever taking out a single advertisement

 

Yeah so I draw a very very  sharp distinction on that one where the technical claims are entirely pseudoscientific and unsupported by any theory let alone measurement. Like I don't even care what the results of a DBT would be...

[jabbr]

56 minutes ago, Albrecht said:

It is a gross mis-representation to call SuperDad's post advertizing. I call BS on that. Essentially you are saying that because Superdad responded to a post with the moniker "SuperDad" that that is an advertisement; and therefore ridiculous.

It's marketing and he's damn good at it. If we were all communists here we would have shortages of plain black USB cables ...

[jabbr]

3 hours ago, esldude said:

Well confirmation bias seems to have extended into the upper reaches of the violinist ranks.  The renowned and revered Strads in a blind test weren't so superlative all of a sudden. 

 

One report/study does not make a fact. Why are you so willing to conclude that the upper reaches of violinists don't know their instruments?