USB audio cracked... finally!

[barrows]

17 hours ago, fas42 said:

 

If a vocalist is singing directly at you, live, no PA - and then slowly turns around as she keeps singing, the same song - the frequency spectrum of her voice will change dramatically, as received by your ears, as she does this. Does the "timbre" of her voice change such that her voice now sounds artificial, or wrong?

 

Good replay will generate the same subjective experience.

Hmmm.  this is a good example of often why a replay system with no visual component may not "get it right" in a sense, when it really is getting it right.

When the example above occurs in a live situation, the visual information, of seeing the singer point in a different direction, alerts the ear/brain mechanism to adjust the perception of timbre, making it seem consistent.  In an audio playback system, with no visual cues to go on, the ear/brain will get confused a bit by the same auditory information, and this may lead to listening difficulty and a loss of timbral accuracy (or perceived timbral accuracy).

Sorry, did not mean to interject really, but it is an interesting example...

[barrows]

I ordered a LUSH, as i have been looking for more USB cables of interest to try.  I know Peter has no love for blurry sound, so I am hopeful for good performance with this cable.

[barrows]

6 hours ago, Speed Racer said:

 

It can. But in this case, it's a cable made out of spec to cause certain things to happen in the transceiver of the source computer. At least that is what I think Peter said the last time he said something concrete about where the "magic" happens. Since there are a bunch of different USB transceiver chip sets and variations within implementations using each chip set, I find it hard to believe that a single "Lush" cable spec can affect all the transceivers in exactly the same way so the "magic" happens.

 

I don't doubt that Peter can build a cable that will do something cool with a particular implementation....I just don't think it can happen in all implementations without tuning. It doesn't make engineering sense. Hell, I remember working on Ethernet cards at Apple and changing anything in the design (trace paths, resistor manufacturers, chip manufacturers, etc.) could wreak havoc on signal behavior. Marginal, but in spec cables, could cause failures as well. We once had to make a card an inch longer to make room for different layout so we could get marginal cables to work consitently.

 

And I am supposed to believe that one specially massaged USB cable will affect all USB implementations exactly the same way? No sir..... 

On the one hand, I agree with you: the transmitter-cable-receiver is a system, and changing any one element will effect the outcome of the transmission properties, so how could a single cable be idealized for every possible USB implementation?  But, it appears that Peter's design approach is different enough from most USB cables, that it will be (nearly) universally "good" sounding, or at least "different" sounding.

It is also true that different USB receiver circuits in DACs are not usually all that different, generally there is one of two parts, an XMOS or Atmel processor with an onboard receiver.  What does Phase use in their DAC?  If it is Amanero based then it is the Atmel.  Then, maybe the interfaces are different in the trace length and routing to the chip, and the use, or not, of damping resistors on the trace to the chip.

[barrows]

OK guys, I am unsubscribing.  I see nothing of substance here about the LUSH cable anymore, just the usual bickering.  When I get a LUSH cable here I may choose to re-engage here with some listening impressions.

I accept already that different USB cables give rise to different sonics, although in my current system the implementation of the USB transmitter and receiver sections (and isolation of USB on the DAC side) is good enough to reduce those difference to very small levels.  I am looking forward to trying the LUSH and seeing what it brings, I am also interested in hearing of others' listening experiences, vs. other USB cables...

[barrows]

1 hour ago, jabbr said:

 

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)

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.

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...

[barrows]

54 minutes ago, pkane2001 said:

 

Since Lush is said not to conform to USB 2.0 spec, it would be interesting to see what exactly it does. RLC, bandwidth, FR, response to square wave at 480MHz, the eye diagram.

 

Ultimately, what's most important is the signal at the output of a decent DAC. For this, a comparison between a good $10 USB cord and Lush in an un-averaged FFT plot would be good. Driven with a J-test signal for jitter analysis, zoomed out, and zoomed in areas around the primary test tone (per jabbr). Possibly a multi-tone FFT, if that helps bring out a difference.  Anything else Peter might think of to demonstrate that Lush actually alters the analog audio output compared to a basic cable. 

 

While I as well would like to see such measurements, what if these measurements show no/little difference?  There are valid reasons why some audiophiles do not put much value in measurements, and no, it is not because (they/we) are living in some kind of world of make believe, it is because most experienced audiophiles have heard a component which measures perfectly well (via the standard set of audio measurements) and sounds absolutely horrific.  Take a look at John Atkinson's (stereophile.com) DAC measurements: there are many where distortion products are well below the threshold of perception (at least commonly understood thresholds), but audiophiles clearly hear (and yes, even "blinded") distinct differences.  

My belief is that indeed, measurements are for sure necessary, at the very least to make sure that something is not terribly wrong with a product, but that the current set of measurements is not enough to adequately equate to audio sound quality, especially at the high end, where minute details at impossibly low signal levels, are what it is about.  As it appears the human ear/brain mechanism is more sensitive to time based distortions, perhaps we should concentrate on measurements in the time domain rather than the mostly static commonly accepted measurements.  Nordost tried to do this  a few years back, but when their protocol (using actual music signals) showed measured differences between cables, the deniers pushed back so hard (saying the measurements are invalid BECAUSE they show a difference in performance of cables, how ironic is that!) that the program was abandoned (at least in public).

Peter, I know that you have stated the LUSH is not built to USB spec, but this could mean many things: clearly at the very least it is not built to the design spec, which requires twisted pairs and shielding, but that does not mean the LUSH does not meet the USB performance spec (for impedance and bandwidth).  Many current high end USB cables are not built to the USB design spec, but many of those same cables do meet the performance spec (and sometimes greatly exceed the spec for bandwidth, looking for improved performance, like Lightspeed)-it appears from his cagey responses that Peter's LUSH may not meet the performance standards for bandwidth/impedance?

My curiosity is about the design of the cable itself, I know Peter appears not willing to mention this preferring to, keep it proprietary, which of course is his right for his intellectual property, but I am curious.  So far the little actual technical build info I have gleaned form his comments is that the dialectic is mostly air (not unusual for a high end USB cable).  Having the LUSH here, the one observation I can make is that the (rather burly feeling) outer jacket seems to be woven cotton... which is fairly unusual for a USB cable.

 

On the listening side, I am wondering what people are thinking: is the LUSH actually performing better than other cables, that is producing all the musical details while eliminating some kind of artifacts (distortions), OR is the LUSH producing a euphonic kind of sound by somewhat softening the focus, perhaps softening the edges and softening the artifacts/distortions at the same time?

[barrows]

Paul: of course, agreed!  As I mentioned, I would love to see some measurements as well, eye pattern, etc...

[barrows]

12 minutes ago, Johnseye said:

Unless we know the engineer's true intention, backed up with measurements and statistics, then we won't know if their product is designed to present the cleanest, true reproduction of the recording or they are shaping the sound to be warm and more analog like or some other signature type.  In the end it doesn't matter if you like how it sounds unless the engineer is misleading you to believe they are cleaning noise and you're getting the true reproduction, when in fact you're getting a DSP modified sound.

I think the above remark represents a "slippery slope" so to speak.  The idea that, "... it doesn't matter" is not actually true in my mind, as here is the problem with this approach:  If one is trying to build a system that is true to the music (recording), then that system should be as neutral (in tonality terms) and as resolving (in terms of musical details and without any artifacts) as possible.  If components are being developed to "shape" sound we are away from being true to the recording, we are actually building a system which leans towards making all recordings sound a certain way.

The problem becomes more apparent when mixing and matching components:  the audiophile ends up in an endless balancing act, mixing "warm/soft" interconnects with cold/hard amplification, and on...  this results in problems every time a new component is brought into the system for evaluation, as the component in question may be neutral and resolving and "perfect" in every way, but in the flawed (shaped) system, this component may seem to not be so perfect...

While we hear much discussion about "musical" systems, and often how systems can be made "musical" by actually softening the sound (SET amplification, etc), in my experience, the very best and most musical systems which I have heard are the ones with the ultimate resolution and the least artifacts, and are not "softened" or "smoothed" to make them more musical.  Ultimately, the system with the most resolution, and the least artifacts, sounds the best, as it relaxes the demands made on the ear brain system and allows for deeper listening (connection with the music).

 

In regards to LUSH, my question would be: is LUSH accurate, and as such is it increasing resolution and reducing artifacts, such that the listener hears more music and makes a greater connection, or, is it reducing resolution in some sense, performing a "cover up" of musical artifacts?

[barrows]

Alex, glad you brought the conversation back to LUSH.  I may post some listening observations soon, as I have close to week of play on the cable now and it is time for a bit of A/B comparisons, but unlike many here, I am not so sure about this cable with my system...

[barrows]

Lush, some listening impressions.  First the DAC is what I call  El Diablo, it uses the (no longer available) Sonore USB interface which is galvanically isolated and has a built in oversampling for PCM to 352.8/384 with a partial minimum phase filter.  Then there is a Twisted Pear B-IIIse, running in sync mode from the Crystek 957 clocks on the Sonore interface.  An NTD-1 does I/V conversion with Clarity CMR output capacitors.  No USB gadgets are used (or needed) and the source is a Sonore Signature Rendu SE.  To any ESS haters out there, this DAC does not sound like your typical ESS implementation, the ESS chips are capable of incredible sound, but you have to use them right, sync clocking, turning off the first oversampling filter, and really good power supplies makes a huge difference here.

At first listen I found the Lush to be a bit dark, and a bit soft around the edges, and perhaps missing out on some details.  So before any further close listening I let it run in for a week.

The Lush still a little on the dark side, but much less than initially, it appears that transient edges are still a little blurred, but this is just a little.  There is definitely more warmth and body in the sound, and this is quite welcome, and gives the music a "larger" quality; no the soundstage is not actually any wider or bigger, but it is set back just a bit more.  So I did some very intensive A/Bing with Diana Krall's "Temptation" from the 24/96 Girl in the Other Room album (I know, I know but it is good test material). Here I could clearly hear a little more ambience/bloom around certain sounds: the "halos" of ambience around individual sounds in the soundstage (easily heard with percussion, but some piano notes as well) was more evident and better expressed with the Lush.  This was interesting, as this kind of performance signals more resolution to me, despite the still present, slight transient blur, a bit curious perhaps.  As of now I am liking the Lush for my system, but more testing is in order.

I would say the Lush is going to be a good choice for anyone looking for a bit more body/fullness in their system, and if it really is letting me hear more low level ambience details, then it might be a good choice for a lot of people.  The price is very reasonable as well, and despite running a very small business, Peter got mine to me very quickly from the EU.

I will be testing at least a couple more USB cables before Rocky Mountain Audio Fest, as I am attempting to find a couple of really good options for Sonora's Demo system there, anyone attending, please remember to come by and say hello, and perhaps listen a bit.

[barrows]

1 hour ago, Speed Racer said:

You might as well put every DAC on your list. EVERYONE says this cable is the second coming. If Alex's ISO Regen did all he claimed there is no way this cable could make the difference he hears.

Please stop posting uniformed opinions with no basis in reality here.  When you actually have some experience with the components under discussion, I would welcome your input, but at this time all you are doing is wasting everyone's time. 

[barrows]

1 hour ago, Jud said:

 

Wish I could be there, but will be moving to a new home most likely in November, so time and budget will have to be concentrated there.  Next year....

Cool, BTW Sonore owes you a dinner, perhaps at RMAF 2018.

[barrows]

I would suggest that good music is necessary, and good recordings are rare, and appreciated when they happen to be of good music.

[barrows]

3 hours ago, Ciukas said:

I thought there could be some sort of confirmation bias

Definitely no confirmation bias here...  I have trained myself well to avoid these problems (and I have been paid to listen test components as part of my job in the past), and, I confirmed my view of the Lush being a bit warmer/fuller than the comparison cable (not stock) via a friend who I did the comparison for (without giving him any prior idea of what I thought about the differences, his initial statement after one A/B was that the Lush sounded "fuller").

Good for you to use electronic music for testing as well, contrary to some audiophile beliefs, I find electronic music (in my case often NIN) can be quite telling on some aspects of sound quality; there is a lot of interesting textural details in electronic music which requires a good system to express fully.

[barrows]

1 hour ago, mansr said:

ROFL

Comments like these speak more about the commentator than they do about me.  You have no idea of my level of training and experience in this area-like most things, listen testing is a learned skill, and learning how to avoid biases of any kind are a big part of that skill set.

[barrows]

That could be interpreted as evidence that blind testing is flawed.

[barrows]

4 hours ago, Summit said:

What exactly make blind testing flawed? Blind test are not the best method for evaluate all SQ aspects, but sure is for checking biased based on reputation, manufacturer and price.

 

All people are biased somehow, some more than others. To understand that we are all living in a context and are all biased because of previous experience and expectations is critical to minimize the influence of bias (this swings both ways and also includes people that are biased against any SQ difference in cables before even testing them).  

 

Bias = Expectations

I am not going to engage in any kind of long discussion about this here, as it has been hashed over elsewhere. Succinctly; blind testing creates stressors which make critical listening difficult, the tendency is that one is testing the subject rather than the DUT when blind testing is done.  If you are interested in more information about this I suggest a Google search.

 

Listen testing is a difficult, learned skill, one of the things one learns is how to avoid expectation bias.

 

@elsdude: Yes, the first step of any listen testing is assuring levels are precisely matched, without this control one is off in the weeds.  this is easy to achieve for USB cable testing, as different cables will not/cannot alter playback level.

 

Can we get back to the LUSH?  Anyone else with some listening observations? 

[barrows]

19 minutes ago, rickca said:

Like a nothingburger? 

http://www.marketwatch.com/story/you-can-get-a-nothingburger-at-mcdonalds-but-it-takes-some-work-2017-08-23

While that is true, a jitter free signal entering the DAC (chip I think you mean) does not guarantee "perfect" analog sound at the output.  There are many ways that noise from the digital input side of a DAC (component) can couple with the analog signal, and cause audible distortions.

Many of the efforts of good high end manufacturers are focused on reducing noise on the input side, and on reducing radiated RF (internally) which can cause analog distortions.  Are these often low level effects, yes of course, that is exactly what high end sound is about.

Additionally, there is little guarantee that there will be a low (it will never be jitter free, no such thing) level jitter signal at the DAC, and that these same noise sources can and will effect masterclock performance.

Optimizing a DAC for all these problems is not easy.  We must remember that a DAC is a mixed signal system, Digital and Analog, and clock signals themselves are analog signals.

[barrows]

6 minutes ago, CuteStudio said:

 

As you have noticed there is some speculation about the possible mechanisms by which a USB cable could achieve these interesting results, I have realised - and feel a bit sheepish about missing - that we are all failing to do the obvious: Ask the inventor how it works!

 

Please then can you explain how the cable manipulates the serial data stream or the timing - or some other factor to achieve the sonic improvements? I.e. what is the electrical and/or physical mechanism? Is there an internal processor in there somewhere?

 

Apologies for not just asking the obvious earlier, many thanks in advance!!

 

I do not understand why you continue to post on this thread.  It appears that you have no personal experience which is relevant here, and that all you are sharing is your belief system.  I have no interest in personal beliefs which have no basis in reality.  Your comments are worthless to me and just add noise.

There is plenty of discussion elsewhere on these forums about the how and why USB cables sound different.  No one, ever, suggests that the cable is changing the data (for example).  If you can make relevant commentary about the LUSH, I welcome that, otherwise, it seems like you just like to hear yourself talk.

[barrows]

3 minutes ago, CuteStudio said:

Is there another mechanism that can explain any of the claims for the cable within the known laws of physics?

You are completely uniformed.  Additionally, it is not the responsibility of anyone here to educate you.  If I suspected that you even cared to learn, I might try, but it is clear you are only here to listen to your own thoughts, and bask in your superior nature.  Please stop wasting our time and adding noise.

[barrows]

19 minutes ago, GeneZ said:

Could it be?   That with the modem disconnected that Windows shuts down certain background activities that creeps into effecting the sound when the audio source itself is found on the hard drive?

Easily could be...  Reasons like this are part of why many choose to use audiophile sources, in my case, an Ethernet connected Audio renderer, getting all commercial computer gear as far (physically and electrically) from the audio system as is reasonable.  Just an ultra low noise Renderer, made for audio, in the playback system.

[barrows]

3 hours ago, Bystander said:

You can't will yourself out of being susceptible to placebos.

Right!

 

But the point which many here seem to be missing is that you can train yourself to avoid the pitfalls of placebo, and expectation bias.  When one has gone to great lengths to train themselves to be a critical listener, to be able to hear small differences with a great degree of accuracy, it is maddeningly insulting when someone else, with quite different experiences, comes along and suggests that they are imaging things.

 

I am a true believer in science, to the extent that it is a developing field of study, where new discoveries are being made every day.  But engineering is different from science, most engineers are taught hard and fast rules, and most engineers do not think beyond those rules.  This is fine for their profession in most cases, as the form of their rules is entirely adequate for most electrical design tasks which they take on, in audio, those rules are not always wide ranging enough to design the absolute best sounding products.  Sometimes it is necessary to go beyond the constraints of ordinary engineering to make advances in performance.

 

Right now it is clear that the standard set of audio measurements is not adequate to describe all aspects of sound quality of an audio systems; I am not suggesting that all aspects of audio performance are not measurable, all I am saying is that the standard measurements commonly used are not enough to do so.  Perhaps some day we will have more/deeper measurements that can do this (and I hope so).  One audio engineer I respect very highly is Bruno Putzeys (Hypex, Grimm Audio, Mola Mola, etc).  He is very much an engineer first, and designs by simulation, maths, and measurements first.  But he has stated publicly that he also listens, and when he listens, if he finds something that does not sound quite right (according to his sims and measurements) he then works backwards from his listening test to try and find the measurable parameter which could describe the sonic problem he heard.  He does not convince himself that his listening test is invalid, because the measurements he initially did do not show a problem.

The lack of respect given to those who listen here, by those who have not even tried the product at hand (Lush cable) is not acceptable in my book.  Listen to the cable for yourself (and measure it as well if you can, I would love to see at the very least an eye pattern comparison between the Lush and some other cables) before insulting those who have.  

[barrows]

40 minutes ago, Jud said:

 

Would love to see a thread about opamps...

Jud, go to DIYaudio.com and you can read about opamp comparisons for DAYS!  I prefer no IC opamps in my system (but my amp does have a discrete opamp implementation in for the input stage).  Sorry for the OT guys, back to LUSH...

[barrows]

6 hours ago, Audiophile Neuroscience said:

I'm not so sure ASA is the mechanism used to distinguish realism in audio playback. 'good hifi sound' , now a pejorative term, is pretty good at presenting multiple discrete sounds spread in the sound stage. Perhaps it is *too* easy for us to de-scramble making it unrealistic and a cue that it is not real.

 

Intuitively I think it has more to do with more fundamental qualities of the sound like tone, dynamics and scale.

 

All I know is when the sound quality is what I consider good my dog and cat also pay attention to the sound. They know when its fake like on the TV and ignore it. Perhaps it just means I have the same taste in music as my dog :-(

I can see it now, dogs at RMAF!  Dogs in every listening test at dealers across the country!  Hahaha.  Classic post.

[barrows]

32 minutes ago, Speed Racer said:

Feed the USB PHY the perfect USB signal. What the "Lush" cable does is add something to the workload of the USB PHY by being out of spec thus changing the sonic signature down the line.

Speed, please provide your evidence for this remark.  Apparently you "know" what the Lush cable does, you must have at least listened to it, or perhaps measured to "know" this, so please share your information.