Thread Rating:
  • 1 Vote(s) - 4 Average
  • 1
  • 2
  • 3
  • 4
  • 5
I'm coming 'off the fence'
#1
Hi guys,

For years now, I've been sitting on the fence when it comes to the cable debate. Reason being; I always had more important things to spend my money on. Yesterday though, I ordered myself a 'high end' USB cable (Straight into the deep end on the controversial scale  Tongue ) and in light of the current argy bargy on Stephen_V's thread I thought I would start a new one here so you guys can abuse me and leave his thread for what he intended it to be.

What I'm going to try to do here is write up my findings in an objective way. Why do I think I can be objective? Well, a few things.

For one, I've listened to a lot of reasonably priced and very expensive systems in the past and have found that price had no influence on whether or not I was impressed with the sound. Proof to me I'm not easily biased. And the cost of my new cable really isn't going to make a dent in my financial situation so there will be absolutely no need to justify my purchase. I'm buying this cable for the purpose of learning and gaining an opinion only.

Two, I don't have any audiophile friends to impress with my 'shiny new cable'. If anything I would probably be laughed at by most if I told them what I 'wasted' my money on.

Third, I don't believe the placebo effect will have an influence on me in this case for one simple reason; I want to hear a clear difference in the sound. If I'm going to have to guess then there is no difference worth mentioning as far as I'm concerned.



A bit of back ground information on me and why I'm open to the subject. After all, ones and zeros etc...

I work on an oil and gas rig as a Measuring Whilst Drilling hand. We use down hole tools that use digital communication to get the message across. 
This is how it works; Drilling gets done with a drill bit on the end of drill pipe. In the bit are a few jets and water gets pumped through the pipe and jets, circulated to surface on the outside of the pipe and with it will come the cuttings so the well stays clean and we don't get stuck. 
In the drill pipe sits our tool taking all kinds of readings. The tool has a little poppet that can partially restrict the flow of the water for a short period. The short restriction will created a small increase in pressure and presto; there's a 1. On surface there's a pressure transducer that 'sees' the increase in pressure and converts that to a change in Voltage; A series of pressure changes at the right time create 'digital' words, with the use of some smart algorithms.

It's quite a simple idea but it gets a bit more complicated than that; There's interference! Whoohoohoooo   Huh

1. The pumps use pistons and valves creating a pressure change of themselves; creating harmonics. With that will come first, second, third etc. harmonics that could all be seen as pulses and therefore influence the proper reading of the original words.

2. The cables are long and run along other cables. This creates magnetic interference, which in turn creates electrical interference, thus creating another obstacle for a clear 'reading' of the words. 


Because the pulse frequency is very low (4-8Hz) I can actually see the pulses on our computer screen. Yes, I can see the shape of the pulses created by the system as a whole. And I can tell you that, in some cases, it doesn't look pretty. And we can freeze the pulses as well and count them. This way we can find out if some words are actually made up of the ones and zeros they are meant to be made up of.

And in some cases, there is a shift in the pulses; Jitter! Huh

Still, we can take surveys that make directional drilling more accurate then the best sharp shooters on the planet; We hit 4 inch targets much further away than the best sharp shooter can hit a person. Pretty cool stuff.

So it's clear that the interference is there but it's also clear that the interference can be dealt with adequately. In this case that is done with very sophisticated filtering and algorithms. (a 1 billion US$ / year budget can buy you some fancy gear)



So how does all this equate to USB or any other digital cable? Well, I've seen proof that digital signal isn't digital at all; It's an analog signal that varies in Voltage (common knowledge), quality (some pulses look really wonky) and timing. It's the timing that I would like to delve in a bit deeper.

In MWD there are rules; The word are part of a sentence in a certain order and of a certain length. So the surface system knows what's meant to be coming and has a pretty good idea of what's rubbish and what's not. If a pulse has shifted a bit, it can be 'moved' so the word makes sense. If the signal is complete rubbish, the system will let us know and will wait for the sync word, sent at the start of every sentence, and disregard the previous sentence.

In digital music transfer, I have no knowledge of such a system. As far as I know, one end spits it our and the other end catches it. There are too many variables in music for the receiving end to know what's rubbish of not. (again, as far as I know)

What I have absolutely no idea of though is how robust digital music transfer is compared to what I use at work. What I do know is that the higher you go in frequency, the harder it is for a signal to stay intact...

...Which makes a reasonable case for 'not all ones and zeros are the same'. Hence, I'm open to the subject.



So, in the coming weeks / months I will write my finding and at some stage I will jump off the fence on one of the sides. Interesting times ahead!
                                                    Lifetime Roon, Mac mini, int. SSD, ext. HDD, tv as monitor, key board and track pad on bean bag as remote,Devialet 200, Od'A #097, Blue jeans speaker cable,                                     
                                                                                                                                                                            Dynaudio C1 MkII.
                                                                                                                                                                              Jim Smith's GBS.
                                                                                                                                                                        Northern NSW Australia.
Reply
#2
Woo-hoo. Getting interesting. Eagerly awaits your findings.
Before: Le200, KEF LS50, AQ Type4, NUC 5i5RYH/8GB/128GB M.2SSD, Roon, Win8.1/AIR2.1.3/RoonBridge, MM/AIR3/RoonBridge, QNAP TS-212P 5TB NAS, AQ NRG-X3

Now: KEF LS50W, NUC5i5RYH/8GB/128GB M.2SSD, Roon, QNAP TS-212P 5TB NAS,iFi iSilencer3.0+DC iPurifier+iPurifier2, Sonos ZP80+SPDIF iPurifier


Location: Cyberjaya, Malaysia
Reply
#3
Nice post Pim.  Always interesting to hear about what other people do for a living (especially if it's engineering -- I am biased!).

Picking up on one point you mentioned: error detection and correction.

My understanding of USB (having written host and peripheral drivers for USB audio among other things) is that each USB packet carries a CRC which enables the receiver to determine whether or not the packet has been received correctly (all bits as transmitted).  In other words, errors can definitely be detected.  However, as far as I know the USB audio standards do not define a mechanism for error correction (other standards, for example mass storage, do define error correction methods).  Putting this into real-life terms, if a DAC receives a bad packet, it know about it but is then on its own in deciding what to do -- maybe output silence for the short period corresponding to the incorrect packet, or maybe do something more sophisticated.

Also, when I hear statements about signal rise time affecting jitter, I am sceptical in the extreme.  Maybe that was/is/can be true with something like S/PDIF, but there are so many more hardware and software layers involved in USB that by the time the payload reaches something that actually knows what to do with it (i.e. play it as audio) the original clock timing is long gone.  (I am not talking about clock drift, which is a significant concern, but short-term jitter e.g. between samples.)  All USB controllers that I've worked with transfer data from the wire into memory in chunks of words, maybe whole packets at a time.  Any relative timing between arrival of individual bits from the wire is surely lost during that process.

The same thing is broadly true of wired Ethernet, except that some of the higher-level protocols (such as TCP) do include mechanisms for error correction.

It should be no surprise that both these protocols have (at least) error detection capabilities, because otherwise they would be unreliable for general use.  We don't agonise over whether "bits are bits" when we connect a disk drive to a computer or use file transfer protocol to send a file across the Internet, precisely because the relevant protocols are engineered to detect errors and correct them, one way or another.

Having said that, I can see mechanisms by which a USB- or Ethernet-connected audio system could be affected by subtle behaviour depending on what cable is used.  The most obvious one (to me at least) being injection of noise via the power and ground and/or due to switching currents.  These are analogue effects, of course.  And I don't see how one could generalise about whether or not they are audible as I'd expect the effects to depend substantially on the design of both "ends" (e.g. computer and DAC, streamer and Devialet, ...) as well as the connecting cable.

I'll be really interested to follow your experiment.  One other thing, though: I wouldn't under-estimate the power of expectation bias, even if you try to take it into account Smile
Roon (Mac Mini), Wilson Benesch Full Circle, Expert 1000 Pro CI, Kaiser Chiara
Warwickshire, UK
Reply
#4
Very interested to hear what you think. Did you specifically not name the USB cable you bought? I am keen to hear which one it is.
Devialet 200 -- Roon Nucleus-- Sonus Faber Olympica 2 -- Tellurium Q Black Speaker Cables --
Chord Qutest -- Niimbus US5 Pro Headphone amp —HifiMan HEK, Abyss 1266TC
Newcastle upon Tyne, England
Reply
#5
With the best will in the world, there’s no way of controlling for expectation bias except by blind testing. People often say things like “I’d like the cheaper cable to sound better, therefore expectation bias won’t make me prefer the more expensive one” or “I’m expecting the cables to sound the same, so I’m immune to expectation bias”. It would be nice if statements like this were true, but sadly that’s not how expectation bias works.

(In passing it’s worth noting that ‘expectation bias’ isn’t exactly the right term. What’s actually in play here is a wide range of unconscious cognitive biases, of which expectation bias is just one.)

In order to fully control for cognitive biases, you would have to be absolutely certain that you had excluded all unconscious cognitive cues that might incline you to make a false judgement. But how could you ever achieve certainty about this? You might be utterly convinced that your judgement was not affected by cognitive cues, but you cannot prove it, unless you remove the cues in the first place, by blind testing. 

You might argue (as you do in your post) that being relatively certain is good enough. Well, it isn’t. Even the smallest degree of bias, which might be totally unconscious, can be enough to skew a set of results completely.

Now consider the state of play in audio cables. We have thousands of anecdotal instances of people claiming to be able to discern differences between cables in non-blind evaluations. However, there isn’t a single halfway-reliable instance of someone being able to replicate those results in a blind test. Doesn’t that strike you as odd? And doesn’t it strike you as odd that cable manufacturers haven’t tried to use blind testing to advertise their cables? 

On the other hand, we have a wealth of evidence that shows how easily our minds are influenced by extraneous cues without our consciously knowing it. The placebo effect is the best known of these. 

It’s important to recognise that no-one is accusing anyone else of being dishonest (with the possible exception of some cable manufacturers). I’d trust you to be as honest as I’d hope I would be in the same circumstances. Nor is anyone accusing anyone else of stupidity. On the contrary, what’s at issue here is the sheer complexity and resourcefulness of the human mind, which is after all one of the most complex things in the known universe.

In these circumstances it seems to me that non-blind evaluations of hi-fi cables should be treated with amiable scepticism. Good on you for being honest about your opinions. But I hope you’ll understand why I couldn’t reasonably treat your evaluations as evidence of anything except your own honest good intentions.

Matt

Sonos Connect (W4S) > DSpeaker Antimode 2.0 > Sanders Magtech > Martin Logan Montis
Sonos Connect (W4S) > Devialet 200 > Vivid V1.5
Silver Phantoms (just the two)
London
Reply
#6
Pim, you have a higher pain tolerance than I do. Cheers! I hope you create a successful honeypot here for the cable skeptics.

Your water-jet communications system is very interesting, thanks for the explanation. Just think of the acronyms Devialet could invent for that. What a cool job you have!

For some background I worked in telecom at Bell Labs so might know a little about optical fiber, communication data transmission, etc. I did human-interface design and testing so have some professional education and experience in evaluating the sound of things (among other aspects of perception). However we're not playing with a long-haul backbone here but rather with audio cables so my technical understanding doesn't necessarily apply. I too was skeptical of an audible difference even after I understood some of the science of audio cables. In addition to having a technical bent I'm a cheap bastard so wanted a cable of sufficient opto/electical specs. to sound as good as anything else to me. I have no attraction to the aesthetics of thick cables, fancy casework, or other audiophile bling. My ideal system would be invisible and cheap. However, hearing very clear differences convinced me to test with my ears. Even when I can hear I difference I want to understand why the difference exists though.

At this point there is plenty of data about why digital cables don't just transfer bits perfectly and thus end the story. If anyone wants that information it is publicly available, or I can provide some links. You got to the core already though, Pim: we're transferring digital bits through an analog physical medium and therefore the properties of the analog world influences the transmission of digital data in audible ways. That doesn't excuse any of the snake-oil salesmanship which is rampant in the audiophile industry and especially in tweaks such as cables, but it does explain why some people hear a clear difference among digital cables.

Even hearing a clear difference is more complicated than one would assume though, given how effective placebos are due to expectation bias, psychoacoustics, etc. Even blind testing is complicated, and conducting a valid blind test isn't trivial. You can easily create a blind test that yields results which aren't repeatable and thus are invalid. There's also no absolute truth to perceptive senses. I've done some studies with color perception. I can display a perfectly calibrated color to a set of people and they will see it differently. The color is scientifically absolute but irrelevant given relative human perception of color due to eye and brain differences. We've done voice-quality transmission testing that found more correlation to the time of day when we tested than to anything else because how humans perceive sound changes during the day. We need to recognize the whole system, including the human element. Back to this hobby, our target isn't necessarily the perfect transmission of a live recording given all the manipulation of recording, mastering, etc. Our target is what sounds best to the person doing the listening. So the valid test for you is what helps you determine what you like to hear. I would be interested in hearing how people like to test what sounds best to them. Since few of us have multi-million-dollar test lab budgets (and even those aren't necessarily useful) we need to figure out how to find the pieces of equipment that sound good to us among the chaff and snake oil.

Since we can't try everything it is useful to hear what others have heard (and thus we talk about equipment including cables). I use the experience of others to decide what I want to try myself, and then find it useful when dealers allow me to take home some options to hear in my room. Unfortunately I don't have easy access to many options and can't afford most of them so have to make decisions with limited experience. I find both quick A/B testing and longer listening useful evaluations. I can concentrate on the analytic differences and am also learning to concentrate on the emotional impacts as well. I use friends to help me avoid some of my biases, either by changing the equipment while I listen or through their perception. If I'm the only one who can hear a clear difference I would certainly question myself. I find it fun to have friends who are musicians play their music on my system, and I like to hear my favorite recordings in different systems/rooms. I also recognize the wisdom of finding something emotionally acceptable and then just enjoying listening to it. After all as interested as I get in the technical aspects I simply love listening to music.

Pim, I hope you enjoy your test, and I look forward to reading what you hear. Please tell us what USB cable you got if you're not too embarrassed (you can wait until after testing it to admit what you bought if you like it).
Reply
#7
USB and Ethernet have protocols to make sure the data is sent correctly. That is why your printed page doesn't differ, even subtly, from the one you sent to the printer...
Frequency is not a problem at the frequencies we are dealing with here. Any jitter affecting the sound will be at internal the receiving end, not caused by the cable, and can be dealt with there before conversion to analogue.

SPDIF OTOH is a bitstream which may have jitter. The waveform can be really terrible, but as long as it has upward and downward parts that can be sensed by the receiver that is all that counts, the beauty of the waveform is of no consequence in the digital domain. Any timing shortcomings again need to be dealt with at the receiving end, usually a FIFO buffer and accurate clock. Some jitter may be added by the cable but only a poorly engineered receiver would be unable to deal with it.

So the quality of engineering in the devices either end of a digital cable would have to be dire for the cable to have any influence on the final analogue output.

Noise transfer is a possibility with USB and Ethernet, possible but unlikely with a normally engineered wired SPDIF input and not possible at all using optical SPDIF. I have seen a USB cable which actually has an aerial within it. This could pick up music related noise from the cables around it, but this is a deliberate manipulation and probably negated by a good receiver design at the DAC.

It is unwise to automatically assume one is immune to the placebo effect.
Placebo drugs are shown to be effective on subjects even when they know the treatment they are having is a placebo, and amusingly/ironically the more expensive the subject believes the treatment to be, the more effective it is, even when it is the placebo.
Our senses play tricks on us.
Devialet Original d'Atelier 44 Core, Job Pre/225, Goldmund PH2, Goldmund Reference/T3f /Ortofon A90, Goldmund Mimesis 36+ & Chord Blu, iMac/Air, Lynx Theta, Tune Audio Anima, Goldmund Epilog 1&2, REL Studio. Dialog, Silver Phantoms, Branch stands, copper cables (mainly).
Oxfordshire

Reply
#8
(03-Feb-2016, 17:05)Jwg1749 Wrote: With the best will in the world, there’s no way of controlling for expectation bias except by blind testing.
...
Matt



Nice post, Matt: kind, open, and respectful. This could be a fun thread.

I agree with much of what you said but I'd caution against putting too much weight on even blind testing. The scientific method is wonderful, but tests are difficult to get right even recognizing all the barriers to objectivity. Your comments about the human brain and subconscious bias are very appropriate. Add in what we know from cognitive and perceptual psychology and you can easily understand why the perfect test is a lofty goal. Science can be as much snake oil as anything else, given the humans behind it. Look at the data about the percentage of even peer-reviewed studies that are critically flawed. So I suggest remaining skeptical about blind tests and other such data even as I've worked extremely hard to create the best tests possible.

Saying that there hasn't been any valid attempts at blind testing of audio equipment (including cables) is dismissive of many good tries. If you're willing to accept the imperfection of humans then we can learn from attempts at rational evaluation. We can attempt to control for the way our minds work through more rigorous experiments, greater sample sizes, etc. This is one path. I've enjoyed reading the Big Sound '15 evaluations of headphone system and their attempts at blind testing. http://www.innerfidelity.com/category/big-sound-2015 in total and http://www.innerfidelity.com/content/big...-i-learned specifically may be of interest. I also enjoy all the measurements of loudspeakers even though many experts recognize that as at best a small part of evaluating a speaker given the flaws in such measurements. Choosing speakers by measurement alone would't be a wise approach (unfortunately for those of us attracted to rational evaluation) even if there's a strong correlation between success and speaker manufactures that do measure their speakers vs. those that don't. I have much more respect for a company that publishes test results of their equipment than those who don't. Measurably lower distortion usually does mean better sound. However as we can see with Devialet measurement without listening can lead to flaws in audio equipment. Devialet firmware updates please some people and displease others but Devialet seems to have no control of that, probably because they measure but don't listen (at least they don't admit to listening). In contrast PS Audio keeps improving their firmware to almost universal acclaim and they both measure and listen when attempting to improve the sound of their gear.

Another path is to accept that the end goal is a happy perception of music so whatever path gets someone there is also valid. Bill Lowe (the head of AudioQuest) says that the value in cables is as much about the relationship between the user and their equipment as it is about the technical details. I love his comments about equipment changes being like bringing home flowers to enhance a relationship, not that the flowers make a technical improvement but rather that your enjoyment changes through participation in the relationship. As someone of very different persuasion I don't enjoy this path and think of it as bringing Bill Lowe flowers in the form of wasted money, but I recognize that others have more of an emotional relationship with their equipment and so may gain more enjoyment through enhancing that relationship with new toys.

There's got to be an appropriate balance where we can be educated enough about the data to avoid the snake oil yet in touch with the reality of human perception and psychology enough to work with people as the not-entirely-rational being we are.
Reply
#9
(03-Feb-2016, 17:17)f1eng Wrote: USB and Ethernet have protocols to make sure the data is sent correctly.
...

Unfortunately for us "the data" that is sent correctly is designed for a machine to use and as you point out we're far from machines (or I could fix my computer with a placebo) so the data we hear can differ. The error correction helps computers receive the data sent, but we're still learning a lot about how to send sound digitally so that the human hearing the sound perceives the sound we're trying to transmit.

When the telephone was invented voice data was transmitted, but it took decades of research to learn how to transmit the right data in the right way to make what came out of the telephone sound like the human on the other end. Then we switched to digital, wireless, and our research hasn't caught up so the sound quality has suffered. We're transmitting more exact "data" now but it sounds far worse because we're paying more attention to the computer communications than the human perception.

Digital music is taking a similar path. When music went digital they promised "perfect sound forever". Whoops. It turns out that digitizing music certainly made it error-free from a computer data perspective, which is very useful when manipulating it but many people found that CDs sounded worse than vinyl. CD players improved, and now DACs are making rapid progress. The improvements aren't to the perfect transmission of bits but rather to other things that affect the sound we hear when those bits are transmitted. It turns out that the digital reconstruction filters are critical as is the control of noise even outside the audio band. All cables act as a filter to some degree. Let me assure you that we'd love to have made a perfect optical fiber but while we've made amazing stuff even optical transmission is imperfect. Those darn analog effects keep our digital data from being perfectly transmitted. Our imperfect understanding of what data needs to be transmitted in what ways keeps the sound from being perfect even if we could transmit the bits perfectly.

We keep improving the data transmission technology (Ethernet Cat 5, 5e, 6, 7, etc.) and one can easily build a robust data network now with ethernet, but that doesn't mean that we understand everything about digital music or making what comes through ethernet sound good. USB has even more technical issues with noise and other detractions from perfect transmission of music data. Luckily we're one our way to understanding them, and if you're interested then modern DAC design makes fascinating reading.
Reply
#10
(03-Feb-2016, 18:14)deviousalet Wrote:
(03-Feb-2016, 17:17)f1eng Wrote: USB and Ethernet have protocols to make sure the data is sent correctly.
...

Unfortunately for us "the data" that is sent correctly is designed for a machine to use and as you point out we're far from machines (or I could fix my computer with a placebo) so the data we hear can differ. The error correction helps computers receive the data sent, but we're still learning a lot about how to send sound digitally so that the human hearing the sound perceives the sound we're trying to transmit.

I would draw a distinction between how to "represent sound digitally so that the human hearing the sound perceives [it as intended]" and transmitting that digital representation from A to B.

Certainly there are plenty of different ways to approach the former and arguably we don't yet have a perfect method.  Yes, different methods of data representation may result in different heard experience for the listener.  But that's a very different discussion, isn't it?

More to the point, I think it's orthogonal to a discussion about digital data transmission.  The requirement there, as I see it, is that the receiver must be able to reconstruct the same bits that were sent from the transmitter, in a "timely enough" fashion depending on what the bit stream will be used for -- be that audio playback, rendering a web page, storing a file, etc.  The USB and Ethernet protocols Frank and I referred to are designed to solve that problem and indeed do so in the real world.
Roon (Mac Mini), Wilson Benesch Full Circle, Expert 1000 Pro CI, Kaiser Chiara
Warwickshire, UK
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)