Computer Audio @ SSI

The Computer Audio 2010 seminar on Saturday was very well-attended—I barely managed to get a seat. The presentation was by Steve Silberman of Ayre Acoustics, with technical commentary by John Atkinson. Silberman took an admirably generic and non-partisan approach, barely mentioning Ayre products, and refusing to answer the question "Should I get a Mac or a PC?" I've taken a wait-and-see approach to the whole computer audio subject, and Silberman did not convince me it's time to introduce a computer into my audio system, but I must say that he did an excellent job of describing the options, and if I were to take the plunge I would certainly use the information on the Ayre web site.

Some of Silverman's statements were intriguing—and sure to give grief to the bits-are-bits folks. In discussing USB cables, he said that the cheap generic cables can significantly limit sound quality. Yes, USB cables have a sound! There are three USB cables that Ayre recommends: Cardas, AudioQuest, and Transparent, none very expensive. For a computer dedicated to sound reproduction, Silberman recommends having 8GB or more memory (for those with 64-bit operating systems; a 32-bit OS will only handle up to 4GB of RAM). Less than 8GB will work, but won't sound as good. Perhaps most puzzling: the sound will be better if instead of a conventional spinning hard drive the computer is equipped with solid-state memory to store the operating system. This is in spite of the fact that the music computer files themselves are stored on an external hard drive of the conventional sort. "Why does this sound better? We don't know. But it does." There are more things in heaven and earth. . . (In later discussion, JA suggested that some of these effects may stem from hard-drive access stressing the power supply and introducing datastream jitter.)

Share | |
COMMENTS
Steven Silberman's picture

Hi Bob-Just a clarification. It is with a windows OS that you need to make sure that your operating system is 64 bit. This is not an issue with Apple. Regards,Steve Silberman

GEORGE's picture

Why would a hard drive "stress" a power supply in a computer. You mean, that now the computer makers don't know how to put the correct power supply in their stuff to allow a hard drive to function correctly? Why is this kind of gibberish made up, now of course there will be "audiophile" computer power supplies right? And it will of course SOUND BETTER with an audiophile line cord on the computer! this stuff never ends does it? If there is a power supply problem in a computer, not allowing a hard drive to access data, you can be sure, there will be errors on other software. Why only in AUDIO, do these errors manifest themselves? Stop the insanity!

Andy's picture

"Stress" a power supply. A computer hard-drive consumes 2 - 5 watts max. This will not "stress" a power supply, considering most all computers have at a power supply that handle at least 350 watts!

John Atkinson's picture

It is not the stress as such, Andy. Perhaps that is a poor choice of words. But even with as many background tasks eliminated as possible, the OS is writing and reading data to RAM all the time, and when it runs out of RAM, it uses virtual memory on the hard drive. While the bits are never changed, George, the flow of those bits from the temporary buffers to the soundcard, USB bus, etc, will never be as constant as it should be, due to all this OS activity. Gordon Rankin of Wavelength did careful listening tests and found a) that any degradation in audio via USB became inconsequential when the internal RAM reached, IIRC, 4GB, and b) that there was no degradation, regardless of the amount of RAM, when a conventional hard drive was replaced with a solid-state drive. Do the tests; report back what you find.

Andy's picture

John, your response makes thing much more clear. "Stress" was definitely the wrong word to use.

Robert Deutsch's picture

The description of JA's interpretation was based on my recollection of what he said about this subject when we were having dinner that evening, and "stress" may well be the wrong word to communicate what he meant. But whatever the technical details, the major point remains: to explain these sonic effects, you have to go beyond "bits-are-bits."

Cybermynd's picture

The bit about running the OS from a solid state disk is interesting. Luckily this is not expensive or difficult these days - especially if you are setting up a computer explicitly for media room use. If your media is offloaded to a server the you can easily get by with a 32GB SSD for the OS.My one comment would be that once you've gone to that trouble shouldn't you cache the music files locally before playing in order to take the network and hard drives out of the equation entirely? I have designed my playback software to download my music files to a temporary file on a USB key and other than a slight delay while the file copies it seems to work very nicely.An SSD solution would be even better because you could cache the files to the C: drive or something and eliminate the need for separate memory.As for USB cables... nope, not going there. I use S/PDIF or optical into my DAC.

nolocontendri's picture

No one serious about computer based audio uses a USB buss for audio data; it's not designed for it. Better to use a dedciated and faster (FireWire) buss (with only your converters on it), or putting in a PCI card with a FireWire out on your machine. Macs are correct/transparent bitwise through their optical outs as long as you keep the internal system volume at max. So from a Mac, you can run optical cable to a decent external converter without having to worry much about jitter (or spending a fortune on cable). Finally SSD's sounding better then hard dirves is ridiculous. As long as you are reading/buffering data faster from storage to memory then you need to send it to/play it back to your D/A, and also not losing sync - then all should be fine. Serious computer audio people store and playback audio files from a separate 7200 RPM sata or esata drive (that does not contain any boot partitions). And SSD drives wear out/fail after a fixed number of writes; defragging them only ages them sooner.

AlexO's picture

I agree with the previous poster that Firewire is by far a better choice than USB for music transfer due to greater bandwidth and faster transfer rates.However, for those wishing to use USB, it is absolutely ludicrous and disingenuous to say that USB cables make a difference in the way the music sounds. Having worked in IT for almost 20 years, I am very confident in my knowledge of digital communication. As long as the USB cables in use meet the USB specs, there will be no issues. A $3 cable will do just as well as a $100 cable or $1,000 cable. Anyone who claims otherwise is absolutely, positively full of shit.

Bauer's picture

> Yes, USB cables have a sound!OMG, this is a joke right? Some know-nothing gives a talk and makes this ridiculous statement, and that's proof enough for you that it's true?

GEORGE's picture

Why do pro machines like Tascam DSD recorder and Korg DSD recorders then have USB to transfer data from and to the units? Download from the units to use their supplied software. What is the "sound" of a USB data flow versus other data? That you CAN HEAR!? Is it because cable purveyors sell "audio grade" USB wires? Like they sell "audio grade" wall outlets? That also advertise in the magazines? Hmmmmm? Any one with an ounce of thought knows it's basic BS. Audiophooles continue to be taken for a ride. There is ZERO facts to back up any of this absurd claims, of course the guy making these claims, doesn't have any tests, data, or anything to support his nonsense. Does he? And i am sure he listened without knowing which USB cable was being used, correct? Or did he listen as he watched the audio cable be changed, to the ""better" one? Sionce all USB cables meet the USB industry specs, therefore it's BS. High end audio, the few industrys, that still get away with scams.

John Atkinson's picture

So "Bauer," "AlexO," and "George," did you ever meet anyone who was as smart as you believe yourselves to be? If you are all as skeptical about audio as you appear to be, why do you even hang out on our site? And George, you do not yet appeared even to try the experiment I suggested to you - about the effect of RAM and SSDs - before you mouth off, perhaps you should actually _try_ the tests others have done before you jeer. You might learn something.

GEORGE's picture

Skepticism is nothing related to intelligence. Being skeptical, is safer. Just because someone claims something at a seminar, doesn't make that the reality either. I seem to recall someone from Nordost making claims about in 6 months he has discovered how wires sound better than others, and the magazine would be testing this claim, how is that coming? Or did you just accept his claims as facts, and move on? And did YOU do a side by side BLIND test of computer audio to prove or disprove if you could hear anything different, between USB wires, or did you just take this claim as gospel. That dear sir, is called being gullible, and a mark. Being skeptical, is logical, especially in high end audio. Have you used your green pen lately, Mapingo discs, Furutech plastic demagnetize, cable elevators,ART magic bowls,"audio grade" wall outlets, freeze any wires, CD's, plugs lately. you get my drift, so much nonsense, so little science. So much advertising, marketing, so many dollars spent on advertising.

Bauer's picture

"Gordon Rankin of Wavelength did careful listening tests" - Listening alone to identify subtle changes in quality is never as good as measuring. It's easy to record the output of a sound card! That will reveal brief dropouts clearly with no chance for human error to muddle the results.

John Atkinson's picture

Bauer, when you wrote "It's easy to record the output of a sound card! That will reveal brief dropouts clearly," there are no dropouts. The data presented to the AC are the same under all circumstances. So if the data are the same, the timing with which those data are presented to the DAC must have changed. Both need to be correct for the data to be converted to analog without old-fashioned distortion being introduced. See, for example, the articles on jitter on this website and the jitter measurements I perform with every test of a digital product. George: without doing your own tests, you are merely echoing your own lack of direct experience. Some things that you might think matter are not so important when you do a listening tests. Others that you initially dismiss do turn out to matter.

AlexO's picture

John,This has nothing to do with "meeting anyone as I think I am". This has to do with making unsubstantiated, fraudulent claims.If memory serves, you made a claim a while ago about wired data transfers via ethernet being superior to wireless data transfers via wi-fi until someone told you to do a bit check on the transferred files to see if they're the same. After that, you acquiesced that wireless and wired data transfers had no effect on the sound.Point being that our hearing is flawed. Anyone can say anything on any subject and claim that they "hear the difference". I'm sure you heard the difference before you ran the tests and realized that the difference was all in your head.As far as "hanging out at your site" is concerned, if you notice, I don't post to your forums anymore and I let my subscription lapse. I don't "hang out" here anymore. I pop in once in a while to check Stephen's band recommendations. Sometimes I stumble on bullshit like this and I post a comment against my better judgment.

John Atkinson's picture

?If memory serves, you made a claim a while ago about wired data transfers via ethernet being superior to wireless data transfers via wi-fi ...you acquiesced that wireless and wired data transfers had no effect on the sound." I don't think that was me, AlexO. And when you write, "Sometimes I stumble on bullshit like this..." you make my point. Because when you write something like "I can see no reason for this making any difference," you are in effect claiming to be in possession of all relevant information on this subject. Which is both arrogant and incorrect. Its is the unexpected phenomenon that leads to progress. Sorry you don't read the magazine any more, but to be frank, I never grasped what it was you valued in our content, given your skepticism.

AlexO's picture

John, I would never argue about analog communication. I may have my doubts about various claims and such, but I wouldn't argue about it because I don't have the expertise in the field and it's too easy to throw technical jargon at me and I wouldn't be able to say what's legitimate and what isn't.I do however have a very good grasp of digital communication and when it comes to claims regarding the "sound of USB cables", I am very comfortable calling the claimants on their bullshit. Oh, and yes, it is bullshit.I gave up reading Stereophile at your urging. I was an avid reader for over ten years in great part because I bought into the premise that audiophiles perfected what consumer market developed. I realize now that this isn't the case. The audiophile market is a self serving microcosm that repackages what consumer market develops. It's really the pro-audio market that perfects AND develops. When it comes to the audiophile market and publications, the emperor has no clothes. Most people see it, about 80,000 do not.

GEORGE's picture

I seem to recall a gentlemen, named DUP, or perhaps a female, that used to question the merits of many of the high end claims and observations of many of the magazines writers. Since many of these "trained" ears claim some pretty outrageous ability, like a chap called Mickey, who said he heard and it was plainly obvious to others that a demagnetized VINYL record sounds better. Now I never seen him in ages, he usually made more sense than most, when it came to being highly skeptical on the crazy and their ideas. I have been a subscriber and reader since back in the original days, pre advertising, and it used to based on reality, it has really gone out to some weird world of claims that are really seemingly and imaginary idea. If everyone tried everyone of these claims, to prove or disprove, why would there be a need for the magazine, shouldn't you as a reviewer do such? But i think Alex is correct, no matter what the claim, the magazine endorses it, and let's the claimants get away with anything they claim.

John Atkinson's picture

AlexO, when you write "I do however have a very good grasp of digital communication and when it comes to claims regarding the "sound of USB cables", I am very comfortable calling the claimants on their bullshit. Oh, and yes, it is bullshit," perhaps, yes, you do about data communications as you claim. But does that knowledge translate to audio? And you have avoided mentioning any such listening tests you yourself have done when you say that "$3 cable will do just as well as a $100 cable or $1,000 cable. Anyone who claims otherwise is absolutely, positively full of shit."At the SSI seminar, Steve Silberman was discussing the effect of different USB cables - none of them expensive, BTW - in the context of the Ayre QB9 DAC. This powers the TAS1020B receiver chip and its associated circuitry not from the QB9's supply but from the USB bus via the cable. Perhaps, if differences are heard, they stem from the cables having different series impedances, hence slightly different induced jitter signatures.

John Atkinson's picture

AlexO write: "I gave up reading Stereophile at your urging."Yes, because I could not understand why you, as someone who appeared to despise high-end audio, were reading something that gave you so much unhappiness. And in your interactions on our forum, you were like a fiber-glass boat owner jeering at those who liked wooden boats."When it comes to the audiophile market and publications, the emperor has no clothes. Most people see it, about 80,000 do not." - AlexOI am sorry, the "most people agree with me" is always conjecture on the part of the one using it. As I said earlier AlexO, it would do you good to try things out before you describe them as shit.

John Atkinson's picture

George write: "I seem to recall a gentlemen, named DUP, or perhaps a female, that used to question the merits of many of the high end claims and observations of many of the magazines writers.""DUP" was a reader named Carl Engretsen whom we eventually banned from our forum, not because of his views, but because his antisocial, hectoring, bullying manner and his continued refusal to accept that other posters might have different points of view and tastes from his own, was driving readers away from that forum.

AlexO's picture

John,USB data transfers are digital transfers. Hence, they fall under digital communication. It makes no difference at all whether the encapsulated data is music or data files. As far as the tests are concerned, the proof is in the pudding. I can transfer various data files, including music files via ANY USB cable that meets the USB spec and these files will have the correct checksum and will be indistinguishable from one another. There's your test.That test actually tests the cables and proves that the cables meet the specs and do not affect the data. If you're telling me that a difference in USB cables can be heard via whatever DAC, then I suggest that the hearing is flawed. This isn't rocket science. Really. Oh, and I don't mind you urging me to give up reading Stereophile. You are absolutely right: there's nothing within its pages for me.

John Atkinson's picture

'John, USB data transfers are digital transfers. Hence, they fall under digital communication. It makes no difference at all whether the encapsulated data is music or data files." As I have said, AlexO, no-one is claiming that bits have been changed by the use of more RAM or a SSD or by a different cable. But what you don't seem to be able to grasp is that any error in the _timing of the conversion of each digital word to its analog equivalent _also_ introduces measurable and audible changes in that analog signal. This is well-documented in the literature, and I have written about it at length in Stereophile. I also offer measurements of the effects of this phenomenon in every issue. Yet you seem unaware of it. So before you continue claiming to be an authority in this area and before you continue to dismiss the results of tests others have done as "shit" it really does behoove you to go back to your books.

AlexO's picture

John,The errors in timing will not be the function of USB cables. A properly designed DAC won't be susceptible to this. Besides, increasing the buffers will alleviate these issues even if they're there. As I said, it's not rocket science.

John Atkinson's picture

"The errors in timing will not be the function of USB cables."I offered a possible mechanism that you ignored AlexO."A properly designed DAC won't be susceptible to this. Besides, increasing the buffers will alleviate these issues even if they're there. As I said, it's not rocket science."And you are now ignoring all the measurements of DAC jitter susceptibility that have published. Once jitter has been introduced into a streaming D/A chain, it can only be low-pass filtered, not eliminated.

Stephen Mejias's picture

I seem to recall a gentlemen, named DUP, or perhaps a female, that used to question the merits of many of the high end claims and observations of many of the magazines writers.For the record, "DUP" and "GEORGE" are the same person, Carl Engebretsen. The comments he posts come from the same e-mail address, and use the same language. Carl has even e-mailed me, signing off as "GEORGE," I suppose not realizing that his real name, Carl Engebretsen, shows up in the e-mail data. The fact that "GEORGE" even defends "DUP" is disturbing and a little bit sad.

John Atkinson's picture

AlexO wrote: "Oh, and I don't mind you urging me to give up reading Stereophile. You are absolutely right: there's nothing within its pages for me."I didn't respond to this right away because I wanted to check my email archive. AlexO, last October you demanded, among other things, that Stereophile stop telling people that LPs could give a better musical experience than CD, failing which you would stop reading the magazine. In response, yes, I did say "If you really feel this way, then please do not read Stereophile. I fail to see what possible benefit you can get from it....I'll be sorry to see you go, AlexO. But you really don't seem to grasp why and what a magazine like Stereophile does. We are not Consumer Reports; instead we are like a forum where ideas are discussed and where products that embody those ideas are tried out and described....I don't believe the role of a magazine like Stereophile is to impose authority from above but to allow understanding and consensus to emerge from below."

AlexO's picture

John, you have urged me to allow my subscription to lapse on multiple occasions. It's ok. Really. I don't hold a grudge against you in the least. You were absolutely correct in that you recognized that Stereophile, its culture and mentality does not mesh with my own. Furthermore, you recognized that way before I did.

Steve Silberman's picture

Hello All-After the discussion I invited anyone who was interested to come upstairs and do some listening. We listened to generic USB cables vs. the AudioQuest, Cardas and Transparent. Everyone felt that the premium USB cables made an improvement. Steve Silberman

Pages

X
Enter your Stereophile.com username.
Enter the password that accompanies your username.
Loading