The State of USB Audio by Alan Taffel

[The Computer Audiophile]

Hi Scott - The premise still holds, no worries.

 

One writer from a traditional print magazine will not decide the fate of USB audio for a whole industry. The future has never been brighter for audiophile sound quality and USB will be a huge part for the foreseeable future.

 

 

 

[johnnyturbo]

sorry. no free lunch: never was and never will be; but one can purchase a solid healthy lunch from a rather diverse and ever expanding and menu at a very reasonable cost. do not lose faith. keep reading the CA as well as the AA PC-audio.

 

 

"If they get you to ask the wrong questions, they don't have to worry about the answers"

Thomas Pynchon

 

 

[Audio_ELF]

I don't disagree that USB can be a great interface for a DAC when done well (Wavelength, Ayre, dCS) but many products have been (IMHO) jumping on a bandwaggon adding a cheep USB interface without really understanding it and in many cases doing a job best described somewhere between mediocre and piss poor. For many audiophiles who want to daddle in computer as a source, they are looking at iTunes or out of the box Media Monkey or J.River or Foobar and don't want to start messing with ASIO, etc. It's these people who buy a Bryston DAC, plug it in via USB and find it poor compared with atavhibgtheir CD transport via SPDIF or AES. And it's to these people that TAS article (which I've not read to be honest) is probably talking about. Yes the details maybe were lacking with deails of FireWire DACs (though to be honest I wouldn't class Weiss as an Audiophile company, more a pro audio company looking to sell to audiophiles with their Jason / Medena / Minerva range) but by his own omission Gordon withdrew his DAC from being reviewed beause the reviewer wasn't competent with Computer systems - though I'm sure a lot of his caomera are darting at a similar place.

 

If you're an audiophile used to plug and go of a CD player, then (most) DAC don't give their bet performance through USB conenction - you onl need to see just how many posts there are asking about Lynx / RME / ESI cards and FireFace / ProFire 610 / AFI1 FireWire interfaces into DACs such as the Bryston which has a USB conenction.

 

Finally some people commented about how the article tars all USB DACs with same brush. Part of that is probably down to how the article was editted rather then the original author. Writing USB is alot shorter than non-asynchronous USB each time. Also as I commented Gordon did decline to allow the Wavelength to be tested whch MAY have given the article a different spin assuming that the author was being unbiased.

 

Eloise

 

[rlodad]

Eloise,

Re: I don't disagree that USB can be a great interface for a DAC when done well (Wavelength, Ayre, dCS) but many products have been (IMHO) jumping on a bandwaggon adding a cheep USB interface without really understanding it and in many cases doing a job best described somewhere between mediocre and piss poor.

 

Absolutely correct, and the root of the entire problem. The other question, though, is whether the reviewer is biased by not having any good USB set-ups available due to his perceived bias by the manufacturers, or just plain uninformed. There is no reason to write asynchronous USB when you haven't heard any. Read the article. Truly nuts, IMHO.

Brad

 

[ataffel_tas]

I am delighted to see so much interest in my recent TAS article on “The State of USB Audio”, even though much of what has been written here and elsewhere takes issue with my methodology and findings, not to mention questioning my competence! Because this is such an important and timely topic, I thought it appropriate to clarify and expand upon elements of my piece, and to respond to accusations that have been leveled against me personally.

 

The article’s perspective: I believe it is important to understand the purpose and intended readership of the article. I consciously approached the topic of USB audio from the perspective of a typical high end audiophile (that is, a typical TAS reader), who is not necessarily a computer whiz. My aim was to determine, from this perspective, the results that could be achieved from USB both on an absolute and relative basis. My goal was not to assess USB in a vacuum, nor was it to determine what a highly sophisticated user might be able to do with it. In general, high end audio does not require those who enjoy its products to have expertise in those products’ inner workings. For computer audio to be broadly relevant to this market, it must meet this criteria as well as sounding good.

 

The author’s qualifications: Although many posts have implied otherwise, I am actually quite proficient in computer technology. I hold a degree with honors in Information and Computer Sciences, and have had a twenty-year career centered around data communications protocols. Suggestions that I might not understand the technicalities of asynchronous USB, for example, are incorrect. However, as noted above, my purpose was not to approach this research from the perspective of a computer expert.

 

Bias against USB: Quite a few posts have alleged that I was pre-disposed against USB, presumably because my results were less than glowing. By this logic, any negative review of any component or technology could be chalked up to bias rather than impartial observation. However, the fact is that I would have absolutely no reason to harbor such a bias, and indeed I did not. On the contrary, I would have enthusiastically reported more positive results, and I was disappointed at the actual outcome. I think the fact that I employed three different PC’s (from different manufacturers and running different operating systems), tried vainly to include a Wavelength example, tested multiple DACs and software programs, and even experimented with expensive USB cables attest to my efforts to give USB every opportunity to shine. Whether or not you agree with my results, know that they were based solely on what I heard. At the same time, it is worth noting that many of those suggesting, without foundation, that I have a bias actually do have a demonstrable, commercial motivation for promoting USB and for denigrating any negative opinions (and their source) about the interface.

 

The choice of DACs: For this project, I tried to round up as many exemplary USB DACs as possible. Audio Research and Bryston were chosen because both firms have solid engineering and build quality, and both understand good sound. Benchmark employs a highly respected, purpose-built USB input module. I sought mightily to include a unit from Wavelength because, as noted in the article, its technology is innovative and I actually do understand its promise. Wavelength refused to participate (more on that below) and Ayre, which uses similar technology, has by its own admission completely severed ties with TAS. Ergo, these latter two units were not available to me. Their lack of inclusion is unfortunate, but cannot be construed, as some have, as a desire on my part not to give USB its due. The units were excluded by their manufacturers, not by me.

 

Wavelength’s withdrawal: I have not been at all surprised to read Gordon Rankin’s reaction to my article, or his statements about my qualifications to review his equipment. In response, I would simply say that if I, with my education and experience, am not qualified to get the best out of his gear, then neither is TAS’ readership. However, I believe the issue of my qualification is a red herring. In our conversations, Gordon became familiar with my background and—although he now states otherwise—indicated comfort with it. As evidence of this, note that it was after these conversations that Gordon sent me a Cosecant for review. Things fell apart over a different issue: my intended test bed. Gordon’s preferences in this area are well known: powerful Macs with SSDs and massive amounts of RAM running iTunes. I did not feel such a configuration would be typical of our readership, and when I indicated the test beds I planned to use, along with my (valid) issues about iTunes, he demanded the return of his unit. As much as I would have liked to have heard the Cosecant, my feeling is that if Gordon’s recommended configuration is a prerequisite to his DAC (and by extension, USB itself) sounding good, it only reinforces my conclusion that the interface is not yet ready for the world of high end audio. Still, I wish Gordon had had the courage to subject his DAC to a controlled evaluation—where it would be compared to other USB and non-USB DACs—in a typical audiophile environment. I am certain the results would have added significantly to the conversation.

 

USB vs S/PDIF: Some posts have argued that USB actually does sound better than S/PDIF, if only the right DAC is used. These posts point to good experiences with DACs from Wavelength and Ayre, and the positive review by my colleague Steven Stone of the Streamers products. The fact that USB can sound pretty good (or even very good under narrow conditions) does not mean it is better than S/PDIF. I think it is noteworthy that none of the aforementioned products offers an S/PDIF input, making a true apples-to-apples comparison impossible. On the other hand, I used DACs that did enable such a comparison and in every case, including the Benchmark—which certainly does not treat USB as an “afterthought”—S/PDIF sounded clearly superior. Please don’t take my word for it, try it yourself—if you can be unbiased.

 

USB vs FireWire: I am glad to see that no one (yet!) has challenged my assertion that FireWire is a superior means of getting audio out of a PC. Several posts have actually pointed to technical reasons why this is true, and I would simply add that FireWire excels without the need for the elaborate technology patches USB apparently requires. My statement that “no FireWire DACs exist” was meant to say “no audiophile grade FireWire DACs exist”, and I owe readers an apology for not being clearer on this point. As many posters have rightly pointed out, there are a number of pro FireWire DACs available. My intent was to challenge high end companies to make a similar product built specifically with high end consumer sensibilities in mind. Perhaps this is one area in which we can all agree.

 

I hope the above serves to clarify what I was trying to accomplish with my report, and puts my results in a clearer context. Within that context, I believe those results are completely valid, and will prove useful to their intended audience. And while I don’t have the time or resources to engage in one-on-one discussions with everyone commenting on the article, I do look forward to reading additional comments.

 

 

[The Computer Audiophile]

Hi Alan - Thank you for posting your comments. There are always at least two sides to every story. While I still disagree with much of what you said, I do respect the fact you've posted a well thought out response.

 

You can count on the readers here at Computer Audiophile to show respect in their responses to your post.

 

[cfmsp]

 

"Here is the biggest advantage that USB computer usage has over the flawed S/PDIF interface. Firewire is used in the Pro market and we al know the pro market is not as good at the Audiophile market. Firewire usage requires additional drivers, which is a no no IMHO and can cause some configuration issues. I also believe it jacks the price up which I despise."

 

LIzard,

your words above are inconsistent with reality, IMO. the 'pro market' has been manufacturing DACs for a lot longer than audiophile firms. They use Firewire - the protocol designed for high bandwidth audio data transmission - becuase it's better than any other protocol (until the recent equalization with Async USB).

 

The only disadvantage that I'm aware of with Firewire is the need to write drivers, In my experience, there are NO configuration issues, just download, install and play - as easy as an iTunes install, and faster. That may not be true on your preferred OS, I can't vouch for lack of configuration challenges of WinDOHs. Indeed, most of the highest quality Pro audio software / hardware I'm aware of is purpose built around Mac OS X (Metric Halo, Wave Editor, etc.).

 

Incidentally, industry luminaries seem to agree with me that audiophile companies have no 'advantage' over pro audio companies when it comes to computer audio. As reference, I would refer you to the dozen or so interviews with the computer audio 'household names' published by Positive Feedback Online in issue number 41.

 

enjoy

clay

 

 

 

 

 

 

[cfmsp]

 

Alan,

 

first of all, thanks for such a detailed explanation into your thinking.

 

I'll simply raise one issue:

 

You state:

"My statement that “no FireWire DACs exist” was meant to say “no audiophile grade FireWire DACs exist”.

 

I believe this is still incorrect. There are certainly 'audiophile grade' Firewire DACs in existence. For future reference, the Metric Halo ULN-2, and the new world beater, ULN-8, along with the Weiss (and others) Firewire DACs, are certainly audiophile grade, Indeed, IMO, they are ABOVE audiophile grade if we are to use the DACs in your survey as representative of 'audiophile grade'.

 

Perhaps you meant to say - no purpose built Firewire DACs have yet been offered by an audiophile company?

 

respectfully,

clay

 

 

 

 

[Lars]

Alan,

 

You reviewed dacs without even understanding the USB interfaces involved. Did you take the time to even try to understand what was involved? Your computers were not worthy of Gordon's Cosecant. You made a ridiculous assumption that people who read about $68,000 speakers would not purchase a decent computer for their high end system. TAS is about sound excellence; you instead took the low road and not only are you proud of your work, but blame Gordon Rankin for your lack of universal accolade.

 

 

 

 

[Audio_ELF]

Alan, I for one would like to thank you for coming and explaining your thoughts behind your article. Unfortunately I'm in the UK so only get to see TAS if there is a copy in my local Borders store but the reporting of this article has made me want to track down a copy if at all possible.

 

I do agree with others that your "no audiophile grade FireWire DAC" is misleading, as others have pointed out Weiss are certainly targeting the Audiophile market with their Minerva DAC and it is (from all I've seen) a product certainly worthy of the "Audiophile" tag. You can question the devices such as the Metric Halo ULN-2 is a Audiophile product having many features beyond the requirements for integration into a HiFi system, however in terms of the SQ this, and many other pro-audio devices (the Benchmark DAC-1 is from a pro-audio company), can be up with the best from Audiophile Manufacturers.

 

Moving on from Alan's comments, we come to Lars ... you say "You reviewed dacs without even understanding the USB interfaces involved. Did you take the time to even try to understand what was involved?"

I think you have missed one of Alan's fundamental points ... that currently to get a good Computer Audio system, you DO need to start to understand the interfaces involved. Whether he does or doesn't understand it himself, he is writing as a layman who has (maybe) just bought a new computer for their work, has his old one laying around (I'm sure I've read somewhere what computers he used for the test and they were previous generation) and either has, or decides to invest in, a DAC with a USB interface. Currently you need to know about how the DAC has been built to decide if the USB interface is worth persevering with, or if you'll get a better quality sound with a FireWire interface offering an SPDIF or AES interface; or "pro-audio" PCI card.

 

Carrying on, Lars wrote "Your computers were not worthy of Gordon's Cosecant. You made a ridiculous assumption that people who read about $68,000 speakers would not purchase a decent computer for their high end system. "

So, what computer is "worthy" of Gordon's Cosecant? If previous generation computers are not worthy for Gordon's DAC, but can produce a high-end system with a FireWire to SPDIF interface, then that isn't Alan's problem. I don't think it's uncommon for people to want to trial a computer audio system with their last computer before spending £500 - £2000 (or more) on a dedicated computer to slide into their HiFi Rack.

 

Finally Lars commented "TAS is about sound excellence; you instead took the low road and not only are you proud of your work, but blame Gordon Rankin for your lack of universal accolade."

Yes, TAS is about sound excellence and currently, with most DACs and all those that were tested, the BEST sound available from them is NOT via the USB connection. There are many posts on here, and other forums commenting how better results (even with 16/44.1, 16/48 CD quality) are obtained using a separate computer to SPDIF interface into DACs such as the Bryston, Audio Research and (yes even) the Benchmark DAC-1. In his rebuttal of people's comments, I don't feel he "blamed" Gordon, he just explained the withdrawal of his product from testing from his point of view as Gordon had already put his own side.

 

Eloise, trying to be balanced

 

[Lars]

Eloise,

 

I purchase The Absolute Sound to be informed on the status of high end audio. If I want the common man's approach, I'll purchase Consumer Reports. Would you play the Wilson Maxx3 speakers with a $200 amp from WalMart? That's the equivalent of what Alan did in his report.

 

Lars, trying to be enlightened.

 

[Audio_ELF]

I'm not sure, but isn't the "status of high end audio" that there are many people who have traditionally bought high end vinyl and CD replay are starting to dip their toes into Computer Audiophile. They are then blindly buying a DAC because it's from a renowned manufacturer and getting good reviews in the press (i.e. the Bryston) and often just connecting a USB cable they have lying around to the USB port. Then they are getting disappointed as it's not up to the quality of they Bryston / Krell / Audio Research / etc CD player. At least the article wasn't blindly advising people to use the USB port which many articles (at least in Britain) have suggested you can do, without having done any tests on them.

 

As for what computer people are using, there are MANY people who do use exactly the same type of computer as Alan did with good results when they connect it in a good way (i.e. via FireWire or async USB). Yes people like Gordon advocate the latest, greatest computer with SSDs, etc. but many people (on this an other forum) are using older computers and getting results as good as they did with CD players, so I really don't think the computers chosen were responsible for the results. As Alan points out, we only have the manufacturer's information (i.e. Ayre and Wavelength) that their DACs sound better with USB than they could with SPDIF as USB is the ONLY available interface (which isn't necessarily a bad thing, but is a valid comment).

 

Eloise

 

[laurent]

Hi to all,

 

I haven't read the article from TAS but I think that it does show that having a good usb implementation is not an easy thing. It is also what Mr RANKIN means when he says that it took him years to developp asynchronous usb.

That is certainly why Alan think that usb DACs do not work as well as their Spdif counterparts.

 

On the contrary, with firewire it is the normal mode to have the computer slaved to the Dac. In a way, asynchronous usb is a way to make usb work like firewire...

 

But asynchronous usb or audiophile firewire DAC are not main stream audiophile products and this article is a warning (justified?) against standard usb implementation. Why would a company like Bel Canto design the usb link if they were satisfied with their usb input in their DAC.

 

If this article had been written been two years ago, it would have been very interesting but now it is seriously outdated since it does not take into account sate of the art products.

 

However like the author, I also wonder why there are so few firewire DACs. It is an easy way to go asynchronous and it can go up to 24/192 whereas usb is still limited to 24/96.

 

Laurent

 

 

 

 

 

 

 

 

 

[enyonam]

I got curious about this cp audio thing a few years back. Decided to aim high. So I bought an EA overdrive , Locus Design Polestar, Macmini with the latest specs and put Amarra on it. And guess what....., compared to my old AudioAlchemy DDS-pro, I now miss the color, micro-dynamics, drive ect. of it . I really wonder how many people HERE know the sound of a very good turntable. Compare that to the misery presented here!! Guess what.... just sell everything ( except for the overdrive )!!

 

[Purite Audio]

I have a very good turntable, TW AC Graham Phantom, Colibri and XV1-S , extremely good turntables tend to sound more and more like extremely good digital sources, what you don't get with digital of course is the tracking and tracing distortion an the increased noise floor, btw Amarra was only released this year so I doubt very much you used that a few years ago.

 

[cfmsp]

 

"the reporting of this article has made me want to track down a copy if at all possible."

 

eloise, if you have paypal, and paypal can handle currency exchange, you can download a copy of the article here.

 

http://www.avguide.com/theabsolutesound/the-absolute-sound-august-09

 

 

 

[Audio_ELF]

Will see if they have a copy in Borders later in the month, otherwise will look at paying for a copy.

 

Eloise

 

[zerung]

I went digital after reading one of my favorite audio writer Jeff Day say his piece on + feedback. I bought the mhdt Havana there and then and was very impressed.

http://www.positive-feedback.com/Issue38/mhdt_paradisea.htm

 

That I went Weiss DAC2 thereafter was a reinforcement on my belief in the Computer based music delivery systems. That I had a Request music server before was to show how desperate I was for this kind of route for music.

 

It amazes me that even till date there is so much of mistrust, mis-information and general apathy towards digital. That the irony of the Computer Audiophile Symposium being just over - with one of aim being to help clear the air surrounding this digital flux. That the tree fell in the audiophile jungle, That TAS did not hear, means that the tree never fell.

 

[Wavelength]

Alan,

 

I would like to clear up a something that Alan may not remember. Here is the email from 9/17/08 that I sent Alan before I demanded the Cosecant to be returned.

 

Alan,

 

People have been tauted with lackluster reviews of Computer Audio for sometime. They want the meat now not the sizzle. Doing a survey now is silly. That means your not going to get into any of the products to any degree. Your not going to get into the technology. Your not going to get into the setup.

 

Believe me I talk to 100 people a week. This is what they are waiting for.

 

Now that I understand it's a survey... I am even less enthusiastic about doing it.

 

Thanks

Gordon

 

Really if anyone read the article they would not have called it a survey, heck not even a Primer. What valuable information would anyone have gotten from these writings Alan?

 

Alan you may not recall but I have designed computers since the late 70's. I have had more than 140 commercial software products based on Windows, UNIX and Macintosh computers. I have more than 3 different full PC platforms for which I have more than 5.4M PC's in the field. Along with that I have more than 68 hardware products for PC's including ISA and PCI cards. My specialty has hardware low level interface, device drivers etc...

 

I would say I know more about PC's and Windows than I do about Macintosh Computers.

 

The reason I use MACs is because it sounds better. As I stated in several emails about all of your hardware that it was pretty sub par compared to what people are using in the field.

 

I use MACS with more memory and SSD's because it sounds better.

 

Jonathan Valin lives down the street from me. I have known Jonathan for many years and the one thing he taught me was that setup was the key to a crucial review. On many occasions I would go there and he would be setting up something new, not to my taste but something non the less. I would come back a week later and it would be sounding a ton better. This setup really helps to describe the capabilities of what the products were.

 

This is where Alan totally failed in all of his articles, he never even understand the basic setup procedures.

 

But really a better understanding of why I think computer audio sounds better with better resources is simple. I think it really comes down to task swaping, virtual memory and computer performance. Virtual Memory is basically non critical memory (usually large chunks) that is stored on a hard drive until required then the OS/Kernel will swap this in putting some other process's memory on the hard drive. When it comes to audio programs there is a lot of virtual memory because large songs require lots of memory. Even more so with Lossless file types as it would require 2x the memory to decode on the fly. This is all virtual memory.... the faster it can come in and out the better.

 

BUT!!!!!! Even better than that... if you have a ton of real memory then the Virtual memory does not go to the hard drive it stays resident all the time.

 

Alan, heck even your old friend Steve Nugent a PC user to the hilt is now feels the best platform is his new MAC Mini with Amarra/iTunes.

 

Anyways... gang after Alan and I had a few phone calls and the editors pushed me some more. I said WAIT**** this is going to end up a big mess and just confuse the heck out of people. This is when I asked for the Cosecant to be returned.

 

I am pretty sure that what I read was exactly what I thought would happen.

 

Alan you never spent the time like Valin would have to really get inside and see what is capable. This is what I derived from talking and the 20 or so emails we exchanged.

 

Thanks

Gordon

 

 

[Wavelength]

Guys,

 

CG just sent me an email that made a lot of sense....

 

These are the guys testing and touting very expensive solutions for everything, but now think that readers would balk at spending $600 for Mac Mini to run their computer based audio system. That much money for an interconnect or power cable is fine, but for a source.

 

Yea look ok let's say you are like me and want to get into this hole hog...

 

$599 Mac Mini

$50 takes you to 4G of memory

$150 Samsung 64G SSD

 

Heck my reference Nirvana T2 SPDIF digital cable cost more than that

 

Let's get real!

Gordon

 

[Lars]

Eloise.

 

Please read the article before commenting. You seem to have missed the poit about our objections to Alan's article. An informed comment from you would be greatly appreciated.

 

Thank you

 

[Audio_ELF]

Gordon,

I think to be fair you need to consider that a lot of Audiophiles used to CD players, etc., would be uncomfortable taking apart a MacMini and adding components themselves. So unless you can get an independent dealer to do it (an authorized mac shop wouldn't I don't think) you can't have a MacMini with an SSD. Also isn't the 64G Samsung SSD you quote an MLC device where SLC ones are recommended? Anyway from the Apple Store (only way to keep your warranty I think) it's more like $1,150 for the MacMini with improved CPU (you said yourself better CPU helps) and 4GB memory. That's over half the price of the Bryston DAC used in the test, and more than the cost of the Benchmark. Yes TAS do review high 5-figure components, but that wasn't the starting point for this article.

 

Finally, in response to your previous comment of "the article turned out as I expected it" maybe it turned out that way because he was unable to get any good USB device to test, i.e. you withdrew your Cosecant DAC from testing. If there was a Wavelength device for him to test, maybe his result would have been "except for the Wavelength, I find that the SPDIF connection into a DAC (via the FireWire to SPDIF interface) to be superior to the inbuilt USB connections". We'll never know because he was unable to test a good USB device and so he's tarred all with the same brush, when we (i.e. computer audio enthusiasts) know that there is more than one type of USB interface.

 

Eloise

 

[Audio_ELF]

Lars writes ... "Please read the article before commenting. You seem to have missed the poit about our objections to Alan's article. An informed comment from you would be greatly appreciated."

You are correct I should read the article... no further comment will be forthcoming from me until I get the opportunity to.

 

Eloise

 

[CG]

Eloise,

 

At least around where I live, the independent Apple stores will all install memory and new hard drives and honor the warranty. If I recall, Apple warranty would also still apply as well. Just don't do anything that requires a soldering iron.

 

The total price would be a lot less than the price you posted.

 

Compared to what most people would likely qualify as a "good" transport with SPDIF output, it's pretty cheap.

 

Yeah, it's not perfect - what is? - but it works pretty well as a music source.

 

 

 

[Lars]

Eloise,

 

Most of the SSD drives being made today are MLC and have superior specs to the original SLC drives. The new ones are faster and hold more data. Do you have a turntable? Installing a phono cartridge properly on a turntable takes more skill than installing a hard drive in a computer.

 

Again, the point of my objection to Alan's article was that he did not research USB audio properly for high end audiophiles. High end audiophiles will buy the proper computer if advised. TAS is not an economy magazine. Please look at the components being reviewed in this issue. This is the point I am making. Sure, if Alan wants to suggest out-of-date equipment that's fine. But I won't pay for this type of review in TAS. This is not what the magazine is about. Never was. Look again at the name of the magazine.