Computer Audio Asylum

Music servers and other computer based digital audio technologies.

Return to Computer Audio Asylum

Message Sort: Post Order or Asylum Reverse Threaded

How can USB performance impact audio quality?

76.89.155.252
Posted on June 12, 2011 at 13:13:55
mr.bear
Audiophile

Posts: 4167
Joined: November 13, 2001
 It seems it *should be* that once you've gotten the data outside of the computer via USB and its safely being read by your DAC, it must be clean, or at least clean-able.

It *should be* 100% the sound of the DAC you're hearing-- how it handles clocking/jitter, the quality of the A/D conversion, filters, analog circuit design and execution being the whole show. You're not listening to the USB, it just sends data packets to the DAC.

That theory has some holes though; for example reliable, inelligent folks report audible differences in USB cables. I don't want to believe that-- what could explain that phenomenon?

Along with this, I want to believe that my DAC won't be severely impacted by minor electrical noise coming in via the USB cable. Can it? How does one deal with that?

If the discussion comes down to nothing but the audibility of jitter, you might lose me. It mirrors too closely the debate over THD percentages in transistor amplifiers in the 70's!

 

Hide full thread outline!
    ...
RE: How can USB performance impact audio quality?, posted on June 12, 2011 at 13:26:02
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 The noise rides along with the bits on the USB cable and the circuitry in the DAC doesn't do an adequate job of cleaning this off. The noise can ride on the signal wires in the form of amplitude and phase variations or it can travel as common mode noise. In addition, even if logic circuitry receiving the USB signal completely cleans the signal this receiver circuitry creates its own digital noise which may couple through proximity, power or ground into downstream analog circuitry (which includes the clock running the DAC). As if that's not enough, the noise can couple out power wiring into downstream amplifiers.

None of this is logically necessary in the sense that some amount of noise is logically necessary with an analog tape machine, but it's still there. More effort has been spent on marketing hype than on the necessary hard core mixed signal engineering.

Of course, I've only listed the "known unknowns". It may be that there are "unknown unknowns" as well. The problem is tough because people can hear things that aren't easily measured, but they don't hear them consistently and so listening isn't by itself sufficient to solve the problem once and for all.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: How can USB performance impact audio quality?, posted on June 12, 2011 at 13:40:03
Ryelands
Audiophile

Posts: 1868
Location: Scotland
Joined: January 9, 2009
 More effort has been spent on marketing hype than on the necessary hard core mixed signal engineering.

Ah, well, never let the chance of good sarc go a-begging, espeically if there's a newcomer to be had.

There is a deal of genuine expertise on this list on USB and matters audio but there is perhaps also more crap talked about it than on almost any other topic that gets regularly aired here.

A good place to start is at the link below. It's not too long, it's amusing and the writer knows his stuff. Of course, he's not talking about high-end audio but the basic issues are the same and are IMHO well explained.

 

RE: How can USB performance impact audio quality?, posted on June 12, 2011 at 18:08:26
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 Yes, I remember that article from a while back. If it wasn't clear before reading it, it should be clear just how absurdly complicated the USB interface is. You basically have to have a computer to talk to it. Contrast that with a simple system such as AES/EBU/SPDIF or the even simpler IIS, where you need nothing more than a state machine.

One has to assume that any digital electronics running in the USB clock domain have been horribly polluted by all of the complex and arcane aspects of the computer. To get beyond that requires specific, engineered isolation and probably multiple stages of isolation at that. Some products have done some of this, but apparently they have not done enough. Still, at this narrow part of the system this is the best place to look for isolation. The USB is just too broad a pipe, logically and electrically, to serve as much of a point of isolation. Its undoubtedly an improvement over a circuit board in the computer chassis, but that's about it.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

It's jitter, posted on June 12, 2011 at 18:09:19
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 Read these papers to get educated:

http://www.positive-feedback.com/Issue43/jitter.htm

http://www.positive-feedback.com/Issue22/nugent.htm

Jitter affects the D/A conversion, creating distortion.

 

Those clowns are responsible for a lot of horrible sound, posted on June 12, 2011 at 19:09:13
Charles Hansen
Manufacturer

Posts: 6984
Joined: August 1, 2001
 On the one hand TI is the only chip company to make an all-in-one USB data decoder that requires no programming. On the other hand those guys are complete morons when it comes to clocks, clock recovery, jitter, or anything like that.

TI has tried *three times* to make an S/PDIF reciever chip. The first was so horrible that it never made it to market. The second two both relied on the "SPACT" technology mentioned in the linked article. That technology is so bad that both S/PDIF receivers using it were discontinued. I think the first one was so bad that it was actually recalled...

Then they had a group that made the (programmable) TAS1020B USB receiver chip. That one worked really well, assuming you had a really smart programmer to write code for it.

The bottom line is that the writer does NOT know his stuff when it comes to clocks and jitter. Otherwise it wouldn't have taken years to get that chip to market and it would have worked far better than it does (~5000 psec of jitter).

 

RE: How can USB performance impact audio quality?, posted on June 12, 2011 at 20:02:54
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
 This has also bothered me for some time, especially when you have an asynchronous interface which is supposed to prevent such things from happening.

After a LOT of investigation and experimentation I now think I have a somewhat decent handle on it, as Tony mentioned it all comes down to good old fashioned analog noise and in particular the spectrum of said noise and how this changes with whats going on in the computer etc.

I and others have built DACs with exquisite high quality power supplies etc which is supposed to prevent any coupling through the PS. BUT this ignores the ground plane. After you get everything else cleaned up you have to deal with ground noise, and this is particularly difficult to deal with.

As to how this is important I'll cover a scenario with USB, I'm assuming this is already an extremely well done async interface.

A packet comes in over the cable and goes in the receiver chip. This chip does a lot of stuff with this packet and eventually puts some bits in a buffer that down stream circuits extract using its super stable ultra low jitter clock. But what does all that processing in the receiver do? It generates noise on both the positive supply rail AND the ground plane. Any good design is going to have separate voltage regulators that prevent the noise on the positive supply rail from reaching any other circuits (such as the ultra low jitter clock, DAC chip, analog stage etc). BUT those don't work for the ground. If you are not EXTREMELY careful, some of that ground noise from the receiver is going to be on the ground pins of the other parts. If you just use a common ground plane you are almost guaranteed to have this. No its not huge, but its still there.

The interesting aspect of this noise is that if you look at the spectrum of it you see a strong 1KHz component, which is the packet rate on the USB bus. This falls right smack dab in the middle of the audio range. It gets even worse. In most circumstances the exact time at which packets are sent is controlled by software, NOT a low jitter hardware clock. This means that things like kernel scheduling policies, interrupt latencies etc have a HUGE impact on the exact timing of those packets. It comes out to close to 1KHz all the time, but there is significant variations on the exact time between packets. This causes that 1KHz component of the ground spectrum to spread out, it has "skirts". This "modulation" is in the low end of the audio range. There is a fair amount of evidence that this type of low frequency jitter can make subtle changes in the listening experience even when it is fairly low level.

Unfortunately this can make USB DACs quite susceptible to all kinds of things that happen on computers. The player software used, what audio stack is used, kernel parameters, timing on memory chips, all kinds of things can cause subtle changes in the packet timing on the bus.

There IS a way to cut down on this, use separate ground planes so noise on the receiver does not get into the sensitive area. This is a lot easier said than done. One way is to completely isolate things, but then you need something like an optoisolator or transformer for the data. None of this is easy and usually adds significant amounts of jitter.

The other way is to make sure there is one and only one connection between ground planes (which is right underneath the data lines). This actually does work and does not require isolating the data, but it DOES require one and only one ground connection. This means it takes separate power supplies.

To REALLY do it right requires multiple domains each with its own separate power supply and very careful ground plane management. Unfortunately this is neither easy or cheap. It takes a lot of work to design this and it takes a lot of money. But when done right the results are spectacular. But it means that any DAC done this way is going to cost a lot of money, there is just no way around that. And that's not even using boutique components.

The fun part is to try and come up with a design that tries to do the above but doesn't actually fully implement it. There are a LOT of different compromises that can be done for various different price points. This is where I think the biggest challenge is going to be, how to do it well (not perfect) without costing a fortune.

So the state of affairs right now is you can spend a lot of money and get something that is fairly immune to whats going on in the computer (not completely, there are some third order affects that are REALLY tough to get rid of), or pay a small amount of money and really have to tweak the computer to get really good sound. Unfortunately there is no cheap and always sounds great, it doesn't exist.

There is a lot more to say about this subject, but this is already way too long so I'll stop now.

John S.

 

RE: How can USB performance impact audio quality?, posted on June 12, 2011 at 20:53:24
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 It should be possible, given the right clock architecture, to do this at a cost which is still less than the absurdly complex (and ineffective) band aids people try to apply inside their computers, software and/or hardware kluges or the cost of boutique transports and cabling.

Your comment about separate power supplies is on the mark. It may actually require more than two if it turns out that multiple stages of isolation are needed for best sonic results. However, with the right interface there is nothing particularly complex about the isolation interfaces, no protocols, just buffers and registers. The various stages of isolation can, and should be, within one clock domain shared with the DAC. The async USB receiver exists in two clock domains and has the necessary synchronizers.

This needs to be approached as an engineering problem. One wants to specify how many dB any (still bit perfect) variation on the input signal must be attenuated before it exits the DAC through the analog output. A good start would be to require that the difference between all recognizable eye patterns be attenuated to the point where all audio signals are affected by less than, perhaps -140 dBfs. (This does not mean that the output result has to be perfect to 24 bits, 1 psec jitter, etc., just that the differences in output caused by different input waveforms are attenuated this much.) This can be accomplished by multiple stages of attenuation, at the cost of very little logic per stage and the greater cost of separate power supply and packing for each stage. It seems likely that more than one stage of isolation may be required to reach this level of isolation. However, it is clear that each stage will provide some isolation, so there is no magic involved in adding stages as needed.

It should be pointed out that with multiple stages of isolation as one moves down the pipeline closer to the actual converters the quality of the components involved has to improve, i.e. all of the parts in the final circuitry, clock included, have to be audiophile grade. This is not necessary upstream after some stages of isolation. So the cost of going multiple stages may not be as great as it appears at first glance.







Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: How can USB performance impact audio quality?, posted on June 12, 2011 at 22:07:05
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 Basically, it is a hijacked and shared interface which contains the problems described below. Because of this, it is the least robust of the audio interfaces.

I have drawn attention to this for quite a while, particularly the sensitivity of async usb sound to cabling and power supply.

However, some of the inmates who are now admitting the the drwabacks have been strong proponents of usb over Firerwire and spdif done well simply because they make such dacs. For me, they should attempt to tackle the issues and not try to discredit doubters.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 12, 2011 at 23:50:40
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 Clowns? If you choose to use a common, cheap, sometimes universal interface for audio, you'll need to put in software and hardware efforts.

You said it was too troublesome to make a good spdif interfac e. Here is trouble .....

 

RE: It's jitter, posted on June 12, 2011 at 23:55:15
mr.bear
Audiophile

Posts: 4167
Joined: November 13, 2001
 Those were very credible, intelligent articles about jitter- thanks. I see that all digital source chains have the capability of generating a sonic signature based on the way they manage jitter. Quantative data regarding jitter however appears not to be standardized, not well correlated to audible effects, and is not made available to consumers of hi-fi gear. The arguments hence all seem a bit fuzzy.

The good news for computer audio is that there are "low-jitter USB interfaces available now that compete with the best CD playback devices," according to Mr. Nugent. The upshot is we're stuck with a certain amount of jitter coming from USB sources. The effects of that jitter can be managed well enough to be inaudible (or at least tolerable vs CD's) on all systems except those with the most extreme resolving-power. That may be a good enough answer (until the next playback system comes of age).

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 13, 2011 at 05:24:50
Ryelands
Audiophile

Posts: 1868
Location: Scotland
Joined: January 9, 2009
 the writer does NOT know his stuff when it comes to clocks and jitter.

In the context of PC audio, the OP noted that the design of USB cables reportedly affected the sound quality of the output and asked "what could explain that phenomenon?". It's a fair question.

Explicitly noting that its author was "not talking about high-end audio", I pointed him to an article written by a chip designer that explained in outline how USB audio worked and what the pitfalls were. Steve pointed him to some of his own material in similar vein. The perceived merits of the chip in question were irrelevant - I didn't suggest that the Burr-Brown devices were high end. But neither did the author:

We thought that there would be many applications for a one-chip USB interface and DAC if it could be provided at a low cost.
Sadly, apart from one of John Swenson's rare but always thoughtful posts, most of us could have written the others in advance and off the top of our heads. You know the drill: either scholarly but airy-fairy abstractions or, frankly, spiteful suggestions that you and colleagues act in bad faith.

I don't think these are likely to help the OP very much.

 

Wonderful explanation., posted on June 13, 2011 at 08:18:32
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 The new Wavelength Crimson HS I have has USB High Speed opto isolated ground between computer and DAC. Transformer outputs as well.

 

Even better than CD transports, posted on June 13, 2011 at 11:04:08
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 I retired my heavily modded (Superclock and other mods) Transport years ago. The jitter from some Computer Audio using USB is lower. These are my white-papers BTW.

 

What?, posted on June 13, 2011 at 11:06:36
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 Waht do you mean "tackle the issues"? I think we are doing this already. Proof is the great reviews and Golden-Ear awards etc...

 

RE: What?-How much?, posted on June 13, 2011 at 11:14:30
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 does your golden ear award unit cost?

 

RE: How can USB performance impact audio quality?, posted on June 13, 2011 at 11:17:26
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 Tony - how many of these clocks and USB interfaces have you designed?

I am evaluating an expensive NEL OCXO with really low jitter specs. Driven from a really high quality regulator, it still has 80-100psec of P-P jitter as measured by my 7GHz Tek scope. It is unrealistic to expect to get 1psec of jitter from any system IMO, much less a clock. It is difficult to achieve speced jitter on a clock in an actual system.

Isolation stages all add jitter BTW, and are useful only in the non-jitter sensitive paths. The only way to keep the jitter really low is to keep the design extremely simple with ultra-short paths.

Individual regulators, and high-quality ones certainly help. I use 7 in my USB converter.

Steve N.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 13, 2011 at 11:22:52
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 "In the context of PC audio, the OP noted that the design of USB cables reportedly affected the sound quality of the output and asked "what could explain that phenomenon?". It's a fair question."

It's obvious. Losses on the cable and the resulting affect on risetimes is detrimental to the probability of detecting a transition at the receiver. If the receiver has noisy voltage (they all do), then the switching threshold is changing in time. If the risetimes are slowed this can therfore result in more jitter.

 

RE: What?-How much?, posted on June 13, 2011 at 11:24:43
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 The latest version is $799.00 Good clocks and good regulators don't come cheap.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 13, 2011 at 11:57:14
Ryelands
Audiophile

Posts: 1868
Location: Scotland
Joined: January 9, 2009
 It's obvious.

I'm sure it is to you but not necessarily to everyone else.

Not to the OP (who seems to have round the block), it wasn't. Not unless you appreciate that USB handles audio unlike other data (i.e. in real time), it isn't. Not to many programmers (who really ought to know better), it isn't. And so on.

I for one remain convinced that "high-end" USB cables typically offer a questionable solution to a problem better addressed, in the first instance at least, by other means.

Most people don't know, for example, that simply configuring a computer's USB ports with a little care can make a marked difference. Until you do that, money spent on fancy cables is, in all probability, unwisely spent.

 

RE: How can USB performance impact audio quality?, posted on June 13, 2011 at 12:31:04
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 You didn't get the distinction I made between jitter coupled from the input signal and jitter in the local oscillator. This distinction is the whole point of the asynchronous USB clock architecture and other clock architectures such as AES/EBU where the DAC sends the clock to the transport. Once one has this architecture it becomes possible to work separately at minimizing these two causes of jitter. With this clock architecture there is no physical necessity for any jitter coupling from the input signal, but there still is physical necessity for jitter in the local clock and the associated clock circuitry up to and including the actual switches that comprise the DAC. If one fails to make this distinction one might end up using the inherent physical limitations of oscillators as an excuse for failing to engineer sufficient isolation from the input signal, but this will be an excuse not an explanation.

Even if the local clock has significant jitter, it will still be possible to measure (and presumably hear) the effect of coupled jitter in the input signal. Measurement can be done, for example, by deliberately injecting jitter that is synchronized to a signal and then measuring the jitter sidebands in the output of the DAC using synchronous detection.

While I have no doubt that poorly designed or implemented isolation stages may fail to isolate jitter, they certainly do not add jitter in a proper clock architecture. (With improper clock architectures this will not be the case. Thus if clock is fed forward from the source and jitter suppressed using phase locked loops or digital variations thereof, there will be eventual buildup of jitter and unstable system operation. This is a familiar problem in communications systems, for example I have worked on system designs that ran into there problems using both analog PLLs and digital PLLs when these were used in series. Even with systems where signal is reclocked at each stage, such as in Ethernet repeaters and the FDDI token ring, there is a build up of timing errors which result in system failure when too many stages of reclocking are used. This can be avoided by the appropriate clock topology.)

I agree with the emphasis on simplicity. If one must start with the USB and all of its baroque complexity (in the classic Intel engineering philosophy) the complex logic must be logically, electrically, and physically isolated from the sensitive analog and mixed signal electronics that are the heart of any DAC. Furthermore, the higher the quality of the DAC the less opportunity its own jitter will have to mask any coupled jitter, and hence the necessity for greater jitter isolation.

When measuring jitter, it's the jitter spectrum that matters not the total jitter energy. Specifically it's the jitter spectrum in and near the audible band that is going to produce sidebands that will be perceived as audible distortion. Jitter at RF frequencies may be important in telecommunications applications and jitter at minute fractions of a Hz may be important in instrumentation applications, but neither of these is very important in audio applications. A single number that comprises jitter is not relevant unless it's very low, thereby ensuring that any possible spectrum with this jitter power will produce inaudible distortion, but this mandates a signal number that is below 10 psec, possibly even below 1 psec and hence probably not obtainable.


Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

I couldn't agree more. nt, posted on June 13, 2011 at 12:46:51
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: How can USB performance impact audio quality?, posted on June 13, 2011 at 13:22:53
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 I just finsihed reading Art Dudley's comments on the Ayre QB-9 and 3 music programs: Amarra, Decibel, and Pure Musicin the July Stereophile.

Dudley's conclusions on the sound of the music programs were the opposite of mine. He found Amarra having the largest soundstage. I have found it to be smaller than that of Pure Music. How is it possible that our conclusions can be so different? His methodolgy was not well documented, so who knows if he ran hog in Pure Music or just Decibel.

I have also found that firewire is not any better than USB in many of these matters. In fact, USB seems to be offering higher sampling rates to 384KH in dacs like the Antelope Gold, and surpassing the specs of firewire dacs.

But in the end, there are many products being made that offer an enjoyable musical experience. The critical views expressed here should not deter people from enjoying USB dacs.

But as fmak commented, we should not be afraid to ask questions.


 

A few things..., posted on June 13, 2011 at 14:00:44
Gordon Rankin
Manufacturer

Posts: 2946
Joined: June 9, 2000
 Steve,

First you cannot measure audio related jitter with a 7G scope. Again for the 1 millionth time... the jitter we are looking for is in the low audio band like from 0.5Hz to 100Hz. These scopes would require terabytes of memory to look at this. You are only looking at jitter above 100KHz which is useless for audio. All that stuff gets averaged out and therefore is not worth looking at.

Second to get low jitter from an oscillator you need to throw away 3 terminal regulators. You need to start designing some discrete ones and have a spectrum analyzer which can test them down to the 1nVrms range below 100Hz. This will tell you more than your scope will. Like on my Crimson I have a battery supply then a local regulator and then a discrete regulator and at 1Hz the noise is 5nVrms and goes down from there.

Tony is correct in that the local oscillator will be effected by other aspects and oscillators being that they are X,Y or Z and non-related to the audio oscillator. The big problem on many of these extra oscillators is that a PLL is used to generate some high oscillator property.

High Speed USB does not work at 1KHz, instead it works at 125uS or 8000 times a second. Still in the audio band but would average out a lot easier.

But the real way to deal with USB is put these extra oscillators on a separate board with an optical barrier between them. Feed the audio oscillator into this board and to the DAC chips and therefore you don't have this contamination from the USB section because they are running on separate supplies and totally optically isolated.

But even then the USB cable still has an effect. Hey so does analog cables.

Thanks
Gordon
J. Gordon Rankin

 

RE: How can USB performance impact audio quality?, posted on June 13, 2011 at 14:04:31
b.l.zeebub
Audiophile

Posts: 9361
Location: 52deg 28'N,1deg56'W
Joined: April 17, 2006
 USB? Higher sampling rates?

My FW400 interface is 6 years old and always managed to handle 12 channels of 24/192 in&out simultaneously without any hiccups.
I bought it 2 years after I've seen the first FW interface doing 32/384 fully duplex.

 

RE: How can USB performance impact audio quality?, posted on June 13, 2011 at 14:14:46
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 I was discussing consumer dacs like the Weiss Dac 202 and Berkley dac.

But obviously, you are correct. The point I was trying to make, and quite poorly I might add, is that the line is blurring between the two formats and they are both subject to many of the same problems.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 13, 2011 at 16:42:17
rick_m
Audiophile

Posts: 6230
Location: Oregon
Joined: August 11, 2005
 "Most people don't know, for example, that simply configuring a computer's USB ports with a little care can make a marked difference."

Oh, it's so nice to be in the majority, it just feels great!

But let's say that in the privacy of my own home that I feel the urge to make a marked difference, what sort of arcane ritual would it entail?

I think I did try to find a port on a root node, what's next on say an XP system? Perhaps I could mark the best port with a dab of chicken blood?

Thanks, Rick

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 13, 2011 at 17:34:11
Ryelands
Audiophile

Posts: 1868
Location: Scotland
Joined: January 9, 2009
 Oh, it's so nice to be in the majority, it just feels great!

I wouldn't know - my hunch is that most of humankind have not been confined to the PC-Audio Asylum even if a fair few do now use computers.

What sort of arcane ritual would it entail?

Looking round the ward (aka an archive search), you'll find posts on the topic over the years from folk such as Presto, John Swenson and Gordon Rankin. Fmak even (useful, too - well, from time to time). For my take, see link.

PS: What does "arcane" mean?

Perhaps I could mark the best port with a dab of chicken blood?

Messy. But, if recreational killing is your bag, . . .

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 13, 2011 at 20:35:14
rick_m
Audiophile

Posts: 6230
Location: Oregon
Joined: August 11, 2005
 Thanks for the 'your take' link, didn't know that info was in there. For a minute I thought you were going to force me to do some actual work to search for the data. Being a man of our age I try and keep modern and thus I want instant gratification and I want it now!

If you really don't know what "arcane" means I would suggest looking it up in your Funk and Wagnall. But since you thoughtfully provided a link, so will I...

The funny thing is that our family dictionary really was a Funk and Wagnall's while all my friends had Webster's, probably accounts for a lot!

Rick


 

RE: How can USB performance impact audio quality?, posted on June 13, 2011 at 21:52:06
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 OK, but why put consumers in quandries by using interfaces that are not audio friendly and are troublesome in use, whilst claiming 'overwhelming' jitterless performance?

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 00:29:04
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 'But let's say that in the privacy of my own home that I feel the urge to make a marked difference, what sort of arcane ritual would it entail?'

You need to learn Intel speak and have patience. I don't actually think that changing USB polling is a solution, or lead to that much improvement.

Disabling USB ports is, to me, a ridiculous solution just because the usb audio makers have hijacked a connection meant for other things.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 00:31:28
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 The link illustrates the disregard of punters for possible website malware warnings. No, I don't want to disable my screen filter just to look.

 

RE: Even better than CD transports, posted on June 14, 2011 at 00:33:57
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 How do you judge what is better?

 

RE: No, this thread has been tremendously revealing!, posted on June 14, 2011 at 00:38:04
mr.bear
Audiophile

Posts: 4167
Joined: November 13, 2001
 I asked what I thought were a couple very basic questions that some savvy digital guy like you would answer in 25 words or less. What instead resulted was substantial, basic conflict amongst divergent schools of thought regarding USB audio, plus widespread lamenting about how lousy it is.

What I was able to discern from this is that there are numerous ways to transmit, isolate and condition the USB signal, and to treat noise ahead of and inside the D/A. Excellent treatment takes a lot of expensive circuitry. Each different design approach is accompanied by tradeoffs affecting musical quality and ultimately there's no clearly dominant design. So, for the present, selecting a DAC comes down to matters of personal taste and prejudice, much like phono cartridges. That's a little disheartening-- I wasted a lot of money trying different cartridges in the 70's.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 01:57:19
Dawnrazor
Audiophile

Posts: 12729
Location: N. California
Joined: April 9, 2004
 Gee Fred you really dont know what you are talking about. Malware that is targeted at computers that have no internet connection??

Here is the link sans the picture and no "malware":


Optimising USB DACs

The plethora of USB DACs now on the market ranges from high-quality but expensive devices to others providing remarkably good sound for modest prices. Most low-end types are non-oversampling, scantily documented and lacking dedicated drivers though they usually work with ASIO4ALL.

A serious drawback of USB technology is that it is often poorly implemented on motherboards. Where designers (charitably) must keep prices to consumer levels, they reportedly skimp on critical design parameters. Intel also notes that poor-quality power supplies degrade USB performance. It is generally systems laden with devices running at high speeds and/or drawing power from the motherboard that encounter the most severe problems. That scenario may not pertain to PC audio but it does illustrate common problems that do. Some simple steps can help to minmise their impact on sound quality, particularly on a dedicated audio PC. Try some or all of the following:

1. Connect the DAC to a back-panel socket, not to the motherboard’s on-board USB headers;
2. Provide the DAC with its own power supply – do not power it off the motherboard;
3. Some after-market USB cables make a measurable difference. However, devices using Analog Devices’s ADuM4160 isolator such as that sold by Circuits@Home significantly improve the sound quality of most USB DACs and are usually better value for money. For best results, provide it with a high-quality 5-volt power supply, by-pass the on-board voltage regulator and connect it to the DAC with an adapter as shown rather than a cable. (The device has been successfully tested at 96kHz. Though issues have been reported with some DACs at rates over 48kHz, they are almost certainly caused by the drivers.)
4. Disconnect all other USB devices when playing music;
5. Disable USB 2 (and 3 if available) in BIOS (enable USB 1 only: USB 2/3 can be re-enabled when needed);
6. Disable USB power management. Select Control Panel > System > Hardware > Device Manager > Universal Serial Bus Controllers and expand (click on +). Right-click each Root Hub in turn and deselect [Power Management] > Allow the computer to turn off this device to save power;
7. Disable unused hubs. Select every Root Hub as above and find the DAC’s (right-click, Properties > Power). Disable the rest (perhaps leaving one or two for housekeeping): right-click and select Disable (not Uninstall as XP will reinstall the hub at the next restart);
8. Remove ‘non-present’ devices. XP notes devices hitherto connected to the USB. To clear the record, launch a command prompt, type set devmgr_show_nonpresent_devices=1 [Enter], then devmgmt.msc [Enter]. Leave the window open. In Device Manager, click View > Show hidden devices and uninstall all USB devices (both ‘ghost’ and current) and reboot with the DAC on. The step may have no effect but can resolve obscure driver issues;
9. Check the DAC has a unique IRQ (see Step 6 above);
10. Check PCI latency timing. The values determine how many clock cycles elapse between a device taking and relinquishing control of the PCI bus. Some manufacturers set defaults too high so that their devices deny prompt access to others. The shareware ‘PCI Latency Tool’ can view and possibly adjust them (depends on mobo chipset): if some are unduly high, try reducing them to 64 or 48. Note that crudely setting the DAC high and all else low is counter-productive – the idea is to enable smooth access by preventing any device from bus hogging;

It is not claimed that the above steps will transform the quality of an entry-level DAC but, taken together, they make for more reliable performance and almost always for better sound as well. The AD isolator in particular (step 3) is a significant upgrade and excellent value for money.

Cut to razor sounding violins

 

RE: How can USB performance impact audio quality?, posted on June 14, 2011 at 02:23:50
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 I guess I'm not as sophisticated as you are fmak. My USB dac works well and is enjoyable to use.

 

RE: No, this thread has been tremendously revealing!, posted on June 14, 2011 at 02:44:23
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 "So, for the present, selecting a DAC comes down to matters of personal taste and prejudice, much like phono cartridges."

Just like every other audio decision you make when you are building a system. This is no different.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 02:50:56
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 And you do?

The steps you mentioned are not the characteristics of a universal interface. If you know so much, why do you have the probelms with usb that you posted just now.

A little knowledge is always a dangerous thing ....

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 03:06:01
Dawnrazor
Audiophile

Posts: 12729
Location: N. California
Joined: April 9, 2004
 Poor Fred,

I was talking about the site that Ryelands linked.

You were scared to go there because you thought it was malicious. THAT is what you are clueless about.

I kindly posted the material on the site so you could see it. I dont use USB for serious audio.

And you probably missed it but the problem I have with usb is when I am ripping. And it is a windows problem more than a usb problem. Since I can rip simultaneously with 4 usb drives and zero problems besides that one issue (which is easy to deal with by the way) USB at least for ripping is turning out to be just as good if not better than the ide buss. THAT doesnt work at all from the mobo...but does from the pci-e expansion card.

You were a huge help by the way to solve those ripping issues. Great advice fred. Which specific IDE or USB or SATA drives do you insist will solve my ripping woes?? I would love to see your expert recommendation.

But let me play prophet here. You wont have one. Somehow you seem to bow out when it actually comes to helping or exchanging anything useful. This time will be no different.

Cut to razor sounding violins

 

RE: You are posting w/o rationale..., posted on June 14, 2011 at 03:41:34
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 as I said, first comprehend, then discriminate, then post what you think in response to what is said.

Why should I, or anyone else visit a site that for months has been flagged up as unsafe? You believe in spreading malware?

 

RE: How can USB performance impact audio quality?, posted on June 14, 2011 at 03:44:04
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 You haven't answered my question. As you said, you have found PM to be better in YOUR system, which is a little out of the ordinary. Others have found differently, and you shouldn't need to be defensive about that.

I am in the same boat.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 03:47:34
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 And oh, these instructions constitute a Clown's solution to making a weak audio interface work.

But then I don't expect you to comprehend what I have said.

 

Yep. Score one for the prophet., posted on June 14, 2011 at 03:59:38
Dawnrazor
Audiophile

Posts: 12729
Location: N. California
Joined: April 9, 2004
 You never cease to disappoint.

No actual advice from Fmak. NONE.

At least you provide a funny post.

Somehow you missed the part about me posting the site here without all that "malware". Yet talk about discerning and comprehending.

I cant wait for the next act!




Cut to razor sounding violins

 

RE: How can USB performance impact audio quality?, posted on June 14, 2011 at 04:05:13
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 I remember the designer from dCS stated that they selected USB as it was a common interface and it worked well. I don't have an answer to your question as I am satisfied with USB..

I'm not technically trained to really answer your question in depth.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 04:23:27
Ryelands
Audiophile

Posts: 1868
Location: Scotland
Joined: January 9, 2009
 Malware that is targeted at computers that have no internet connection??

I think you miss fmak's point - I assume he had some sort of malware warning and doesn't want to visit the site. His post is an example of the weakness some punters have for passing on false malware warnings.

But thanks for posting the notes on USB, most of which came from other list members. To one of the points:

Most low-end types are non-oversampling, scantily documented and lack dedicated drivers though they usually work with ASIO4ALL.

I'd now add something like: "but the AQVOX (Ploytek) generic USB driver, though not cheap, gives much better results".

I'd also draw attention to "Step 7 - Reduce USB Polling" elsewhere on the same site which I and others found to be an effective "tweak".

The points (as is obvious from the site) aim at getting the best from a low-end DAC driven by a computer used only for audio replay. I don't know to what extent, if any, they might apply to up-market kit with bespoke drivers though I understand that Wavelength Audio endorses the ADuM4160 where appropriate.

Most of them can easily be used to good effect on general-purpose computers though dedicating a hub to a DAC can be fiddly. Some users report good results from connecting via a dedicated PCI or PCIe USB card. I've not tried it.

** I see that fmak has just replied to your post to say:

And oh, these instructions constitute a Clown's solution to making a weak audio interface work.

Ignoring the tiresome abuse, that's exactly what they are. With a little care, USB can be made to work rather well for audiophiles on a budget. I don't know why he is so resentful of this.

(If the likes of Mercman are to be believed, it works pretty well up market as well though I sometimes wonder if he reports only what his dog tells him to . . . )

But then I don't expect you to comprehend what I have said.

You're right - I don't.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 05:01:18
theob
Audiophile

Posts: 3180
Location: ann arbor michigan
Joined: November 4, 2000
 You said:

...the likes of Mercman are to be believed, it works pretty well up market as well though I sometimes wonder if he reports only what his dog tells him to . . .

I resent your bias against listening to one's dog....Coco told to say that.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 06:53:30
rick_m
Audiophile

Posts: 6230
Location: Oregon
Joined: August 11, 2005
 "Disabling USB ports is, to me, a ridiculous solution just because the usb audio makers have hijacked a connection meant for other things."

Yea, I know what you mean. On the other hand using a GP computer to play music seems like the epitome of that to me and yet I find both the concept and results pleasing. The amazing thing is that it can do it at all let alone so well.

I'm comfortable with making adjustments and selecting player software to achieve good sound with internal audio and it usually causes little or no impairment to 'normal' usage so I hope for something along the same line from tweaking the USB system.

Regards, Rick

 

By George! I think you've got it! , posted on June 14, 2011 at 07:12:49
rick_m
Audiophile

Posts: 6230
Location: Oregon
Joined: August 11, 2005
 "Each different design approach is accompanied by tradeoffs affecting musical quality and ultimately there's no clearly dominant design."

Wow! That is the best description of home audio I've ever read.

It's also true of almost every other field that's tightly coupled to our senses and cognition. It describes binoculars, hearing aids, cameras, scents, even cars, mates and houses.

And political systems, lawnmowers and fireplugs. Even systems of measurement and money.

There is only one common denominator...

Regards, Rick

 

RE: You are posting w/o rationale..., posted on June 14, 2011 at 07:32:15
Ryelands
Audiophile

Posts: 1868
Location: Scotland
Joined: January 9, 2009
 Why should I, or anyone else visit a site that for months has been flagged up as unsafe? You believe in spreading malware?

Just for the record, there is no "malware" on the web site in question - you need to look to your security software. Please stop spreading mischievous rumours.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 08:02:32
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002



Why do you think Lucy is the Musical Director?

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 14, 2011 at 08:05:56
Dawnrazor
Audiophile

Posts: 12729
Location: N. California
Joined: April 9, 2004
 Hey Dave,

I knew what Fred was alluding to. But I think he misses where the malware on that site supposedly is, and why he is getting the false positives. It is not from visiting the site.

Anyhow no good deed.



Cut to razor sounding violins

 

You have..., posted on June 14, 2011 at 08:16:18
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 may be up to 12 usb ports on a general use PC. Do you want to disable 11 and change polling for the lot?

And, do you want to remember enabling them for use when you want to?

 

You are providing a guarantee?, posted on June 14, 2011 at 08:18:06
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 It is up to web providers to conform with rewuirement. Otherwise you can complain to Microsoft.

 

RE: You have..., posted on June 14, 2011 at 09:10:21
rick_m
Audiophile

Posts: 6230
Location: Oregon
Joined: August 11, 2005
 "Do you want to disable 11 and change polling for the lot?"

Beats me, it's all TBD.

Isn't this in the same vein of not listening to expensive hardware because you might decide you like it but really don't want to shell out the money?

I think the first step is finding out what sort of differences things make then one can decide their value vs what they cost in dollars, time or hassle. Right now I'm just kicking the tires out of curiosity, maybe I'll learn something...

Rick

 

I have never had the opportunity to compare USB2 with FW using the same source & dac., posted on June 14, 2011 at 09:50:29
Norm
Reviewer

Posts: 31024
Joined: September 6, 2000
 But I have never heard any USB2 connection with any dac that reaches the level of what I hear with FW into the Weiss Dac202, from whatever source, using either FW, SP Dif, or toslink. Now with a computer server based on Macs, I wish I had a either a SP Dif source output or either a FW or USB input on my dac.

 

A few more things..., posted on June 14, 2011 at 10:26:29
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 "First you cannot measure audio related jitter with a 7G scope."

I said P-P total jitter, not audioband jitter. This is what is speced in the specsheet. Just because it specs 10psec, in a real system it can easily be 100psec.

"Second to get low jitter from an oscillator you need to throw away 3 terminal regulators."

I already use discrete regulators, only the best.

Regarding optical isolation, these conversions just add jitter. I get great results without it. We will see how much jitter my products have when JA makes his measurements. I've been doing this a while too.

Steve N.

 

RE: How can USB performance impact audio quality?, posted on June 14, 2011 at 10:33:11
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 "You didn't get the distinction I made between jitter coupled from the input signal and jitter in the local oscillator."

I do understand. I've been doing this for more than 30 years. It is possible to have jitter in the input USB signal coupled by power supply modulation in a shared supply, or by RF coupling or by ground bounce coupling. All indirect effects.

"While I have no doubt that poorly designed or implemented isolation stages may fail to isolate jitter, they certainly do not add jitter in a proper clock architecture.

I would agree, with the caveat that the isolation take place prior to the master clock domain.

"When measuring jitter, it's the jitter spectrum that matters not the total jitter energy. Specifically it's the jitter spectrum in and near the audible band that is going to produce sidebands that will be perceived as audible distortion."

I know this also. I have written a few white-papers on the web that echo this, most about 4-5 years ago. They are still on positive-feedback.com.

Steve N.

 

RE: Even better than CD transports, posted on June 14, 2011 at 10:36:31
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 Mostly by using the tool that is used in the end-product, the human ear.

I know your game. You are trying to discredit me for not using 100% measurements to gauge quality. Well it doesn't work. It doesnt work with amplifiers or CD players or digital audio either. I trust my ears. They have not failed me yet. It's not by chance that my room at RMAF gets best of show.

 

RE: How can USB performance impact audio quality?, posted on June 14, 2011 at 10:41:21
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 Well said.

I personally have found Amarra to be better in my system, however only certain versions run in certain ways. The EQ and volume control are top-notch, and worth it for this alone IMO.

Steve N.

 

RE: You have..., posted on June 14, 2011 at 22:01:01
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 Try it, it's a hassle.

Changing the Polling frequency just changes the way the signals intermodulate. Will that be an improvement for every system? Not likely.

It can also make your usb devices less responsive.

 

RE: You have..., posted on June 15, 2011 at 01:44:05
Ryelands
Audiophile

Posts: 1868
Location: Scotland
Joined: January 9, 2009
 Changing the Polling frequency just changes the way the signals intermodulate.

And your evidence for the "just" bit is?

I ask because the link suggests that, despite being confused about what was involved (you muddled it with the very different gamers' trick of increasing HID polling rates), you supported the suggestion.

Note that the RegEdit change discussed lowers the rate at which USB ports are polled from 1KHz to 200Hz and is applied per controller. I see I suggested that:
. . . the default MS polling interval is one millisecond, i.e. its frequency is 1 KHz, bang in the middle of the mid-range. If the activity, esp with consumer-level DACs, exacerbates jitter even slightly, the effect of moving it to the less-sensitive 200 Hz region might well improve the sound.
It can also make your usb devices less responsive.

Sure. Drives take forever to copy, scanning slows to a crawl, the sky goes dark, cows starting herding beneath trees and there's an eerie silence.

Solution? Don't perform the change on controllers driving HDDs, scanners - or cows.

But I'd wager that neither you nor anyone else has ever spotted a performance difference. And were Rick do so, he might just have the gumption to reverse the change.

You're making this up as you go along.

 

You do not comprehend nor have an understanding, posted on June 15, 2011 at 04:02:24
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 what I posted

nor

do you know how intermodulation works, whether at 200 Hz or 1 KHz.

It's better to keep quiet if things are beyond you.

 

RE: You do not comprehend nor have an understanding, posted on June 15, 2011 at 04:08:51
Ryelands
Audiophile

Posts: 1868
Location: Scotland
Joined: January 9, 2009
 You do not comprehend nor have an understanding . . . what I posted

That's true - but neither did anyone else.



It's better to keep quiet if things are beyond you.

Good Advice.

 

RE: No, this thread has been tremendously revealing!, posted on June 15, 2011 at 13:10:40
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
 Most decent DAC's sound very similar when driven from a low-jitter digital source. See TAS review of iDAC in the current issue.

 

Comparisons with other interfaces, posted on June 15, 2011 at 15:04:24
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
 After reading what has been said on this thread about USB and other interfaces I'd like to make some comparisons based on the same issues I brought up with USB.

First off some definitions, note these are mine, not industry standards:

First order effects: these are jitter and noise from PLLs, frequency sysnthesizers, noisy power supplies etc. They have strong impact on final sound.

Second order effects: Noise on power and ground traces on a board due to devices running on that board and connected to it through interfaces. Note I'm making a distingction here between noise of the power supply itself and noise on board traces due to devices running off the supply.

In many boxes the first order effects are much greater than the second order ones, almost completely swamping them.

So, on to the comparisons. In my first post I was focusing on USB devices that had almost completely eliminated the First order effects, in this case the second order effects are now clearly heard. Its not that things such as asynchronous interfaces are implemented poorly, quite the contrary, they have done their job well and almost completely eliminated the first order effects. Now you can clearly hear the second order effects.

Now compare this to other interfaces. Firewire, ethernet, PCIe, PCI, ALL have very similar second order effects, why are people not screaming about them?

Most firewire implementations (not all, but most) still have very high levels of first order effects. One very popular FW interfece has 700ps of jitter on its local clock, they are even proud of that since their competitor has 1200ps! The second order effects due to timing variations on the bus are completely swamped by the first order effects so you don't notice the dependancies on computer timing.

Ethernet implementations are split, some still have fairly high first order effects and some have quite low first order effects. BUT ethernet receivers are usually implemented with a general purpose processor running linux or some other OS which has a lot more going on than JUST decoding audio ethernet data. The result is quite a bit of second order noise. The contribution of timing variations on the bus is still there but is being swamped by the noise generated from all the other processing going on.

PCIe and PCI sound cards are also affected by second order timing on their busses, but they are so tightly coupled to the noise from the computer itself that they are again being swamped by that noise.

From this analysis it seems to me that far from being the worst as has been postulated by several posts on this thread, USB is now the best of the computer interfaces. The engineers have worked hard and have almost completely gotten rid of the first order effects, and significantly decreased the second order effects. This is bad?

The situation HAS brought to light that things are still not perfect, that there is more work to be done in decreasing the second order effects. But even so one of these devices when plugged into a computer that has not been optimized is still MUCH better than devices that still have significant amounts of first order effects. Yes you can tweak the computer and get even better sound, but how does that mean USB is inherantly BAD? It just means that we are in an intermediate stage of development.

Until designers had gotten rid of the first order effects the impact of the second order effects were not taken into account, now they are getting looked at, some designers are working on ways to deal with them. Others are working on ways to tweak the computer to decrease these second order effects with existing equipment.

Eventually the designers will succeed and these computer tweaks will be unnecessary, until we find out about the third order effects, whatever THEY may be!

John S.

 

Another excellent post, posted on June 15, 2011 at 18:22:40
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 Thank you John.

 

+1 (nt), posted on June 15, 2011 at 19:55:34
Charles Hansen
Manufacturer

Posts: 6984
Joined: August 1, 2001
 nt

 

question is:, posted on June 16, 2011 at 00:31:21
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 'In many boxes the first order effects are much greater than the second order ones, almost completely swamping them.'

Are you comparing usb, firewire and others boxes with what you call 1st order effects largely eliminated?

And; are usb cable effects 1st or 2nd order when they have such an influence on the sound quality of triple isolated async usb devices?

There is an interesting article with measurements by Paul Miller in the June issue of HiFi News on the Musical Fidelity CLic in which first order effects appear constant. Comaparing usb, spdif and wired interfaces, usb has the higher distortion and jitter plots over the others (academic since we are looking at very low levels).

Frankly, for less than really serious listening, a box such as the £1250 MF CLiC would seem the much easier and better solution than countless software and hardware mods on a computer replay system.

This would not be my personal choice but for some it seems sensible.

Mercman, have a look at your new 'opto coupled ground?' here:

http://www.inalp.com/technotes/about_optical_isolation.pdf

 

RE: A few things...you are right, posted on June 16, 2011 at 07:16:08
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 There are 2 articles in HiFi News June issue on the wow equivalent of digital jitter at <2 Hz.

I also agree that there no really good 3 or 5 terminal regulators.

 

RE: question is:, posted on June 16, 2011 at 07:26:58
rick_m
Audiophile

Posts: 6230
Location: Oregon
Joined: August 11, 2005
 That's an interesting link Fred, I have seldom seen so much cluelessness in one place. It's not ALL wrong however. What did you find of value there?

Rick

 

RE: question is:, posted on June 16, 2011 at 08:32:43
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 Googled. Meant to show that opto coupling isn't the final answer.

I don't like some opto-coupled dacs.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 16, 2011 at 09:17:03
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 Which specific URL are you worried about? What anti-malware program are you using?

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: question is:, posted on June 16, 2011 at 09:42:36
Mercman
Audiophile

Posts: 6581
Location: So. CA
Joined: October 20, 2002
 Seems to work well for the Wavelength and Ayre prodcuts.

 

RE: A few things..., posted on June 16, 2011 at 09:48:05
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 "But even then the USB cable still has an effect. Hey so does analog cables."

Since the effect may not be measurable, I believe it is necessary to concoct some clever experiments to narrow in how the differences and cables may be coupling into the audio system. That would have to be the first step in solving these problems (or as John suggested, finishing with the second order problems so as to be able to move on to the third order problems).

I don't believe it will require expensive instrumentation to do these experiments, but after the problem has been isolated then it may require expensive instrumentation to be able to select circuits that minimize the unwanted coupling. But at least one will know where to work. (It may be that some know the answer here, and are keeping quiet for commercial reasons. If I were in that position I would just keep silent, because even to admit there was an answer would be to help the competition.)


There are some things that probably do not affect the sound. I'm not sure where the line is drawn, but I think it is safe to say that the Ethernet cable connecting the HDtracks.com download server to the nearby hub is not going to affect the sound quality experienced by any of their customers. :-)


Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 16, 2011 at 10:56:30
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 http://www.cicsmemoryplayer.com/index.php?n=CMP.10Soundcard#OptimisingUSBDACs

Just IE8 and Windows Firewall with Smart Screen filter engaged. Disengage that and its ok.

But this site has been like this for months and there has not been an attempt to clear it. Others have found the same.

This doesn't mean that I hadn't looked at it, but those who preach the faith should take it seriously, as one never knows.

 

RE: question is:, posted on June 16, 2011 at 13:06:15
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
 I've actually yet to find a FW interface where the first order effects are very low. I'm sure they exist somewhere, but I haven't found one yet. The Metric Halo units are probably the best candidates at this point, but I don't have one (they only work with Macs and I don't have Macs).

I have several ethernet interfaces with very low first order effects, but the second order effects due to proceesors etc are pretty high.

S/PDIF is an interesting case. Almost all implementations have significant first order effects (the PLL used to recover the clock). There IS a way to get around this, send a clock or S/PDIF stream from the dac to the source. In this case the data is synced to the DAC clock so it can be reclocked with the local clock, so you don't need to use the recovered clock. This does a good job of getting rid of first order effects, BUT they still use a PLL based receiver. The jitter from that receiver is still creating ground noise which can get into the rest of the system. Note that a pulse transformer on the input to the S/PDIF receiver makes no difference to this.

There is a way around this: don't use a PLL based receiver when you are syncing the source to the local clock, its unnecessary. I've actually done this using an FPGA to do the S/PDIF decoding, it works very well.

I've not done work on whats going on with cables and why they make the differences in modern DACs so I can't really answer that question. So far I've just been looking at how changes in the computer affect ground plane noise in the DAC.

In reading the posts I see I have not actually described my testing setup. I built a very low noise differential probe which I plug into a spectrum analyzer. The probe "tip" is a short length of shielded twisted pair. I directly solder the two wires to different locations on the ground plane under test. The probe electronics are battery powered and heavily shielded.

I then plug the DAC into different computers and try various different settings on the computers. I do this for several different locations on the ground plane and with different DACs.

John S.

 

RE: Those clowns are responsible for a lot of horrible sound, posted on June 16, 2011 at 13:35:15
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 I don't use Internet Explorer. I consider it to be unsafe. I don't use the M$ blacklist, either. It's not the fault ofcics that Microsoft has blacklisted the site.

In addition, I use other techniques to practice safe computing.



Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

Some random thoughts..., posted on June 16, 2011 at 14:55:35
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
 It may not be covered by the SPDIF specification, but I'll bet most SPDIF transmitters when synced to an incoming clock have a constant output frequency and phase (plus the usual jitter due to cables and electronic devices). This means that one can use synchronous detection if one can arrange the phase (strobe time). For short cables it is probably not even necessary to provide the necessary adjustment. In other words, there would be no need for an SPDIF decoder, not even one that didn't use a phase lock loop. Of course the data rate with this encoding is twice as high as would be needed with a simple NRZ code, such as can easily be used with the separate clock line. That would be even better as it would simplify the synchronization at the decoder. If I were going about to isolate a DAC from a computer, I would be inclined to a two box solution, the first box comprising the computer interface (e.g. Async USB) and connected to the second box with an IIS interface with clock coming from the DAC and data going to it. If variable phase delays were needed at this point they could be achieved by tuning the length of the cable between the two boxes or by building a variable delay into the converter box, again keeping the maximum amount of logic aware from the box that contains the clock, converter and analog circuitry. (This style I/O interface is hardly original and was developed by Seymour Cray in the 1960's for the I/O channels used in the CDC 6000 series computers.)

I am inclined to take this even a step further and remove all the DSP processing from the DAC box. This is certainly possible if one uses a 1 bit output at which point any required PCM conversion to bit stream could even be done in the computer system. With such a system there would be no digital logic in the DAC box with the exception of data buffers and clock buffers and the analog switches and the box would contain only a single clock domain. The metric for "good design" in such a system would be to minimize the number of gates. Unfortunately, I've not yet convinced myself that a one bit design is going to work well to my standard as I don't like high frequency noise. If you go to multibit you are talking FPGAs or custom chips to do the necessary dynamic element balancing (e.g. DCS ring DACs and ESS Sabre chips) and one has complex digital logic in the box, i.e. problem not solved.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: question is:- have you, posted on June 16, 2011 at 22:35:18
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 looked at Weiss or dCS units that use Firewire? Daniel Weiss argues quite strongly in favour of this.

You are looking at Ground Plane noise and this isn't the definitive factor when it comes to sound quality (as how systems deal with this overall appears to be the key).

All I know is that usb devices as everyday audio is problematic. I lost my usb3 audio connection yesterday thru some unknown event (playing DVDs) even though Win 7 said it was all fine. In the end I had to uninstall the hub and driver and reinstall to get it back. I only knew this w/o disassembly because the ports worked perfectly under dual boot XP.

If inmates are prepared to use usb audio on a machine as an audio playback device only and limited computer functionality, then my objections are not valid. But one pays dearly for this and in the extreme we are talking of $20000 plus.

But the obervation that one has simply turned the machine into boxed audio through complicated and sometimes black magic means still remains.

 

RE: I have never had the opportunity to compare USB2 with FW using the same source & dac., posted on June 18, 2011 at 14:56:27
endust4237
Distributor or Rep

Posts: 205
Location: europe
Joined: February 12, 2006
 Soon the new pro Mytek DAC will be available with USB2 and FW and DSDIF, SPDIF, AES interface, quite a bit of comparison possibility. :)

The three best DAC I have heard at home were the MSB Platinum (USB2), Weiss DAC202 (FW) and Sonic 304/305 (FW). So I do not think we can have a kind of ranking based on the interface only, the whole implementation probably more important than the PC connection itself. Hope to get the latest Weiss INT203 (FW) as well and to connect it to a Devialet amp, to see how it can go against M2Tech Hiface EVO (USB2) and the Bryston BDP-1 file player through the AES input of the Devialet. Devialet promise a wifi - based streaming from PC/Mac/iOS(?) sometime late summer up to 192kHz/24 bit which can be comparabale any other way of connecting and streaming. At least according to the promise of the manufacturer. Just to make our life more exciting :)

 

Page processed in 0.048 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]