Are all asynchronous USB inputs similar?


I was wondering if they were all the same, or were certain designs better than others?
koestner
I would expect that different asynchronous USB interfaces will generally differ significantly in terms of how much noise and jitter (timing fluctuation) ends up being present on the clock (the timing signal) that is provided to the D/A converter chip itself. That in turn can be expected to have significant sonic consequences.

Even if the electrical designs of two asynchronous USB interfaces were identical (and in general they won't be), just differences in mechanical layout, signal routing, grounding, power distribution, etc., would be likely to result in significantly different amounts of jitter.

Regards,
-- Al
Al is correct. Each design is different. Its not just jitter either. The output driver impedance matching and voltage is critical too.

Steve N.
Empirical Audio
The one thing that they all have in common is the poor sound quality. USB is for connecting your printer, not your DAC. If you notice the new era of servers, they have abandoned USB due to its poor sound quality. It was originally used as people just did not want to open their computers, so, since USB is like a belly button (everybody has one), thats what they used. It was a convenience thing and never meant for high end applications. This just snow balled with the manufacturers (it was easy and CHEAP)as they all embraced the interface. You can now see how the new designs are not using USB due to a myriad of reasons, one being the natively horrible jitter factor. You need to spend alot of money just to (try) to tame USB jitter.
Dont believe everything you read. One thing that all PCI interfaces have in common is poor sound quality. The power and ground planes inside a PC are just too noisy to achieve a low-jitter master clock. Best to get this outside the computer and on its own low-noise power supply as in Async USB. Computers are built as cheaply as possible, including the PCI infrastructure. They are barely good enough to operate, but not good enough to support a low-jitter clock. I know because I worked as a designer and design manager at Intel corp. for more than 17 years and at times in the platform group.

The right USB interface on the right power system will always beat a PCI card inside a computer, as well as 99% of CD transports.

Many more rooms at shows get awards with USB interfaces that those with PCI interfaces.

Steve N.
Empirical Audio
Tell that to baetis. Just cause there was USB at more rooms is a silly response. The world is flat again? USB is done. Thank the lord. Puters are built cheap but that ol USB output is made of gold. Ithink the Brooklyn bridge is still for sale.
John Mingo of Baetis Audio makes a compelling case against USB use if the best sound quality is the objective. He expresses his perspective in the February TAS review of his digital music server. There are two sides of this debate, it seems he just listened and made a choice. I know of no better way to decide.
Charles,
Here is a better way to decide: customer posts on various forums:

http://www.empiricalaudio.com/news-and-reviews/off-ramp-converter

Steve N.
Empirical Audio
A better way to decide than to listen for yourself? Now, I have 'heard' everything!

Steve,
For someone to compare very well regarded USB products such as yours and the Baetis server,surely listening to both is the best way to make a choice.
Charles,
Certainly listening is the best way. I offer 30-day money-back, less shipping, so the risk is minimal.

I was just commenting that one vendors review is not as good as many customer unbiased posts.

Steve N.
Empirical Audio
They both have trial periods. But given the complexity and the pretty different approaches not sure comparisons are that easy. One is a computer and one uses a computer you have.
The issue is which pathway offers the best sound quality. Baetis eschews
USB in favor of dedicated high quality SPDIF connection vs Steve's high
grade USB products approach. Someone with a computer and Empirical
Audio USB parts compared to Baetis with SPDIF using the same music
source.
Charles,
I welcome such a comparison. I have not lost in a fair shootout yet.

Steve N.
Empirical Audio
From my research I am not convinced taking the SPDIF output from the motherboard is the best approach, and I was never convinced USB was either, although I have used it extensively in the last 6 month because of the convenience. Not that guys like Nugent and Rankin do not know what they are doing, they do in my opinion.

From my perspective I would fundamentally agree with Cerrot that a well designed sound card should provide the best option, but I do not really feel there are any worthy of trying, not even the Lynx which I have heard extensively. Forget [email protected] as well, although it is reasonably priced versus others I have seen.

Overall, the best I have heard in my system is where the digital output of the transport is slaved to the DAC. Right now I just started running Ethernet into my Resolution Audio Cantata and have to say I am quite impressed. Wonder why more people do not do this. If I was going to use a SPDIF interface I would be more inclined to try the Sonore Audio Ethernet to SPDIF converter than the Baetis in my system at this point.
Ethernet is the best possible option. We will all migrate to it eventually. It has the same benefits as Async USB, but avoids the issues that occur in computer software, making the computer/ipad a "don't-care".

I don't expect Ethernet to improve sound quality significantly over good USB interfaces, but computer optimization will hopefully not be necessary anymore, which is a big win for audiophiles.

Steve N.
Empirical Audio
FWIW, assuming comparably good implementation of the various possible approaches IMO the most ideal one would be a **reliably functioning** wifi link, with the computer (and the digital and switching noise it generates) located nowhere near the audio system, and powered from a different AC branch.

Wired ethernet would be my next choice (with the computer powered from the same AC branch as the system, to avoid possible ground loop issues).

The following paragraph from this paper by Steve elaborates, persuasively IMO:
Networked audio (Ethernet), both wired and WiFi is a unique case. Because the data is transmitted in packets with flow-control, re-try for errors and buffering at the end-point device, it is not as much of a real-time transfer as USB, S/PDIF or Firewire. The computer transmitting the data packets must still keep-up" the pace to prevent dropouts from occurring, but the real-time nature of the transfer is looser. Unlike with other protocols, there can be dead-times when no data is being transferred. Networking also avoids the use of the audio stack of the computer audio system since it treats all data essentially the same. This avoids kmixer on XP systems and the audio stacks on Mac and PC Vista. Because of the packet-transfer protocol of Ethernet and data buffering at the end-point, the jitter of the clock in the computer is a non-issue. The only clock that is important is the one in the end-point device. Examples of end-point devices are: Squeezebox, Duet and Sonos. This would seem to be the ideal situation, which it certainly is. The only problem that can occur is overloading the network with traffic or WiFi interference, which may cause occasional dropouts. The problem for audiophiles is that the majority of these end-point devices were designed with high-volume manufacturing and low-cost as requirements, with performance taking a lower priority. As a result, the jitter from these devices is higher than it could be. It should be the lowest of all the audio source devices available.
Regarding USB, I have no relevant experience, but my perception has been that a lot of Audiogon members utilize asynchronous USB DACs with results that are satisfactory to them, while only one member consistently criticizes it. FWIW.

Regards,
-- Al
The USB audio interface dominates the audiophile world because the USB interface dominates the computer world. Whether or not it's the best interface is somewhat irrelevant since as a practical matter for most it's all there is. We should be thankful that it works as well as it does.

BTW, I use firewire and a dedicated soundcard in my systems. I was an early adapter of wireless and found it not very reliable.
Onhwy61,
The USB dominance is not debatable and the Baetis Audio designer openly concedes this fact. His position is it not the best sounding and he simply offers an alternative he believes is superior in sound quality. That's it really, he doesn't deny that many enjoy USB linked systems.
Charles,
In order to make broad conclusions about USB IMO, they must have auditioned every USB interface with every server configuration. This is highly unlikely. They can only rely on their own experience, just like the rest of us.

There are a lot of poorly conceived USB interfaces, even big names. Most of them on popular DACs are not that great. This is the reason why some people have a less than stellar experience with USB.

If you ask Antipodes, whose server is excellent, they would likely make the opposite claim, that with the RIGHT USB interface, USB is far superior to S/PDIF.

I, on the other hand have had a LOT of experience with many different high-end PCI cards, Sonos, Squeezebox2, 3, Duet, Touch, Firewire interfaces and several USB interfaces, adaptive and Async. I have designed or modded all of the above. My USB interface is 5th generation (Off-Ramp 5) and 6th is in prototype stage. I have used or modded servers including Naim, Qsonix, Soolos and Antipodes as well as modified Macs and PCs. I believe I am in a better position to determine which technology is superior. Empirical Audio has been in business for 18 years and I have been doing digital design for 38 years. Async USB driving S/PDIF or I2S can be every bit as good if not better than PCI bus driving S/PDIF, providing a good design and implementation. Driving I2S it should be better because it avoids S/PDIF conversions.

Steve N.
Empirical Audio
Steve, I'm with you, if computer sound card audio was better I would think more than one vender would use it. The USB is more convenient and you can use any computer you have in your house. I personally am not interested in trying the one product offered.
To be fair (unlike cerrot), Bryston uses a [email protected] PCI card in their server with very good results, although not quite as resolving as the best USB IMO. I thought their system at RMAF was better than any other all-one-manufacturer system. They even have their own speaker now.

Steve N.
Empirical Audio
For sake of accurate information in this interesting discussion, there's no sound card for the Baetis as it's directly connected to the mother board. I agree with the convenience of USB but that doesn't necessarily correlate to superior sound, two different arguments. It's good to have choices among high quality offerings.
Charles,
We can all see the original question. While I posted my usual, smug, remark, the responses (thanks, Charles), did conjure up some nice discussions (finally), where there's finally talk where the USB interface may just be more convenient than better sounding.

Happy to be a part of it.
I have heard quite a few different async converters and they are all different.

Thanks
Bill
OK, some excellent responses here. My reason for the question was that I currently have an Auraliti PK90 sending USB to a Blacknote DAC30 which has async USB. It's a few years old now, and with the digital world that's more like a decade. So I was wondering if the newer (and cheaper) USB dacs were passing it by, or were they designed with the same technology? Thanks everyone.
Is there anything out there that confirms that one USB Conversion chip in itself is better than another?? i.e
Xmos, Cmedia, Crystal, Tenor, FPGA ???
I believe that there are most likely excellent examples of most
Koestner - some USB interfaces are constantly improving, even my own. I have done at least 2 mods to the Off-Ramp and added an optional power supply in the last year. It is 5th generation, Off-Ramp 5.

The thing to realize is that an external USB converter like an Off-Ramp 5 driving your USB DAC with S/PDIF will undoubtedly beat your USB interface and probably your transport too. I have lots of customers using the Off-Ramp 5 to drive USB DACs, abandoning the built-in USB interface. 30-day money-back, less shipping.

Off-Ramp 5:

http://www.empiricalaudio.com/products/off-ramp-converter

Steve N.
Empirical Audio
Timlub - these different Async implementations have more to do with the drivers or lack thereof and maybe support for DSD.

The jitter performance of each of these is really independent of the base chipset, assuming the designer knows what he is doing. A lot don't.

This is the reason why I don't expect the SQ of my newer XMOS implementation to be much different than my current M2Tech implementation. If anything, the galvanic isolation will make the difference. Galvanic isolation can be designed into the M2Tech implementation as well, but I didn't.

Steve N.
Empirical Audio
Thanks Steve, Has anyone heard that Microsoft will ever make a USB Audio 2.0 Compliant driver?
Steve N.,

I looked at your Off Ramp and it said to use drivers for MAC and Windows, but my platform on the Auraliti PK90 is Linux. Do I need a driver for that, and how would I install it?
Koestner - Some Linux servers have put the edits in to support M2Tech/Empirical products, others haven't. I know that Sonore and Antipodes have these mods in them. I used Antipodes at last RMAF.

If your server does not put in these edits, then you may have to wait for Off-Ramp 6, which supports Linux without custom drivers.

Steve N.
Empirical Audio
I'm very happy to be off "the bus" and on the fast track with UPNP and DLNA over hardwired Ethernet. It doesn't do away entirely with PC platform hassles but I can deal with software and power supply optimization vs. the interaction of USB and MOBO hardware.
You will all see the transformation away from USB quickly.
We are in an industry where we will send back $50,000 amplifiers to have them rewired with another $20,000 of wire. We upgrade fuses and have our houses rewired. How, then, do we just accept a standard USB output on a consumer grade PC to be the be all and end all in computer audio? I know, we have Steve at Empiracle Audio with his 30 day money back guarantee, to make it all better. Me, I would rather not buy off the rack in a custom audio world.
Davide256, I am with you on this one. I did like my set up with my USB converter, but UPnP via hard wired Ethernet is amazing me. Mac Mini with JRiver as the media server and library, Resolution Audio Cantata as the renderer, using either PlugPlayer or JRemote as the control point. Would like to try wireless version via a bridge but am told Hi-Res files could be problematic with that configuration.
The thing to realize is that an external USB converter like an Off-Ramp 5 driving your USB DAC with S/PDIF will undoubtedly beat your USB interface and probably your transport too. I have lots of customers using the Off-Ramp 5 to drive USB DACs, abandoning the built-in USB interface. 30-day money-back, less shipping.
This has been my experience also. I think just about anyone who has used a usb to spdif converter wouild agree that it is better (in almost every case)than using built in Asynchronous usb on the DAC.
I use a Musical Fidelity "V-Link" as a way to escape the usb interface. It will only do up to 96 but it is inexpensive and most of the music on my computer is from ripped CDs so that is adequate fidelity for me. I'd like to try something like the Berkely Audio or the Off Ramp but both of these cost more than my DAC an Oppo 105) and they just don't make sense from that standpoint.
In my own experience with the Cambridge Audio DACMagic Plus and the Audioengine D1, USB 2.0 works fine in the office but my home PC is a gaming rig that is *heavily* overclocked (it requires water cooling to remain stable).

In that application, USB 2.0 is unlistenable and I rely on toslink SPDIF from the mobo to get clean audio.
Everyone seems to think that the DAC is more important than the master clock device (USB converter or reclocker). Not so. The jitter of the source is more important, particularly if you goal is to eliminate harshness and get more detail, better slam and imaging.

I would much rather have a less expensive DAC and a good USB converter.

Steve N.
Quote" Everyone seems to think that the DAC is more important than the master clock device (USB converter or reclocker). Not so. The jitter of the source is more important, particularly if you goal is to eliminate harshness and get more detail, better slam and imaging"

Steve, you convinced me of this at least a year, maybe 18 months ago. I have used 3 different inexpensive converters and even at this level, converters made a difference. I wish we could all afford your products, but Life is what it is...
Why is it that Top quality USB Converter/re clockers are not typically built into the DAC itself, or why can't it be or maybe why is it that outboard USB converters seem to always be better than an onboard unit? Power supply?
"Why is it that Top quality USB Converter/re clockers are not typically built into the DAC itself, or why can't it be or maybe why is it that outboard USB converters seem to always be better than an onboard unit? Power supply? "

I'm afraid that its just experience and design skills that are lacking. Most USB interfaces on DACs are either new designs by designers that have never done this before, or they are designs by third-party contractors with more interface experience, but little product experience. Product experience would allow them to select the best clocks and other components, optimize the power supplies and the ground-plane infrastructure in the DAC to achieve really low jitter using USB to I2S.

The built-in USB interface (USB to I2S) on my Overdrive DAC is now actually better than the Off-Ramp 5 with Dynamo power supply that I sell.

Totally uninteresting tracks like "Gimme Shelter" from the Stones Let it Bleed now sound fantastic, with weight, depth, slam and great tonality. I can actually understand most of the lyrics now.

Steve N.
Empirical Audio
What about I2S directly from the computer to the DAC. I know a company that offers a I2S PCI card with a HDMI style I2S output.

This company also makes DACS with I2S input of course.
Mordante - I2S is not particularly helpful if the clock is still powered by the computer. The main objective is to reduce the jitter of the master clock. The best way to do this is get the Master Clock out of the computer and on its own power system/supply.

Also, S/PDIF can be superior to I2S if it is galvanically isolated from ground returns. I2S and HDMI I2S is very difficult to galvanically isolate. I have both HDMI and RJ-45 I2S outputs on my products. I was the first to put HDMI I2S on my Off-Ramp 5 USB converter.

Steve N.
Empirical Audio