USB DACs with 24/192 via USB

Marco (Jax2), the TAS article has indeed been controversial, and Gordon did answer why he withdrew his DACs from the review (questioning the competency of the reviewer). See the following threads for some more information:

Audioasylum: Did I miss the discussion re: the latest TAS article on USB DACs ...

Audioasylum: An Open Letter To Robert Harley

Audiogon: Absolute Sound Article on USB and Firewire

AS much as I respect efforts from Wavlenth, Imperical Audio, Bel Canto, Lavry, etc… using a standard (gosh) high end DAC being fed by a good to very good sound card via AES, BNC, or coax sure has it’s benefits. Not to mention the abilities to process more than adequately the higher ‘numbers game’.

I don't see any benefits of introducing an extra step of a sound card conversion from USB to AES/coax. In fact, I really don't see the advantage of using the badly designed AES/coax SPDIF connection that lacks a separate clock transfer and barely can make 24/192 (most SPDIF inputs are limited to 24/96 too).

One of the big advantages with asynchronous USB is that you can place the clock right next to the DAC chip and slave the PC to the DAC clock. That gives the lowest jitter and something that is not at all possible with AES, coax or any other traditional conversion schemes. Finally, the only limit to data transfer rates via USB is due to the lack of drivers. If a company is willing to write drivers for their DAC then 24/192 won't be a problem. And I don't see USB disappearing from PCs anytime soon.

There is something to be said for Firewire too - the best computer DAC implementation I heard to date uses Firewire (the Weiss DAC2/Minerva). But that requires extra drivers as well for 24/192 operation.

Bindjim - here is some nice info on USB and the driver requirements in plain terms:

24/192 high rez via USB - Drivers, EMU0404, etc.

The biggest problem with USB right now is really the mediocre and after-though implementations of many current DACs that just add USB as a second interface without using the advantages of USB (the TAS review just reflected that). Some of the DACs that have a dedicated USB interface of course do have good implementations, except then you throw all eggs into one basket which is not always a good thing.

Therefore, I presume that if it receives a 24/192 digital bitstream, it will decode it in native mode as there's nothing to upsample.

Not all DACs can receive a 24/192 stream even though they upsample. In particular 24/192is only possibly via USB if the unit requires special drivers to be installed on Windows or OSX (see for example EMU0404 USB). A USB input on a DAC does not necessarily mean it goes all the way to 24/192.

For example:
- Bryston BDA-1 (only 16/44 and 16/48 via USB)
- Bel Canto e.One DAC3 (only 16/44 and 16/48 via USB)
- Benchmark DAC1 USB (16/44, 16/48, 24/88.2, 24/96 via USB)
- PS Audio DLIII (only 16/44 and 16/48 via USB)

This websites list the USB input rates for different DACs with USB input (if available):

The Well Tempered Computer

After I get the DAC, I'll compare at 24/96: 1) Lynx soundcard with an AES cable, 2) USB into my Empirical Audio Turbo USB-to-AES converter, and 3) USB straight into the DAC. Should be interesting... Also, looking forward to trying 24/192, which was the original target.

On the Playback design website it mentioned that USB only goes to 48kHz - I guess you'll go high rez 24/192 via the Lynx only? I don't think the USB comparison will be very meaningful - given the implementation my guess would be the USB input is again not close to the Lynx.