USB, Firewire or Optical

[earflappin]

"If one is better than the other, in my opinion, the evidence points in the other direction."

 

Clay, I agree with your statement. It seems from all of the chatter about how cables and server platforms (CAPS vs Aurender vs PC vs Mac) seem to impact the SQ of USB tethered DAC's, and a corresponding lack of chatter about the impact of these variables on FireWire tethered DAC's it seems that FireWire perhaps has some advantages as a realtime audio interface?

 

Witness also the MyTek 192 thread where there seems to be a consensus that the FireWire interface outperforms the USB.

 

Perhaps because FireWire is a more mature audio interface? Perhaps because it uses a dedicated chip set? I don't profess to have the answers, but what I can say is that I have experimented quite a bit with my LIO-8 FireWire interface and I have not found cable quality or CPU load to provide a statistically significant SQ impact and I have done some DBT with friends. I do believe I hear an improvement with integer mode, but I need to do some more evaluation to be sure of this.

 

If FireWire had these type of performance dependencies one would think discerning pro audio engineers like Barry D would have already surfaced them.

 

It is exciting to see that FireWire will live on with Thunderbolt and with an optical option.

 

[DM]

>If one is better than the other, in my opinion, the evidence points in the other direction.

 

May be you are right, I really don't know.

 

My friend is a senior electronic engineer in a consumer/pro audio company, he is designing both USB and Firewire devices. I remembered he help designed the BridgeCo USB/Firewire interface module (the firewire part is used by Weiss).

 

As all implementations are different, seems it is impossible to get an apple to apple comparison. Even on the BridgeCo module (which have both inputs), difference versions of firmware sounded different for the USB and FW1394 inputs.

 

So his comments should be read with "in theory" in front, but how "theory" transforms into real product is another matter.

 

 

 

[barrows]

The Mytek is not a good example for a FW vs. USB comparison. Clearly the USB interface used on the Mytek is not SOTA. Here is a quote from Michal at Mytek from the Mytek Thread:

 

"Now, this DAC also has USB1.1 interface intended for quick hookup without drivers. This input is synchronous and followed by hardware SRC which removes USB jitter. This USB1.1 solution is exactly the same as in Benchmark's DAC1."

 

The USB input is not asynchronous on the Mytek, and they claim that the FW input is.

 

 

 

[orgel]

barrows quoted Michal/Mytek:

 

""Now, this DAC also has USB1.1 interface intended for quick hookup without drivers. This input is synchronous and followed by hardware SRC which removes USB jitter. This USB1.1 solution is exactly the same as in Benchmark's DAC1."

 

You're forgetting the other USB port, which is USB2.0 and asynchronous. Both the FW port and the USB2 port require installation of Mytek's drivers. See the manual.

 

--David

 

[barrows]

Do you have exact details of this? My understanding is that the USB receiver is Rigisystems USB Pal. So far, there has never been any confirmation that this approach is actually asynchronous, and Rigisystems does not claim asynchronous operation on their website. Additionally, there have been reports (unconfirmed) that Audio Research also uses a Rigisystems approach in the DAC 8, and some have found the USB interface in that product lacking (better performance using an external USB-SPDIF converter). Michal has not yet answered my questions re the Mytek, and when they say "asynchronous" it is not entirely clear if they are referring to the USB interface itself, or asynchronous re-clocking in the ASRC, or in the ESS chip itself...

I am not trying to critisize the product in any way, I am just curious as to the exact details of its operation.

My experience is that the best true asynchronous USB implementations leave nothing to be desired when it comes to performance, both measured (jitter) and sonically. Especially when used in HOG mode and integer playback. For measurements, look at the jitter spectra in Stereophile's review of the dCS Debussy-perfectly clean for USB.

I have nothing against FW, my point is just that either approach (FW, USB) will provide "perfect" performance when properly implemented. FW does have an advantage in that it by nature has an "isolated" bus, but with proper set up USB can achieve the same thing (choosing proper USB port, or using an SOtM uSB output card, using HOG mode).

 

[orgel]

@ barrows:

 

Nah, I'm not privy to any special info. I'm quite sure that Michal or one of the other Mytek folks has stated that the USB1.1 port is driverless and synchronous and that the USB2.0 port (and the FW) both require proprietary drivers (which I know are available on the Mytek website) and are asynchronous. There's also a July version of the product brochure -- which may have disappeared from the website since they posted the manual -- that states this. I'm attaching that.

 

I'm hoping Michal will chime in (after he makes sure they've shipped a bunch of units to Vintage King, so I can get mine) and clear this up.

 

I know you're not being negative about the product, and my expectation is that FireWire will be the better-sounding interface ... although I plan to try both hook-ups.

 

--David

 

Edit: The brochure exceeds the file-size limit, so I'm just attaching the relevant page.

 

Edit2: 5MB is the limit for attachments? I think not. Had to work some magic to get it to stick.