Why use a sound card?

[Steve_S]

Just curious, I see a lot of reference to the Lynx AES16 on this site. With good USB DACs on the market, why use a sound card at all? Is it just a matter of preference between SPDIF and USB, or is it something else?

 

I have a MiaMidi PCI card but I don't use it because I can output via USB directly to my DAC. Part of my decision to use USB is because the USB imput of my DAC has 1ppm TXCO where the coax and toslink inputs are 12. Another part of my decision is because using USB eliminates another piece of hardware from the server that uses power and produces heat. But I'm curious why others choose to use sound cards in their systems.

 

[The Computer Audiophile]

Hi Steve - The Lynx card has been around for a while and in heavy use by the recording industry. In a way it's already been vetted. Also, it supports sample rates like 24/88.2 that some USB DACs do not and sample rates up to 24/192. The Lynx is very flexible in terms of clocking internally or with an external clock. AES/EBU also allows a longer cable run than USB or FireWire and can input into many more DACs than USB. The jitter spec is very low on this card as well.

 

[BEEMB]

 

Steve,

 

Chris' comments are of course completely correct - but you don't actually need one. Some modern DAC's do the job fine on their own. I tend to think of the Lynx card as a luxury though some of Chris' builds are cheap even including the card.

 

Matt.

 

[MahlerFreak]

Purely IMO, the leading edge of computer audio will move away from separate sound cards per se. There are just too many variables that can impact the performance of a sound card in otherwise generic PC environments, as many of us are discovering.

 

To avoid these variables, the obvious alternatives are either completely engineered single box solutions, where interactions can be controlled by design from ground up, or external box solutions communicating via asynchronous I/O, like the Weiss and Amarra Firewire boxes, or the Wavelength and forthcoming Ayre asynchronous USB DACs.

 

Standard synchronous USB communication, used in almost all USB DACs other than the aforementioned, will always be somewhat problematic, because the first digital audio clock in the system is still the computer clock. There are different approaches to removing the jitter introduced by the computer audio clock, but none is absolutely ideal. Asynchronous communication, whether via Firewire or specialized async USB, allows the external box to contain the first digital clock in the system. The computer itself then becomes effectively just a data server, shipping another packet of data across the communication channel whenever the external box requests it, but otherwise having nothing to do with audio per se (assuming bit-perfect playback software).

 

[Steve_S]

Thanks for the replies, that clears things up.

 

Mahler, asyncronous communication sounds like a great idea, but is anyone actually doing it in a home computer application?

 

[MahlerFreak]

Well, Firewire was designed basically as an asynchronous media streaming protocol, so all of the units that use firewire interfaces, like the Weiss, Amarra, and lots of pro-gear work this way. USB can be used that way, but only with very specialized software. The only USB DACs I'm personally aware of that do async USB are the Wavelength Audio DACs, and the forthcoming Ayre QB-9. A very clear explanatory whitepaper on the Ayre web site goes into more detail, if you're interested:

 

http://www.ayre.com/PDF/Ayre_USB_DAC_White_Paper.pdf