Would a Thunderbolt audio interface decrease latency (in theory)?

Question

I've always been curious as to the latency limits of USB 2.0/3.0, Firewire 400/800, and now thunderbolt on audio interfaces. I'm really excited to see a thunderbolt audio interface on the market, but will that really allow for any improvements if I'm only concerned capturing say 8 mic inputs @ 24bit/196khz (for constant variables) at once?

I've always tended to stay away from USB 2.0 in favor of firewire because I know it's faster, but does it really make a difference in terms of audio interfaces?

Answer 1

Yes, a faster transference decreases latency, but as far as I'm concerned it wouldn't be any less latency than USB 3.0 or firewire if you are not transferring a large number of tracks. Latency relies more on hardware and processing power than the transfer itself. In the case of computer audio, the latency depends more on the driver and the transfer protocol, but I'm no expert. It would be hard to tell until we have reviews of latest audio equipment.

Answer 2

Yes, GBs of data going from point A still has to be decoded and processed at point B. I can't see how it's done without a large amount of parallel processing in the hardware.

Answer 3

• Question 1: In that case the used hardware components and the hardware drivers have substantial more influence on the signal latency than the bus/connection. So the answer is: Theoretically Yes but, practically No. Because it is absolute negligible.

• Question 2: Yes it make some differences, but the reason is not the bandwidth (the nominal speed) but rather the interface latency. FireWire has a very low latency because of a flat topography and a very small protocol overhead. Also the devices can communicate directly with each other and have direct memory access. - The reason: FireWire (aka M(usic)Lan) was specially designed as low latency interface for music and video devices.