24 Mar
2009
24 Mar
'09
10:21 a.m.
Hi,
On Tuesday 24 March 2009 13:18:37 Frank Barknecht wrote:
Arnold Krille hat gesagt: // Arnold Krille wrote:
I know that clemens knows his stuff. But he is talking about the usb-protocol
here. The same can be said about the firewire interfaces: Because of the way
firewire works all have the same latency[*].
But still that is the latency of the bus/transport. You have to add the
latency of the actual hardware and I am very certain that there are different
converters/chips in use in the different usb audio devices. And even with using
the same chips, the programming in the firmware can be different resulting in
different latencies.
[*] Probably this statement is not as true for firewire as it is for usb as
firewire seems to allow synchronuous and asynchronuous transfer, so its
probably two different protocol-specific latencies in the fw-world. But for the
details you have to ask ppalmers or wagi or Jonathan Woithe who all know more
about firewire than me.
I just know from actually programming (or at least trying to program) an
atmega with an ad-converter connected by a bus that inside the device you have
an additional latency, no matter which device.
Arnold
I _seriously_ doubt that all usb devices have the
same latency. Indeed
they do alls have the same latency introduced from the usb-protocol added
to their specific latency, but that specific latency still depends on the
actual hardware and the driver. And yes, there are different usb-audio
drivers in the linux kernel, just search for *usx2y* for example. And the
hardware latency is determined by the converters and the processing chip,
just like with any other sound device, be it pci, pcmcia or firewire
based...
I'm no expert on this, but IMO ALSA-dev Clemens Ladisch is one. He
wrote:
"Due to how the USB protocol works, all USB audio devices have the same
latency."