A solution to latency problems with an aging USB Interface

[Andy Peters]

OK.... what I was thinking of before fat fingering to remove half the last post was, that yes, bandwidth should be bonzer, but when more channels get TDM'ed into the stream, that more delay is expected (to be compensated properly by the DAW, buffered, latency compensated et cetera)...

So my question, is whether the driver is padding a bunch of extra channels into the USB stream? probably does not have much to do with buffer sizes?

USB devices can specify different "alternate interfaces," each with their own bandwidth requirements. I've seen this done to support different sampling frequencies and word lengths, and in a couple of cases, only "some" of the channels on a device are "enabled" because perhaps bandwidth is limited. This last case could be where the device complies with USB 2.0 and supports High-Speed transactions, but it could have been plugged into an older computer which only has a USB 1.1 Full Speed port. It still works but not on all channels.

The selection of the alternate interface is done whenever an application needs to use it, so the bandwidth isn't wasted when nothing needs to access the device.

Having said all of that: if the application selects an alternate interface which enables 10 channels at some sample frequency and word length, that bandwidth is reserved. So even if the application isn't actively listening on input 6, the device still sends (blank) data for it and the host driver takes care of dealing with it.

-a