On Thursday 08 Oct 2015 15:54:46 Clemens Ladisch wrote:
In theory, the driver supports this, but there are known to be some bugs. And in any case there should be no problem getting the capture stream to start (it's the playback stream that would have a slightly wrong sample clock);
Thanks Clemens,
That was my reading of it too. My latest quirk tests for playback use explicit feedback, which seems to mirror what the Windows driver does.
I'm still having no luck with recording though.
it's possible that this device needs some vendor- specific commands, or a different order of commands. (The interesting USB transactions would be everything except the isochronous ones.)
I've been looking at the usbmon dumps I have of recording under Windows and I don't see anything that stands out, but I'll keep looking.
Cheers,
Keith