If you don't specifically set the asynchronous sync type in the endpoint
descriptor, the usbaudio driver will assume that the endpoint is either
adaptive or synchronous. In other words, for the render side, if you leave it
as unspecified the driver does not expect there to be any feedback pipe and
synchronization is taken care of within the device and the system will feed
data to the device at the nominal sample rate. If your device cannot handle
that adjustment internally then you need to implement the feedback pipe.
For the capture side it is assumed that this is an asynchronous endpoint and
the system will simply take the data at the rate it arrives.
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx] On Behalf Of Alan Davis
Sent: Wednesday, November 04, 2015 8:12 AM
Subject: [wdmaudiodev] Re: usbaudio.sys and descriptors
I've gone back to not including any sync and it enumerates and installs ok.
I also have it appearing in the sound applet now and data is coming through-
A next challenge is to deal with synch, currently it gets glitches after some
On 4 Nov 2015, at 21:05, Clemens Ladisch <clemens@xxxxxxxxxx> wrote:******************
Alan Davis wrote:
Ok I tired setting to adaptive
Which one? Both? The capture endpoint cannot be adaptive without
How does your device actually work? You cannot blindly choose some
random setting; you must correctly describe the behaviour of your
Try Linux ...
Post message: mailto:wdmaudiodev@xxxxxxxxxxxxx
URL to WDMAUDIODEV page: