On Sat, 30 Jun 2018, R0b0t1 wrote: > Hello, > > My code is a userspace application using libusb for now, but this > seems the best place to ask. > > I repurposed some USB audio class device code. At this point I've only > changed the descriptor to report as vendor class so nothing else grabs > the device. I can receive and play PCM audio from the device, but > there is popping and clicking. The program I am using is attached; it > is written in Python. The data is saved to RAM and then written to > disk. > > The device does not implement the synchronization aspect of the USB > audio class and when it registers as a microphone a normal userland > can read it properly. > > During testing I think I noticed behavior related to my issue - for > some reason there will be long runs of zero returned or no data > returned. Usually this happens when a single transfer is scheduled. > Many transfers in a row tend to report the data properly, but from the > audible results I get it seems that runs of zeros still seem to slip > in. > > The problem seems more noticeable when using the Python libusb > bindings but it still exists when using libusb directly. Can anyone > suggest what to look into?
Have you tried using usbmon to capture the data as it is received by the kernel? Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
