While playing with my M-Audio Transit USB device (0763:2006) I saw a synchronization type that doesn't exist in the USB audio class. I get audio out but often the start is garbled with white noise and cracks. Wondering what's going on.

more /proc/asound/card1/stream0 M-Audio Transit USB at usb-0000:00:14.0-2, full speed : USB Audio

Playback: Status: Stop Interface 1 Altset 1 Format: S24_3LE Channels: 2 Endpoint: 3 OUT (ADAPTIVE) Rates: 48001 - 96000 (continuous)u Interface 1 Altset 2 Format: S24_3LE Channels: 2 Endpoint: 3 OUT (NONE) <<<<<<< illegal Rates: 8000 - 48000 (continuous) Interface 1 Altset 3 Format: S16_LE Channels: 2 Endpoint: 3 OUT (ASYNC) Rates: 8000 - 48000 (continuous)

The only definitions in the spec are Asynchronous, Adaptive or Synchronous...

lsusb tells me that there seems to be a synch endpoint for this setting:

Endpoint Descriptor: bLength 9 bDescriptorType 5 bEndpointAddress 0x03 EP 3 OUT bmAttributes 1 Transfer Type Isochronous Synch Type None Usage Type Data wMaxPacketSize 0x0126 1x 294 bytes bInterval 1 bRefresh 0 bSynchAddress 131 AudioControl Endpoint Descriptor: bLength 7 bDescriptorType 37 bDescriptorSubtype 1 (EP_GENERAL) bmAttributes 0x01 Sampling Frequency bLockDelayUnits 1 Milliseconds wLockDelay 100 Milliseconds Endpoint Descriptor: bLength 9 bDescriptorType 5 bEndpointAddress 0x83 EP 3 IN bmAttributes 1 Transfer Type Isochronous Synch Type None Usage Type Data wMaxPacketSize 0x0003 1x 3 bytes bInterval 1 bRefresh 2 bSynchAddress 0

I couldn't figure out what the code does in this case. I didn't find any sync URBs being retired so wondering if this is considered as a synchronous or asynchronous endpoint based on the additional sync endpoint? Any idea on how to debug further? It looks like there could be a second error in the descriptors: if this is a synchronous endpoint then the sync endpoint is meaningless, and if it is asynchronous then the information on locking time is irrelevant...

Regarding the noise on startup, should a delay be inserted after the mode is set? the spec recommends that the host sends silence periods for adaptive and synchronous modes. I see a set of mdelays being added in the quirks, should a new one be added?

Thanks for any pointers, -Pierre