At present, 10-bit samples are transferred across the USB link as 16-bit words, so there are 6 unused bits per sample.
How about including a counter in the unused bits that increments every N samples, where N is fairly large? That way the DomesdayDuplicator program could detect lost samples in regular capture mode, as well as in test mode.
I've put a prototype implementation of this on my seqcounter branch. It seems to work - checking the counter takes about 5% extra CPU time on a slow machine...
At present, 10-bit samples are transferred across the USB link as 16-bit words, so there are 6 unused bits per sample.
How about including a counter in the unused bits that increments every N samples, where N is fairly large? That way the DomesdayDuplicator program could detect lost samples in regular capture mode, as well as in test mode.