09:55 pH5: tomeu: nice progress!
09:57 pH5: I haven't seen avg bias being used yet on imx8mp.
17:21 tomeu: I'm not sure of the exact value, but here looks like they start using it if there is any non-zero value that has more than 7 occurrences
17:21 tomeu: Pass: 548, ExpectedFail: 12, Skip: 40, Duration: 2:05, Remaining: 0
17:22 tomeu: I think input and output channels work already, but when the kernels get so big, there is a bigger chance of hitting a bug in the encoder