OK. As I watched the tumble-weed pass by, I found out what the problem is. However. I'm unsure of the driver behind the behaviour. It seems that there are indeed two channels sampling the data and channel 2 is "out of phase" by 1/fs. These two channels are then interleaved into the read buffer as if...

Hi All. I'm trying to acquire a single channel analogue input using the I2S DMA. This is somewhat successful in that I can configure the I2S and read data correctly and it gives me sensible values. However. I seem to be getting two channels of the same measurement, separated by 1/f, instead of one. ...