I am currently using the CX3 along with a sensor that outputs 10 bit RAW pixel data, and I am using code similar to the UVC example code. Is there is a way to tightly pack the pixel data received?Instead of getting 2 bytes for every 10 bit pixel, I can get 4 pixels for every 5 bytes of data. Does simply setting the GPIF to get 8 bits instead of 16 work?
I think that is not possible, it will depends if your image sensor has that ability.
I've just work with a 10bit raw image sensor, and the UVC example code. The code and the project are easy to understand and to adapt to a new sensor, the big problem is the image sensor, usually you don't have enough information to configure it properly.