Originally posted by carolina_sky I would definitely feel more comfortable if DCU exported RAW (i.e., a single RAW with the merged/processed image rather than the 4 image RAW) rather than TIFF.
There shouldn't be too much of a difference between a 16-bit TIFF file and a special PixelShift RAW file (in which RGB information is present for each pixel).
The TIFF file has white-balancing, gamma encoding, and a tone curve baked in, but with 16-bit information there should be a lot of leeway to make corresponding changes after the fact.
Having said that, it would be nice to be able to store PixelShift images using the "Linear DNG" format, leaving all major "developing" decisions for later.
Originally posted by carolina_sky I don't know if it makes any practical difference in image quality, but there would certainly be a size difference.
Are you sure?
Remember that part of the size advantage of regular RAW files comes from the fact that they only store either red, green, or blue at each pixel position, instead of full RGB information.