Is it possible to convert a video tape (of the case type) into a series of digital still frames and then back to video tape without being detectable? If not, what is the approximate PSNR or whatever metric of error of the current best process?
