>> concern about large versus small pixels in relation to seeing (snipped) Yes, the system with pixels of smaller angular size on the sky should yield higher precision in principle. One way of improving things for systems with large pixels is to take multiple images. For instance, if I were running such a system, I would have it take back-to-back pairs with a command to shift the camera slightly (10"-30") between exposures. I would want to take pairs of frames anyway to guard against chip flaws, cosmic rays, etc. Another route (at least in survey mode) is to arrange the field centers to provide center-to-corner overlap (the center position of one frame corresponds to the corner of a frame taken in any diagonal direction). This provides the multiple coverage and random pixel sampling without reducing sky coverage so much (I wouldn't consider a little redundancy as wasted time in variable-star photometry). \Brian
vsnet-adm@kusastro.kyoto-u.ac.jp