Greetings Stan. If you assume the CCD has a linear response, there is no need for calibration sequences fainter than the usual mag. 10-12. For southern observers, the Harvard E-region standards are the thing to use, since they along with the Landolt equatorial standards define the system in practice. In re attaintable accuracy, you'll find that getting the kind of accuracy you're used to with photomultipliers is pretty much impossible with CCDs. There is almost no all-sky CCD photometry that has real external accuracy below about 0.015 mag. (per observation sigma) published in the literature. The usual limitation beyond doing the observations and reductions carefully is flat-fielding. The best CCD work I've seen is by Alistair Walker (formerly of South Africa but now at CTIO) and by Arne Henden (USNO-Flagstaff). As discussed last night (western US time!) I wouldn't push too hard on getting all five UBVRI colors, just V and some color (probably V-R). You'll have to measure in two colors to determine V anyway. Yes, it would be nice to have all the colors, but the purposes of rapidly getting zero-point stars around a transient event, you want to minimize the work in order to get results out as soon as possible. Good luck with it. \Brian