Just a couple more questions for Stan, plus a comment. Does the ST6B have an actual electromechanical shutter, or is it using the frame-store of the TC241? There is a simple method to check the shutter accuracy by using constant-voltage LEDs which I discuss in the CCD book. I'd suggest using an off-center aperture mask to remove the secondary mirror obscuration. The best system for what you are doing (looking at bright variables where there are no nearby comparison stars) is to use a wide-field telescope, something like the TASS Mark IV. Then you *do* have comparison stars in the same field. Otherwise, you have to use only photometric nights (which are often hard to predict). Plus, if you are only looking at comparison stars every 30 minutes or so, this is much too long between measures even for PEP work. Where exactly are you located? I thought you originally stated Auckland (my memory must be failing!). Another technique for comparison stars, which will save you some time, is to take *two* exposures of the same field with the same filter, back-to-back. Make the second exposure much longer than the first, and use the fainter stars for your comparisons. Then, at least, you have the same airmass and are much closer in time to the variable's measure. You could even sandwich the variable measure inbetween two deeper exposures to monitor transparency variations. The R&I flatfields should look pretty much the same (except for dust, of course). The B flat especially will look quite different because of the different QE in the blue. That is why your R&I data turn out more consistent. A 30 arcsec aperture is *really* big for CCD work. The usual rule-of-thumb is 4 to 5 times the fwhm seeing disk. If your seeing is 2arcsec, then you should be using an 8-10 arcsec aperture. The larger aperture will include neighboring stars, more sky noise and lower the accuracy of your results. You don't have to include exactly 98 percent of all of the flux, it just has to be the same percentage of flux from image to image. You indicate that binning your CCD doubles the number of stars you can observe. I'd think that there is considerable dead time just in rotating the filter wheel and moving to a new position, such that the 25 second vs. 6 second read time shouldn't be a major difference in your efficiency. Using full resolution will permit longer exposures before saturating the ADC. Arne