The M67 paper does cover the AP-7, which lists a canonical "SITe back-illuminated CCD," in this case the SI502AB. The coefficients I determined were b(v) = 0.2600 and b(r) = -0.2716 (that is, the SI502AB peaks between V and R). These don't match Seiichi's values, and like Brian, I wonder how he measured the slope since his slope varies so dramatically during a night. I presume Seiichi's "k" is the same as my "b(v)". I prefer to use a letter besides "k" since this is traditionally used for the extinction coefficient. The natural systems of reflectors and refractors are quite different; not only do lenses have different absorption at different wavelengths, but they are not achromatic at all wavelengths. As I mentioned in the paper, unfiltered measures are like extremely wide filters; very difficult to transform accurately into the standard system. The traditional method of reporting unfiltered magnitudes has been CCD or CCD(C), but that doesn't adequately describe the results, since it depends highly on the comparison star magnitudes used in your differential photometry. Were they V-band or R-band, or (groan) GSC? Did you attempt to transform onto the standard system, or just use the comp stars to zero-point your results? I don't like unfiltered measures, and use them only rarely in my research. Vsnet-chat sounds like a good list to discuss how such measures should be reported to make it as clear as possible what methods were used in the reduction. Arne