In addition to observational problems, a peculiarity in the response function of the detector should also contribute to the errors in the photometry and positions. As described above, a rough model of the average response of a pixel was included in simulations run at b=25 and b=10. A PSF with FWHM = 2.5 arcsec was used, producing frames with the forementioned dither pattern and no dead pixels. The results are shown in Figures 12 and 13; comparisons can be made with Figures 4 and 5 from the standard run (no subpixel response.)
The average ``best'' photometric error seen in the Kamphot photometry, at
the brighter stars, is roughly 
mag. There is also a
significant increase in the positional error. Thus, subpixel response (as
currently modeled) can introduce 1-2
of the subpixel response is not as sharp as modeled, the effect might be
somewhat less on PSFs whose FWHM is close to the width of a single pixel.