An undersampled point spread function may interact with the microstructure of a solid-state detector such that the total flux detected can depend sensitively on where the PSF center falls within a pixel. Such intra-pixel sensitivity variations will not be corrected by flat field calibration and may limit the accuracy of stellar photometry conducted with undersampled images, as are typical for Hubble Space Telescope observations. The total flux in a stellar image can vary by up to 0.03 mag in F555W WFC images depending on how it is sampled, for example. For NIC3, these variations are especially strong, up to 0.39 mag, strongly limiting its use for stellar photometry. Intra-pixel sensitivity variations can be corrected for, however, by constructing a well-sampled PSF from a dithered data set. The reconstructed PSF is the convolution of the optical PSF with the pixel response. It can be evaluated at any desired fractional pixel location to generate a table of photometric corrections as a function of relative PSF centroid. A caveat is that the centroid of an undersampled PSF can also be affected by the pixel response function, thus sophisticated centroiding methods, such as cross-correlating the observed PSF with its fully-sampled counterpart, are required to derive the proper photometric correction.

20 pages, 14 postscript figures, submitted to the PASP