Crowded-field photometry from HST-imaging

We present a thorough investigation of stellar photometry based on HST imaging of crowded fields at $85\hbox{$^{\prime\prime}$}$ and $10\hbox{$^{\prime\prime}$}$ from the centre of the high-surface brightness elliptical galaxy M 32. The Principal Investigators of the present archive data have elsewhere presented an impressive colour-magnitude diagram of the field at $85\hbox{$^{\prime\prime}$}$. Based on the same data we enlarge on their photometric analysis and supplement with error estimators that more clearly show the implications of severe image crowding on the stellar photometry. We show that the faintest stars ($I \gtrsim 25.0$, $V \gtrsim 26.0$) are found too bright by several tens of a magnitude. For the field at $10\hbox{$^{\prime\prime}$}$ we conclude that it is not possible to obtain reliable stellar photometry, standard deviations being larger than 0.4 mag. Artificial-star experiments show that only very few of the brightest stars of the luminosity function can be expected to represent single objects, the majority being either spurious or not as bright as measured. The paper as such introduces and demonstrates basic guide lines which may be used when dealing with stellar photometry of severely crowded fields.