Conclusions

Our conclusions can be summarized in the following points:

1. The aberrated HST PSF modifies the brightness profile of an elliptical galaxy up to 2 arcsec from the center, in a way which is not sensitive to the exact shape of the brightness profile, ellipticity, major axis position angle, and small PSF variations.

2. The brightness profile derived from a Richardson-Lucy deconvolved image reproduce the original brightness profile with accuracy higher than 0.15 magnitudes arcsec at radii smaller than 0.2 arcsec, and higher than 0.05 magnitudes arcsec at larger radii. The performance of deconvolution is not noticeably affected by the intrinsic shape of the brightness profile nor by other parameters such as ellipticity, isophotal twist, or deviations of the isophotes from pure ellipses. In addition, if the PSF used for deconvolution is chosen wisely, i.e., close to the center of the galaxy, deconvolution is robust against small PSF variations due to jitter, spectral shape of the point source and position of the PSF on the chip.

3. Studies by van den Bosch et al. (1994) show that more care is needed when deriving the ellipticity, the major axis position angle, and the deviations of the isophotes from pure ellipses, especially in the case in which these parameters vary with distance from the center. In general, it is safe to assume that ellipticity, position angle, and Fourier coefficients are recovered correctly by Richardson-Lucy deconvolution up to 0.5 arcsec from the galaxy center.