Our hope is that from these phase retrieval calculations it will be possible to derive better mirror maps for the HST/WFPC and HST/FOC optical systems. These improved maps then could be used in Tiny Tim or other optical modeling programs to compute PSFs that would be useful for a wider range of HST image restoration problems.
Blind deconvolution may allow observed PSFs to be used for deconvolution in a way that allows the uncertainty of the noisy PSF to be properly incorporated into the image restoration iteration. The ideal blind deconvolution algorithm would doubtless incorporate a phase-retrieval iteration as part of the PSF estimation (see paper by Schulz in these proceedings for such an approach.)