A critical appraisal of the SED fitting method to estimate photometric redshifts

We discuss the stability of the photometric redshift estimate obtained with the SED fitting method with respect to the choice of the galaxy templates. Within the observational uncertainty and photometric errors, we find satisfactory agreement among different sets of theoretical and empirical templates using the Hubble Deep Field North as a target galaxy sample. Our results suggest that, especially at high redshift, the description of the physical processes of photon absorption in the interstellar and intergalactic medium plays a dominant role in the redshift estimate. The specific choice of the template set, as long as this includes both normal and starburst galaxies, is in comparison a minor issue.