Constraining the (gamma,pi) amplitude for E2 N->delta

A recent analysis of pion photo-production multipoles using p(g,pi0) data from Mainz is repeated, successively adding constraints from other observables and varying the number of fitted partial waves. The original analysis is shown to have been underconstrained and suffered from ambiguities. The inclusion of additional observables in our analysis results in a very different, but stable, ratio of E2/M1 multipoles at the delta that is consistent with the Mainz data set.