It can be hard to use metacalibration photometry for validation as we know Gaussians are not a good fit to the underlying galaxy profiles. It would be useful to have another pipeline module that fits standard ngmix models (e.g. exp, model, bdf) that can be run either before or after mcal. As long as we use SOF it should be fairly fast in comparison to mcal.
It can be hard to use metacalibration photometry for validation as we know Gaussians are not a good fit to the underlying galaxy profiles. It would be useful to have another pipeline module that fits standard ngmix models (e.g. exp, model, bdf) that can be run either before or after mcal. As long as we use SOF it should be fairly fast in comparison to mcal.