photometry of moderately crowded fields using SExtractor + PSFEx

[navii98]

Hi,

I am trying to implement a photometric pipeline for a survey. The PSF varies significantly and I've found that PSFEx models that reasonably well.
So, I've been trying photometry using SExtractor + PSFEx.
I am dealing with moderately crowded fields, where we don't need the repeated loop of "find, model psf, subtract, find again" similar to daophot.
One pass of detection using SExtractor seems good enough.

I had some queries which I could not find in the documentation. I've listed them below in no particular order -

1. In particular how PSFEx deals with 2 stars that are very close, in order of 1-2 FWHM. For example, DAOPHOT groups these kind of stars and fits them simultaneously.

2. during the PSF photometry is the PSF fit within a certain fitting radius? If not is there any way to achieve this behaviour?

3. When using a different measurement image and detection image with WCS headers is there some kind of tolerance when the measurement is performed? e.g. let's say the WCS is accurate up to 0.5''. When PSF photometry is performed on a source detected in the detection image, can the fitting position of the PSF model vary?

4. Is there any way to perform photometry on detected sources from a list of coordinates? akin to forced photometry.

5. Is the volume of the output PSF model from PSFEx normalized to unity?

6. in PSFEx output, how is the FWHM calculated?

7. In the docs it is written that in MAG_AUTO a crude subtraction of neighbors is performed, is the same done for MAG_APER as well?
   The MAG_APER differs significantly from photutils for stars that have neighbours.

8. Does background type local apply to PSF photometry as well?

9. What is the difference between PSF_mag and pointsourcemodel_mag?

10. I understand it may not be the right place to ask this, but are there any other modern softwares/module that perform better with PSF photometry and crowded fields? Maybe something that makes use of PSFEx models?