@hirata10 and I were looking at the OpenUniverse PSF model (left) compared to the Piff fit model on ffov images (right), and we notice a flip in the orientations/coordinates.
We think that this should come down to the fact that the project's PSF model is in the exit pupil coordinate convention, whereas the OU and science images will be in science coordinates. I think these are rotated 180 degrees with respect to each other, you might be able to find a useful description here: https://github.com/Roman-HLIS-Cosmology-PIT/PSFSim/blob/main/docs/coordinates.rst
I can try to take a look later this week, but if you see an obvious fix, please feel free to go ahead.
@hirata10 and I were looking at the OpenUniverse PSF model (left) compared to the Piff fit model on ffov images (right), and we notice a flip in the orientations/coordinates.
We think that this should come down to the fact that the project's PSF model is in the exit pupil coordinate convention, whereas the OU and science images will be in science coordinates. I think these are rotated 180 degrees with respect to each other, you might be able to find a useful description here: https://github.com/Roman-HLIS-Cosmology-PIT/PSFSim/blob/main/docs/coordinates.rst
I can try to take a look later this week, but if you see an obvious fix, please feel free to go ahead.