psf question

patB wrote on Jun 28, 2007

Hi, I am trying to build a good psf for an image im working on and when i display the output from the substar routine after my zeroth order psf, the cores of the psf stars look strange. Random patterns of pixels are removed quite well, but there are still very bright pixels in the cores of the stars. The result (when zoomed in) looks kind of like a random checkered pattern of light and dark pixels within the psf star cores. Is this okay? I am using a guide by Phil Massey and he says the cores should be removed cleanly after this first substar application, but that there may be residuals from neighbors. However, the bright pixels left over in MY image are obviously not from neighbors. I know this is an iterative process, so I wonder if my zeroth order psf is actually okay with these residuals? If not, are there any suggestions as to where the problem might come from? I have been fiddling with my datamin and max values, but have had the same results. Thank you.

---

Francisco Valdes wrote on Jun 28, 2007

Hi,

I'm afraid the author of the task has left NOAO and I am not very knowledgable on the PSF task. I hope you might get a reply from some other reader with more insight. If I were to hazard a guess, what you are seeing is the higher residual Poisson noise in the core which would be normal in displaying a star subtracted image. It could also have something to do with how well sampled the star are. If the sampling is good then the fainter stars should show less of the effect.

Good luck,
Frank Valdes

---

Last post on Jun 28, 2007