iraf.net

I'm using continuum to fit out the blaze function in some echelle spectra. When switching between certain apertures, the task will crash and throw a floating point overflow error. For some reason my red chip data tends to go farther before crashing (for some observations it doesn't crash!) than the blue chip data. If I try to read the continuum normalized file that was produced, I get an end of file encountered error, presumably because continuum didn't finish writing the file correctly. Any ideas?

Please upload the problem image file to the anonftp at ftp://iraf.noao.edu/pub so we can try to reproduce the problem. We'll also need to know what platform you're using (Mac, Linux, etc), the IRAF version and the parameters you used. Otherwise there's not enough information to comment on what might be going wrong.