Notes on Spectra Reduction

I’ve examined the flats to determine the image portion of the chip, and the bias portion of the chip.  These correspond to:

TRIMSEC [10:1198, 5:800]

BIASSEC [1205:1230,5:800]

Next step is to combine all bias frames with zerocombine.  Question: what does ‘to clean up’ mean in the observing notes? Perhaps this was a bias image taken to be sure the CCD was cleared?  I will avoid their use.  Bias images correspond to images 0076 through 0127.  Hmm, the standard zerocombine simply uses all with type ‘zero,’ perhaps I should hide all ones that I don’t think are necessary (i.e., the ‘clean up’ zeros).  Removed all ‘clean up’ biases, for fear they would contaminate the average.  can be put back in again if needed.  This included image 0154, which was not labeled as ‘clean-up’ but have been, for it was done directly after dewer refill.

63 zero images were used

ccdproc was run

combine flat-fields, moved aside those with < 25 000 counts.  not useful.

Flats are broken into ones with 30 sec exposure, and those with 40 sec exposure.

had to change ‘rdnoise’ in ‘flatcombine’ package to 3.5, this was not included in the header, so needed to do it manually

flatcombine done: 65 flats used

Normalized the flat field using ‘response’ package.  The flat field seems really bad, needs a order 6 polynomial to fit to it.

ran ccdproc with normalized flat field, however there is still a gradient left in the images.  Need to map the light correction function. hmmm not sure if i did this properly..


my initial pass through the above reductions steps seem to be done…okay.  I would like to go through them again, with a little more understanding.