Colors not correct after Flatdivision

[pfile]

this is a bit of a threadjack, but i did some experiments as suggested by NGC7789.

i compared my EL flats to my sky flats by dividing them and i can see kind of a circular artifact in the frame. the EL flats were taken with the panel laid directly over the aperture of the telescope, which is a 6" RC design.

i thought maybe this had to do with the central obstruction being so close to the light source, so i put my dew shield on the ota and placed the panel on top of that. now the panel was about 24" away from the front of the OTA. dividing those flats into the sky flat looked a little better, but there is still an artifact.

i should mention that even the sky flats leave some artifacts in my subs, so comparing these two types of frames does not say too much.

i also used DBE to fix the vignetting in an uncalibrated frame and then divided the different calibrated frames into the DBE'd frame. all types of flats seem to give artifacts with the sky flats being best, then the EL flats from 24" away and then the EL flats directly on top of the OTA. but the frames calibrated with sky flats are bad enough that i really have to do DBE to each sub before registering and integrating them or i end up with difficult-to-solve gradients in the final image (i have images from both sides of the meridian and do not have a rotator.)

anyway, i then thought i'd make a 'real' flat box out of the panel but the box is very shallow and the panel is only a couple of inches from the aperture again. the flats are similar to the ones taken with the panel right on the aperture.

i took flats with the box at 0 degrees and 90 degrees and sure enough the flats do differ. the panel is, as i mentioned, an elcheapo ebay model.

anyway, none of this points to a fix for my real problem which is that even sky flats seem to give artifacts. i just finished collimating this OTA and the analysis seems to indicate that the collimation is good. however it's obvious just from looking at the flats that they are not "smooth". this is of course a cheap telescope and so i'm willing to believe that the problem could be reflections and/or stray light inside the OTA, though it does have baffles and is all flat black on the inside.

rob

[Christoph Puetz]

Dear PI-Community,

as promised before, I did some further tests with the flat fields 
Well it may be not really surprising, but I found that really the "correct" exposure time of the flats is the most important thing.
As some fellowers stated in this thread before, you really should avoid noise (from too short exposure times) or gradients
from bad flat field generators (or the sky itself)

For me I found the following rules suitable:
- for my Canon EOS 1100d and 450d (both modified for Ha) I use an ADU of 4000 for flats.
- for my ATIK 383L+ an ADU of 50% of max. ADU value (= about 32.000) gives the best results.

If I "underexpose" the flats, I also had some adverse effects, i.e. clipped black regions (from the OAG) become "pure" white 
So this was very instructive for me.

Unfortunately, the "correct flat exposure times" were not that relevant in general astrofoto-literature, that I read before.
So testing and comparing the results was the best way to find out