ASI2600MC flat problem

[SkyHiker]

I saved my lights in Ekos as lights, then saved my flats as flats then stacked them in DSS. Apparently when saving flats, the format is different. Clearly the flats were not applied but it looks like the whole debayering went wrong, the features seem to be twice the size as on the lights and shifted.

I am reading a lot of the same problems on CN in recent posts, but no clear answer.

Admitted, DSS tells me that it thinks it has to apply the Canon 350D DSLR format to my flats but I don't see a choice for ASI2600MC in the pulldown. So clearly there is a mistake.

I suppose I could have avoided the problem by not telling Ekos that I was taking flats and simply save it as color just like I always used to. Now it saves it as gray scale images with a weird pattern, each 4th pixel is bright.

Has anyone encountered this, and what is the cure?

[SkyHiker]

I unchecked the DSLR checkbox in DSS that tells it what to do for gray scale images, that could simply be it. Re-stacking the whole bunch, I will keep you posted...

BTW that's pretty severe vignetting, and the collimation is off, or at least the illumination - maybe the secondary is not quite square with the focuser. I haven't touched that part yet.

The Paracorr 2 seems to be doing OK though, so I see some good sides. It will all work out, just getting to know this camera.

[Gordon]

I unchecked the DSLR checkbox in DSS that tells it what to do for gray scale images, that could simply be it. Re-stacking the whole bunch, I will keep you posted...

BTW that's pretty severe vignetting, and the collimation is off, or at least the illumination - maybe the secondary is not quite square with the focuser. I haven't touched that part yet.

The Paracorr 2 seems to be doing OK though, so I see some good sides. It will all work out, just getting to know this camera.

I think you are on the right track with unchecking the DSLR box. Be sure to let us know if it works....

[AstroBee]

I believe you are correct about the secondary not being square to the focuser. based on that strange vignetting pattern.
Good luck with the other issue, I've never used DSS so not much help there.

[SkyHiker]

When I checked the FITS header of one light and one flat in Nebulosity I noticed there was a difference in focal length. I recalled that I found out during the session that I had used a focal length of 731 mm (of my Mak-Newt) while the 12" that I used had 1200 mm focal length, so I changed it during the session. If DSS is smart enough to correct for focal length, that could be the issue.

I tried again last night under windy circumstances without doing PA (I figured it was a worthless night anyway) so I got 41 subs with only 1 light that DSS would allow to stack, lol, and 22 flats. This time DSS had no trouble with it. So it looks like changing the focal length mid-session is a bad idea because it makes your FITS files inconsistent.

[SkyHiker]

When I checked the FITS header of one light and one flat in Nebulosity I noticed there was a difference in focal length. I recalled that I found out during the session that I had used a focal length of 731 mm (of my Mak-Newt) while the 12" that I used had 1200 mm focal length, so I changed it during the session. If DSS is smart enough to correct for focal length, that could be the issue.

I tried again last night under windy circumstances without doing PA (I figured it was a worthless night anyway) so I got 41 subs with only 1 light that DSS would allow to stack, lol, and 22 flats. This time DSS had no trouble with it. So it looks like changing the focal length mid-session is a bad idea because it makes your FITS files inconsistent.

This argument turned out to be completely wrong. The difference between the two sessions is that I had used a nonzero offset with the series that failed. If you use a nonzero offset you must use bias frames or else the calibration will fail. The DSS people recommend taking bias frames, always. Although apparently you can get away without them sometimes (all the time for me with zero offset until now).