M33 and Background Removal

[Smashtie]

I've been doing a little experiment, and last night was my first attempt at the technique, on M33. I wanted some opinions from the more seasoned veterans here.

I've recently started using a Celestron f6.3 reducer/corrector on my Nexstar 8. This has some serious vignetting with my 1.25" back. I know I could/should switch up to 2", but I'm cheap, and it got me thinking about flats. I really want to try to remove not just optical errors in the scope, but also background glare (not too bad where I am in Hertfordshire, but not great either!)

I'm using the standard Nexstar 8 mount, unguided (cheapskate that I am) so I have to take short images (10 sec max).

Rather than shoot flats, darks and biases, last night I swung the scope across a short distance between each shot, and took a sample of images of random bits of nearby sky.

Then I made a background frame in fitswork by just taking the median for each pixel, un-registered (in RAW). I attach a jpeg showing how this looks. Promising, I thought! I used 5 batches of images, 1 of lights, then a background batch, then back on target for another light batch, then another background batch, and finally back on target for a last set of lights. Total about 200 images, taken with a Nikon D7000. All the while the moon was starting to rise.

Then I stack by manually discarding blurred frames and just subtracting this background from each image, debayering, then registering and adding. Total number of lights used was 107, which is about 18 minutes of exposure. I'm pretty pleased with the result given the equipment. It's much better than the uncorrected results I was getting before, although certainly not completely flat yet!

Next step in the experiment would be to try using only the nearby (in time) background frames to correct each light frame. This should help remove errors across the batch because of the increasing illumination from the moon, and the change in elevation of the target object.

So, there it is. Any comments? Is this a common technique? Any tips on how to improve the correction, especially the colour which seems a bit off in the outer region?

Anyone have an optimized D7000 debayer file for fitswork?

Rupert.

[Smashtie]

Oh, here's the background image I promised.

[ollypenrice]

The background image does look like a flat but I'll be honest and say that the main image looks as if it's in desperate need of flats so I think I'll be sticking with the longhand method!

My take on this? You need flats. T'was ever thus, especially on a massively vignetted SCT. The F6.3FF is only so-so (at best) at evening up the illumination.

Olly