photo help-green haze?

[graham.hobart (Graham stevens)]

Hello all, I took these two photo's Friday night. The sky was quite cloudy so I only got about 20 frames duration between 45sec-1minute. I used Astrophotographers Tool software to control the camera and PHD to guide with autodither on for the first time.
I processed with DDS then StarTools. Calibration frames for the same duration (shorter flats) were used.
I have these two green hazy areas at different points on the photos. I thought I might have caught some nebulous gas but it's funny how it's in a different area on each of the photos.
The camera was an unmodified 450D with a CLS CCD filter.
Any ideas? (Scope was an 80mm refractor)
Graham

[irwjager (Ivo)]

Hi Graham,

The only thing I can think of, is possibly a Wipe anomaly (did you use it?), caused by dead (sub) pixels in your data. See if you can spot a dead pixel in one of the channels at the center of the green patches. Dead pixels should usually all be eliminated after stacking and calibrating dithered data, but they may indeed remain in individual subs.

If all this is indeed the case, upping the 'Noise Filter' parameter in the Wipe module should fix things.

EDIT: Yep, I can see dead pixels at the center of the pages (check the stars at the center of both patches). The dead pixels will cause Wipe to basically throw its hands up in the air for that particular area of the image, as a dead pixel doesn't fit with what it knows about the background level - it's way too black!.

Cheers,

[graham.hobart (Graham stevens)]

Thanks Irwin- that fixed it- I think my mistake was doing a complete wipe because I liked the fact it stabilised my RGB levels because I am over that blue cast I am getting with the CLS filter (despite using a custom white balance on the camera).
This time I made a single pixel mask and then grew it a few times then used that as a wipe basis- did clip it to black though so maybe a tad dark. When I increased the headroom M43 didn't stand out as much, not that many photons in this set though as was grabbing what I could through the murk.
Thanks for the tip as always!
Graham

[irwjager (Ivo)]

Originally Posted by graham.hobart

Thanks Irwin- that fixed it- I think my mistake was doing a complete wipe because I liked the fact it stabilised my RGB levels because I am over that blue cast I am getting with the CLS filter (despite using a custom white balance on the camera).

Yeah, the color calibration is an added bonus. I usually don't need to play around with the colors any further.
Full Wipes usually give decent results, but it depends on how 'natural' your data is. Worst case, you may indeed have to work around any anomalies (dead pixels, stacking artifacts, trees, mountains - anything that is darker than the 'real' background level). Small defects can be filtered out by the noise filter, larger ones you can indeed work around by by simply masking them out.

This time I made a single pixel mask and then grew it a few times then used that as a wipe basis- did clip it to black though so maybe a tad dark. When I increased the headroom M43 didn't stand out as much, not that many photons in this set though as was grabbing what I could through the murk.
Thanks for the tip as always!
Graham

Nice find - that'll work too. As a matter of fact, it's very similar to what PixInsight's DBE does. The downside is that the algorithm only gets a few samples to work with, so modeling may be less accurate and some clipping may ensue (which you can indeed offset by the 'add headroom' option). The latest version of ST will let you choose a blend between 'Clip<0' and 'Add Headroom' results, instead of forcing one or the other. It's a feature inspired by an animated discussion I had with Jase on clipping...

Cheers,