NGC7000 N.American Nebula Processing help needed please

[Notty]

I wanted to try my modded 1100D on an emissions nebula for the first time.  I only managed 8x300s subs after I decided after the week I'd had I needed bed.  Conditions weren't ideal I guess still not quite pitch black at midnight I think, but I got what I got, with 25 or so bias, 30 flats using my usual method of tee-shirt stretched over scope with laptop which screen held at the end, histogram peak 1/3 from the left, but I only managed 4 darks as my camera battery ran out.  Anyway veeery strange behaviour in DSS today.

This is the raw output WITHOUT flats; I haven't done anything to it but my point is it looks fine in terms of sensible alignment etc:

This is what it comes out with WITH flats.  Not just the weird shapes but it doesn't seemed to have stacked properly:

This is the master flats; it's very dark initially but stretching the levels etc you can clearly see the bunnies and vignetting (should it matter that I've played with it to be able to see it on the screen?  I thought the maths behind the scene didn't care but maybe I'm wrong)

So flats aside, Importing the stacked image minus flats into PS to have a play, and there is the most unholy vertical streaking in the image close up.  Is this my low number of darks, or the flat fails do you think?

Any advice on this dogs breakfast VERY much appreciated, especially anything on the flats oddness!

EDITED to add... and the fact I don't even know what this DSO is really meant to look like makes it difficult to ascertain if the color profile is correct, or whether the dark corners are ALL down to vignetting, etc

[glowingturnip]

Initial thoughts are that something strange is going on with the debayering, or lack of, on the flats.

If you look closely at the flat, you can see a fine regular grid which is what a monochrome image taken through a bayer grid looks like before it's been debayered into a colour image.  Not sure why your flat should be undebayered (is that a word), or whether that's what's causing the weird stacking, though that's probably likely.

Don't stretch your flats before applying them by the way, you'll end up over-correcting - flats are applied multiplicatively, and a stretch is a non-linear transformation of brightnesses - in short it won't work.

Not sure I can see the vertical streaking you're referring to.  I can see some dark lines vertically below some of the brightest stars, which I believe is a 'feature' of some sensors, similar to blooming you might see on some CCD's.

If you like, load up a zip file of all your raw lights and calib frames onto dropbox, and I can see if I can calibrate and stack it for you better in Pixinsight.

Cheers

[Notty]

Hey that is is really kind of you.  I'm not an expert on Dropbox, I'm uploading a zip for each set, but not sure how I share it with you? 

[glowingturnip]

there should be a share link which will just give you a url to each one.

I'm at work at the mo, will have a look later

[MartinFransson]

I've had DSS do strange things like that also. Some of the problems were correctes by lowering the number of detected stars to 100-150. Do this by adjusting the star detection threshold.

Sent from my phone using Tapatalk

[wxsatuser]

First off........are you using the very latest DSS?............if not get it.

Stack the lights without any calibration frames and see how it looks.

The NA neb looks like this, excuse my processing, to the right is the pelican.

Think yours is the other way up, mine was done with a camera lens.

[Notty]

"Excuse my processing?" That's gorgeous! I can't believe it's the same DSO - how long/many subs was that? Alas yes I have the latest version of DSS.

If I stack just the lights I get basically something that looks like the first pic in my post.

[wxsatuser]

My image is 12x300secs, Canon 60Da EF 200mm L f/4 at ISO 1600.

I think you have more in your image, it needs teasing out.

[MartinFransson]

Alas yes I have the latest version of DSS.

If I stack just the lights I get basically something that looks like the first pic in my post.

Then I'm 99% sure it has to do with the star detection limit! I've had the exact same thing happen. DSS mistakes noise or hot pixels for stars and everything gets warped. Apply the median filter and tighten the threshold to make the number of detected stars smaller.

Sent from my phone using Tapatalk

[Notty]

Martin, thanks a bunch that was it!  DSS has now stacked it all and it almost looks sensible... I'll try some processing later and see what i can get out of it.

Thanks again!

[MartinFransson]

Martin, thanks a bunch that was it! DSS has now stacked it all and it almost looks sensible... I'll try some processing later and see what i can get out of it.

Thanks again!

Well, you know... been there, done that [emoji3]

Sent from my phone using Tapatalk

[Notty]

Just to finish this off, fairly happy ending thanks especially to MartinFransson and GlowingTurnip I managed a decent stack, flats did their thing and I've done some rough processing to get something that looks a bit nebulousey. Didn't realise I'd need a much wider field view to get the whole North America shape in, might have a go with my DSLR's lens.

[glowingturnip]

nice one