Quote:
Originally Posted by CaptainCook
Hi Tony,
Thanks for your comment. I have added some details below which I hope you find helpful but feel free to ask anything I may have missed.
1) Imaging: I was consistent with each panel. I settled on 2 mins at f/2.8 using this lens as a good compromise between SNR and vignetting. I set myself a minimum of 32 min integration-I have sometimes got panels with more data (e.g. Vela) but excluded it from the final mosaic so the noise is consistent which helps the compositing imo and definitely a more consistent result. The 32 min panel has proven just about achievable even with the bad weather- sometimes I was getting only one 2 min sub in a night! To also be consistent I never imaged below 40 degrees even though I could probably have gone down to 30 before obstructions etc.
2) Planning: I have used Telescopius which I can definitely recommend. This allows you to enter (or prospectively set) your sensor angle, grid size and overlap (which I set to 18%). You can get this angle from plate solving/astrometry etc and check at any point eg you know how the lens focus ring is very stiff and once or twice I moved the whole lens accidentally. Even at 18% the composition can struggle where there is no obvious structures to match in hindsight maybe use 20%.
3) Composition. I use Microsoft ICE (Image Composition Editor) it is absolutely fantastic and I dont know why this is no longer supported but you can find a download if you go looking. Incidentally, I have used this for terrestrial and MW panoramas too. I even used it to put my virtual backyard into Stellarium.
Sometimes the automated tool fails where you have panels that basically contain fresh air, which meant I to separately do that section with numbered panels before manually loading it into ICE to give it a helping hand. You will get what I mean if you start using it. Then I added that small panel to the rest of the images. In other other words my 56 panels might end up being 60 images input into ICE.
4) Processing. I did this in PI but kept it very simple. The steps I used were ABE, noise XT for each panel prior to ICE. Once the mosaic was completed I used Starnet (again in PI) and curves/layers in gimp to add back in. At that point you are talking 1.5 GB- Starnet takes 40 mins on my laptop. I then reduce 50% and save as jpg etc
I look forward to seeing your mosaics on here!
We need more black and white widefield stuff 
Cheers
Gary |
Hi again Gary,
Thank you for your detailed reply. It seems my approach is quite similar to yours as far as the planning & imaging although I'm shooting at F2 with a full frame OSC camera. The larger sensor means less panels are needed but the down side is more distortion in the corners.
Initially I was using ICE for the composition but it failed because the panels were too dark & needed more stretching I think. Originally I thought it would be best to do the majority of the post processing after the full image was composed but I realise that was a mistake now & you've helped to clarify that.
I'll give ICE another go now using Background Extraction in Siril & noise ext. in Photoshop as I don't have PI.
Thanks again for sharing your work flow Gary, it's been very helpful to compare notes. I'll post an interim mosaic shortly.
Cheers,
Tony