Siril / ASTAP Colour Changes

[WolfieGlos]

I don't want to get into a discussion of "realistic colours" on this one, but I would like to know if anybody has any information or knowledge on how Siril and ASTAP interpret colours from an OSC (I'm using a DSLR), and why this is occuring.

First up, I'm shooting with an l'enhance filter.

So constant cloud cover is meaning that this project is taking a LOT longer than I had been hoping for, and forecasts of clear nights turn out to be total lies. If the forecasts had been right, this would have been finished two weeks ago. So to pass the time I've tried stacking and processing the first of two panels for the Heart Nebula in a 2x1 Mosaic (possibly increasing to include the additional nebulosity to the "left" of the image so making a 2x2 - weather dependent!).

Initially I had a really good 5 hours for the first part, which I stacked and which produced these colours with Siril by just opening up the image, running Background Extraction and cropping - whilst viewing on Auto Stretch:

 

So not processed at all, no Asinh or histogram stretches. But after a 2 week wait for more clear skies, I finally got some subs the other night to start the next panel, albeit with a 75% Moon and I only captured 3 hours due to intermittent cloud, so not enough but I wanted to try creating the Mosaic to ensure all was coming together.

So here is the same view in Siril for the other half of the nebula:

Which looks the same in terms of the red colours, so no issues there.

But then putting the two together with ASTAP to stitch a Mosaic, I get this:

 

Somehow I now have a completely different colour of red. What I'm not understanding, is how ASTAP is changing the colours in the image? Especially given that all of the options are greyed out when using the Image Stitching option as shown below, suggesting they are not used for this.

I have tried changing the stack method, from Image Stitching to another option....which then allows the greyed out options to be changed. So I tried this, changed the colour options in the De-mosaic method.....and then coming back to Image Stitching, but it has made no difference to the output.

I don't have APP to try the mosaic option through it (I used a free trial a year ago, didn't like the software) so I might have to stick with ASTAP for now. But I'd like to know how / why it is changing the colour of stacked images, which it is then only stitching together? I can understand if I had stacked each panel in ASTAP using the options to tweak colours as it allows?

[Rallemikken]

Color balancing of different elements in a mosaic is not easy. Especially when they are shot on different evenings. Consistent stacks with the same settings/options. Looks like you use Siril. First I'd say you have a green tint in both. Used the "remove green filter"? Another option is photometric color calibration.  I've found out that the green filter gives a very consistent base background color without adding noise or altering the overall colors. Then the fine-tuning in the image editor of your choice. I use Gimp. The lower panel is obviously darker than the upper. Have you feathered the intersectioned area?

Lots to add, but the two most important: When doing mosaic's, try to shoot the panels close in time (best is same night) and don't touch the camera. Nor focus or rotation. If possible, set up a sequence  where you shoot first the first panel, then the second, and then next and so on. After collecting 20-30 subs of each, you start over, and keeps on until daylight. Repeat next night, and the night after that if you wish. This will give you more consistent stacks of each panel. Easy if you have the possibility to plate-solve. Switching between targets takes only 3-5 minutes.

Second, do the final fine-tuning and stitching in a image editor. Takes some effort, but gives 100% control. Import each panel as a layer, arrange them, and do the final color correction on each.  Finally you use the gradient tool to feather the intersectioned area. When all looks good, merge down.

[WolfieGlos]

Thanks for the response. I'll try to respond separately; 

11 hours ago, Rallemikken said:

Color balancing of different elements in a mosaic is not easy. Especially when they are shot on different evenings. Consistent stacks with the same settings/options. Looks like you use Siril. First I'd say you have a green tint in both. Used the "remove green filter"? Another option is photometric color calibration.  I've found out that the green filter gives a very consistent base background color without adding noise or altering the overall colors. Then the fine-tuning in the image editor of your choice. I use Gimp. The lower panel is obviously darker than the upper. Have you feathered the intersectioned area?

Yes, there is a green tint because I haven't run SCNR on those images yet, they are just a preview on "Autostretch Viewing" following a crop and background extraction.

To be fair, I didn't try PCC so I have just tried it on the individual stacked panels, and the colours change to almost exactly match the mosaic that ASTAP has created, so it's almost as if ASTAP has applied a colour change during it's mosaic creation. Despite not having any options to control this.

Correct, the lower is significantly darker. I captured the lower panel during the start of the Moon period. The upper was with a 75% waning Moon, about 2 weeks apart, so a significantly brighter sky. Initially I was forecast several clear nights in succession so the intention was to capture 1 panel per night. In reality....I had 1 night only! Sadly, that's just been the terrible weather we've had. Even before you noticed it, I have been considering re-capturing it unless I can somehow blend and control the join in the software. I did ask this question on SGL but noone replied, so decided to go for it anyway. Oh well.

11 hours ago, Rallemikken said:

Lots to add, but the two most important: When doing mosaic's, try to shoot the panels close in time (best is same night) and don't touch the camera. Nor focus or rotation. If possible, set up a sequence  where you shoot first the first panel, then the second, and then next and so on. After collecting 20-30 subs of each, you start over, and keeps on until daylight. Repeat next night, and the night after that if you wish. This will give you more consistent stacks of each panel. Easy if you have the possibility to plate-solve. Switching between targets takes only 3-5 minutes.

That's not a bad idea actually - I'll put this down to a learning experience and consider this for next time, especially given what I just said regarding the forecasted clear skies. Like you say, at least each panel would have the same levels of brightness, especially if imaging over successive nights.

I set this up with NINA's framing and sequencer, and saved it so I can go back to it quite easily on successive nights. The scope is left alone, imaging train not touched nor is the rotation.

11 hours ago, Rallemikken said:

Second, do the final fine-tuning and stitching in a image editor. Takes some effort, but gives 100% control. Import each panel as a layer, arrange them, and do the final color correction on each.  Finally you use the gradient tool to feather the intersectioned area. When all looks good, merge down.

I might give this a try. With the difference in sky brightness between the two frames, I could see that it needed some finesse to blend them together. I might try this before deciding whether to re-shoot the upper panel.