I'm a big plus one for LRGB stacking too. It’s the biggest thing I wish SharpCap had.
However, I'm much less concerned about scripting automated filter wheel control. I just want to be able to see live the live integration into colour in an EEA session and have some control over it.
Something I like to do is take a Luminance mono image and then overlay H-Alpha in the red. This allows you to see interesting details in comparison. Highlighting the star forming regions in a galaxy, for instance. Obviously it is the same with other narrowband colours too - turning channels on and off shows different details in structure.
What I’m thinking about should in principle be quite simple. Basically I would like to save 3 FITS image stacks and overlay them as different colour channels in the software, rather than in Photoshop, since this takes away the live experience.
In my imagination there would be an extra “Multispectral” tab added. This tab would allow you to show each channel separately, or merge two or more to give a colour image.
If you are merging the RGB as a whole, the curves in the UI would work exactly like in Photoshop. The gamma and white/black points would just be taken from the histogram tab. If it is difficult or complicated to add in curve manipulation as in photoshop, I don’t think it is that important. Really it’s just about the levels and the gain — the same controls as already exist.
I really hope you don’t mind, but I’ve done a very quick UI mock up to hopefully clarify and explain.
- Multispectral_mockup_s.jpg (43.01 KiB) Viewed 6204 times
I would hope the UI mockup is self explanatory! However, — just to explain what I’m meaning:
Display
Here you have the option of selecting what is shown on screen.
The default is “Luminance / OSC” and this is how SharpCap works currently. Here the image would plain greyscale or full colour, depending on the camera.
This display choice here would change the viewing image and be persistent as you switch between tabs.
You can pick one or more channels to display.
You can turn on and off colour channels. This will have a live effect on the viewing image.
Each channel on its own would show the image as two tone in the relevant colour.
Select two and they would be overlaid on top of each other.
Select RGB and it should give you a full colour image (assuming your images are correct).
Behind the scenes I would think this dialogue box would essentially be turning on and off colour shaded greyscale overlays. Each tinted overlay would be an already integrated LiveView FITS stack. That stack would be created in the same way that SharpCap works now — acquired and level changed in the Histogram tab.
Acquisition
Here you select the channel for acquisition.
You can only pick one of these at a time.
This would normally be on Luminance/OSC. In this case everything in SharpCap would works exactly the same as it does now.
If you selected R, G, B then the program would start a separate FITS stack for that colour channel.
This will keep adding images to the channel stack until you press the “Pause” button or switch Acquisition channel again.
Pressing the “Clear” button in the main Overview will clear all channels in the Multispectral image.
If you change the Acquisition channel the software would:
- Automatically pause stacking, as if you have pressed the Pause button.
- Check if there is an existing stack for that channel
- if yes, it adds images to that stack,
- if no, it creates a new stack and stacks images in the new channel stack.
- Deselect all current display channel boxes. Then re-select only the relevant colour box (ie. select “Red”, and the display will only show the red channel image, regardless of what was on screen before)
- (The user could then re-add the other channels by ticking the Display boxes)
- If you return to a channel with an existing stack, SharpCap will continue adding to that stack.
Behind the scenes here the user is basically just telling the software to save the separate FITS stack with a channel identifier, presumably something like an “_R” “_G” or “_B” suffix on the file name.
The software would then just integrate new exposures into the relevant stack.
The drop down dialogue button here is to allow for 2 things.
- Clear Channel Stack
- Select From File
Clear Channel Stack — the user can clear the individual channel stack (similar to pressing the “Clear” button in the main Overview section, but for a single channel).
Select From File — the user can select a pre-existing FITS file from another session rather than the current one. This way, the user could use previous integrations as a base for the LiveStack. For example, one night I may only create a Luminance, the next night I may want to shoot H-Alpha and watch it stack on top of the existing image.
It would then grey-out the Acquisition toggle selection for that channel and replace the name (eg. “Red”) with the file name.
This image would need to be checked for alignment to make sure they will stack properly. The user could also override the alignment check if desired.
Using previous files would be a great feature, but if it adds a lot of complication, it would be a lower priority.
Channels
This is very much like a photoshop layout and would change the amount of weight given to each colour channel.
Each channel has a gain control and white point/black point controls. Again, these settings would just be taken straight from the histogram tab.
The user will want to manipulate these to get the most useful / pleasing colour balance.
Photoshop style curve manipulation would be great here, but a lower priority since it possibly adds complications in implimentation.
When stacking in a single channel, any changes on the main Histogram tab would transfer over to the Multispectral tab in the relevant channel and vice versa.
There are Master Gamma, Master Gain and Master Saturation sliders. They would effect the whole image across all channels.
There is a readout that tells you how many images you have stacked in each channel and what the exposure is.
This is a guide to show your relative exposures per channel.
The actual colour channel integration I assume for ease of implementation would be linear. People may have other preferences, but I would think that linear would be the easiest to implement.
I would imagine nothing else would need to change. The software would stack and save each as it does now. I would think this is mainly UI addition and a display change. I don’t think there is much changing in the processing, other than the colour overlays in the display.
Perhaps you may want a Multispectral channel identifier on the Livestack Overview section on the left, just to remind people they are stacking in Multispectral mode, and clear confusion as to why the image is tinted a colour.
Fitting the "multispectral" reference in here was tricky so I did a very quick move around to fit this in on the mockup. I actually think this indicator is probably unnecessary, and it may just be extra work for little advantage.
Overall, this would be a great addition to those of us with filters!
And thanks for writing great software!
