If you think it is worth it, I can run a script and post the script here. Maybe that would eliminate some misunderstandingsâŚ? The problem is that I cannot set in a script âImage pattern Alignmentâ as registration method.
So no, you donât understand.
It is not luck imaging because you cannot have fps at 50ms by only taking picture. Between each pictures, how many time the DSLR is waiting?
And to come back to your issue. I donât understand why you see anything. Just tested it works.
If the algorithm fail to align you should see only 0.
-1 means no calculation.
I can stack the thing without any problem, as long as I select âallâ in the drop down menu in the stacking tab. If I switch to âqualityâ, the big blue âStart stackingâ button becomes gray and unclickable.
I assume that this is because no Quality was computed. In fact, clicking âOpen Frame Listâ in the Sequence tab gives me the list of images, with different XY shifts in the G channel, -1 in the others, and âN/Aâ as quality for the 3 channels.
Plot tab is completely empty.
I see -1 in the R and B channels for all images (but -1 could also mean âshift 1 pixel up/leftâ, this is why I wrote that it is a bit misleading. Same for 0, but I donât see 0âs).
I did the whole thing by hand instead of using the above script, and now I can see the quality in the G channel, and I can select quality in the drop down menu.
I can also add Normalization in stacking without getting errors.
That means that the above script does (or doesnât do?) something that it shouldnât.
Yet, my script is an extract of Siril bundled scriptsâŚ
Unfortunately, the Console does not show the actual commands that siril performs.
Go to the registration tab, select âImage pattern alignmentâ, select big enough region (about 20x the area of the planet with the planet in the center) in the Green channel of the image, press âGo registerâ. The option âRegister all images from sequenceâ is activated.
When Siril ends, no quality is computed, and Plot tab is empty.
There should be something wrong in the above script. Maybe layers should be debayered at conversion time?? This is the big difference when done manually.
I donât understand why the debayer should be made at the conversion stage. Quality computation should also work then. If not, I think Siril bundled scripts should be fixed to avoid this issue, making the debayering at the convert stage.
Yet, there is some strange error message. When I select the quality, I get an error message:
The reference image is not in the selected set of images. To avoid issues, please change it or change the filtering parameters.
This does not make sense to me since the quality is always smaller than 1.000, which is the quality of the reference image.
I used darks, so how should I proceed then? If I donât debayer at conversion stage, then no quality.
Scrips in bundle work because there is no quality selection. If you want to modify the script to add quality, then debayer should be done at conversion stage.
No. I think you are doing wrong something.
You need to debayer after the preprocessing stage. Then, at this step you do the alignment and the quality thing. I do it right now. It works.
Ooops⌠I thought this was done by Siril, since the console says: âBest frame: #139â.
It took me a lot to find how to select the reference image (in Open Frame list). Since this has to be done after registration, should the registration be recomputed afterwards? Or this is what the âRecomputeâ option in the Stacking tab is for?