×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

Announcing StellarSolver 2.0

  • Posts: 2876
  • Thank you received: 809
Right, my goal is to try to identify where in the chain of events an issue might occur that would end up having the Red channel be the one used since now it should use the green channel if it is RGB.

Now I am thinking it would be really useful for stellarsolver to put in the log something like "Using the Green Channel" or "Exporting the Red Channel". I will have to add that, but it won't help now. It would have helped us verify that it is the latest version and that it is really trying to use the green channel.

When you said you build RPM Packages and install them, does that mean you are building from the latest master on Github or from a latest release? I did make a release with the new capability, but it hadn't been released when you first posted. Also note that the changes were in a branch until just recently so the latest master didn't have them either. Maybe it would be good to try again to be sure you have the code in there. But then again, maybe I should add those messages before you do so you can be sure it really is running that code.
1 year 10 months ago #83022

Please Log in or Create an account to join the conversation.

  • Posts: 1009
  • Thank you received: 133
I'm building from github master. I had switched to the colorChannels branch for a while, but the last one was master agai, after you had announced you integrated it there (commit 9d4efa44a01)
And StellarSolverTester (built together with the rest) does show the channel selector.

I'l keep experimenting, but maybe I should put the OSC on a better optics. I'm still not convinced it's not the lens that is producing part of the issue, as the star images do have weird halo/wings that might irritate the routines.
1 year 10 months ago #83023

Please Log in or Create an account to join the conversation.

  • Posts: 2876
  • Thank you received: 809
Gotcha, ok, yes, please keep experimenting.
1 year 10 months ago #83025

Please Log in or Create an account to join the conversation.

  • Posts: 1009
  • Thank you received: 133
Hmm, found something: Settings->FITS->3D Cube. Tooltip says 'if unchecked, only first channel is processed'. And it's unchecked here.... :-P
Unfortunately, I'll not be at the telescope tonight, but I'll try ASAP.
1 year 10 months ago #83034

Please Log in or Create an account to join the conversation.

  • Posts: 219
  • Thank you received: 41
After a long pause for weather and mount problems, yesterday I was able to test Ekos 3.5.9 with its "green by default" handling on Stellarsolve routines. I can confirm that it works as good as expected on my OSC images. Now the three channels has a comparable amount of focus and I'm able to keep seeing faint stars on the B channel (now I've to fix the guiding problem with the HEQ5... but, hey, one problem after the other ;))

Thank you very much Rob!

1 year 9 months ago #83431
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 219
  • Thank you received: 41
Very bad weather until now.... but it's not clear to me what's happening here. Yesterday I return to the "R it's the best focus channel" scenario without changing anything on my Ekos configuration.
1 year 9 months ago #83629

Please Log in or Create an account to join the conversation.

  • Posts: 1009
  • Thank you received: 133
Did you do an update of kstars? I've seen that change some settings (like toolbar visibility) now and then, though I don't really understand how that happens (unless options get renamed?).

I also only had one run so far with the new scheme, and focus seemed clearly better than before. I'd like to get some more data, including focus debug messages....
1 year 9 months ago #83630

Please Log in or Create an account to join the conversation.

  • Posts: 219
  • Thank you received: 41
Really I don't understand what's happening here. I thought that the changes to use G channel by default was in place, but after a long pause for weather and health problems I've been again under stars with bad results.

I'm using 64bits Stellarmate with KStars 3.6.0 (the latest stable one).
On FITS viewer, I've configured 3D Cube checked and WCS checked (this latest is not needed).
On the focus module, I've configured detection SEP and SEP profile to the standard 1-Focus-Default. The algorithm used is Linear 1 Pass (but I think that doesn't matter)

With this configuration I expect that the focus algorithm will use the green channel or the average of them (I don't remember which one was the final decision). But what I've found is that it keeps using the R channel with bad results for the other two. This is the top left corner. You can see how the spikes of the bright star begins to duplicate and separate, also, this will show coma on G and B channels:


Red


Green


Blue

So my question is: are there any way to influence which channel is used that I'm not aware of if?. Are there any information logged out about the kind of image StellarSolver is using here?

Thanks a lot
1 year 5 months ago #86673
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 219
  • Thank you received: 41
Rob, do you have any clue about this? How can I be sure that I'm using the channel I hope to be using?
1 year 5 months ago #86845

Please Log in or Create an account to join the conversation.

  • Posts: 1009
  • Thank you received: 133
Two things: Last time I tried focusing with my OSC camera I also got bad results, so you're not alone.
Second, I had the impression the stellarsolver lib does what it's supposed to, i.e., when getting a color cube image, use the green channel. If that doesn't work, the culprit (IMO) must be EKOS. Either it debayers but then sends only the first channel, or it sends the raw file (but there green should definitely dominate).
1 year 5 months ago #86851

Please Log in or Create an account to join the conversation.

  • Posts: 219
  • Thank you received: 41
Thanks Peter to confirm my bad results. Now the question is, how can we improve the results with OSC cameras? Any clue about how to proceed?

Regards
1 year 5 months ago #86854

Please Log in or Create an account to join the conversation.

  • Posts: 1009
  • Thank you received: 133
Hmm, a quick solution could be to put the camera into a grayscale mode (LUMA or monobin) for running the focus routine. Not sure though if that would work for automatic refocus, though.
1 year 5 months ago #86856

Please Log in or Create an account to join the conversation.

Time to create page: 0.960 seconds