Great!! one bug less in INDI but ofcourse there is clouds tonight, I'll test as soon as possible!
Thank you!!

Long time since I saw any stars at all, including our local star but got a chance to test last night
The OAGT checkbox is nice
But still can't resolve without changing parameters, now I also noticed that trying to use the local resolver stopped working, but I have to test that a couple of more times to be shure.
And "Rapid guide" is still not working, but maybe there is no change done yet.. ?

Br
/Markku

The FOV is not calculated correctly now or what? Also, what do you mean by rapid guide not working, can you be more specific?

Yes, the FOV is correct now.
The problem with "Rapid Guide" is that when I have selected a star to guide on and try to use "Rapid guide" the green box jumps away from the selected star and often finds a 'white' pixel instead, very few times it stays at the selected star, but normal guiding works fine but takes quite a log time to get the guidepicture over the internet.

Perhaps Peter can shed light on this

Hi, try defocus guider a little bit. This should make real stars look better for used algorithm Peter

IMHO, the problem may also arise due to "hot pixels" which cannot be removed by defocusing.

Algorithm is matching airy disks 7x7 pixels. It works perfect for long focal lengths and/or small pixels and worse for wider FOVs and/or bigger pixels but defocusing should help to match real stars instead of hotpixels. Peter

I think that there is no problem with the algorithm because it is possible to guide on the worst possible scenario, take a look at my worst guidingstar ever.



I've tried guidepics from the worst to perfect focus.
I've been thinking about the problem for a while now and here is my two cents..
For me it seems (maybe I'm wrong, in that case please correct me.) like when "Rapid guide" is selected it starts to search for a guidestar from left upper corner to right and so on, finds a white pixel and stays there, and that I understand because it'snot good to loose the star.
I think that it would be better to pass the area (cordinates) of the green box for the starsearch to find the guidestar to get a smaller area to search on ?
My point in this is that the user have already found a suitable star and there is no need to search the whole picture to find one.

I'm not shure at all that I'm correct in my thinking but I don't understand C++ good enough to test anything myself

No, it is searching for the best airy disk match in rectangle +/- 20 pixels from the last star centroid position. It may lock on hotpixel only if there is no better match and it may happen on wide-field perfectly focused image (star = single pixel). That's why defocusing should help.

Last night the sky cleared up a bit so I did a small video about the behavior..
It's really hard to get get round stars with my oag, the more I defocus the more they look like ovals.. but take a look anyway.

Well, I just tried with Rapid Guide and it's working. Your image is very noisy, I wonder why?!! I have the exact same CCD and this is what I get (just a few minutes ago recorded):



I am doing 600 seconds exposures, so I am now waiting for 1st image to download to judge the quality of the guiding, the drift graphics looks OK but a bit jumpy.