Had a go tonight, it was a lot more stable. this time it only crash twice out of the hour, so much better. When I moved around the sky, it crashed twice, but the fits images are now good.
Thanks, that fixes the crash, it seems the wcs thread runs for quite a while after the image is already on screen and if I close it immediately, the window won't close until 10s or so has passed. I dug the code a bit and it seems it shouldn't be necessary to precalculate RA/DEC for every pixel on load but to calculate the few that are actually needed on the fly. There doesn't seem to be that many places which actually use the data (as a test I removed the function to get the array and checked where the compiler gave errors) so I'll see if I can get them fixed cleanly and make a pull request for that.
WCS is used in quite a few places in KStars, but you can live without them. What I've done is that in Resource Limited mode, WCS is not processed. You can still use the PAA tool, but you won't see the Equatorial grid lines. I'll commit shortly.
I got all other places fixed to work without the array quite easily, the only more complicated one is to find RA/DEC bounds for equatorial lines, especially near poles where just checking image corners doesn't work, but there should be an analytical solution instead of just brute forcing all image pixels. I also did some optimizations that WCS info is read only once and stored in the FITSData object instead of being reloaded every time pixelToWCS() for example is called.
I have 2 QHy cameras - a QHY9/filter wheel and a QHY5L2 as a guide cam on a seperate 60/280 guide scope. I want to use the QHY5L2/guidescope combo to do the polar alignment. as it meets the wide field requirement. If I select the QHY5L2 as the camera to do the polar alignment, the "Start" button stays greyed out. It only becomes active if the QHY9 is selected - but this is not useable as it is on a scope with a longer focal length.
Am I doing something wrong?
Even if I disconnect the QHY9, and only connect the QHY5L2, the button stays greyed out
sorry Jasem, I assumed that OAG referred to an Off Axis Guider not a seperate Guide Telescope So I never actually hovered over it and saw the tool tip.....
Guess I'm a bit literal....always thought I had a bit of Aspergers Syndrome !
Hi Jasem, I've been working on an algorithm to align the scope with the polar axis using the guiding software.
If you are interested it could incorporated with your software. The general idea is that rather than using platesolving to determine the centre of rotation, I use the guide software. If the guider can determine a star position to within a faction of a pixel it should be possible to calculate the arc with considerably less rotation than 15 degrees and faster than plate solving. Testing so far is encouraging. It is also possible to automatically slew in dec to align with the polar and report on cone error. Once centred, plate solving can be used to determine the alt and az adjustments to align the pole to the centre of rotation.
If you are interested I'm happy to share my design. If not that's cool and I'll just keep going ahead for my own curiosity
The following user(s) said Thank You: Jasem Mutlaq
Hey Kengs, thanks for the suggestions. Let's see how the current method fares (still in Beta). Your algorithm can be an additional method to do the polar alignment offered to users. Thanks for your work!