good to hear that the tweaks to the collimation tool are coming in useful - the final image certainly looks good from edge to edge and corner to corner
I am struggling to see with the second image in the post above - are there any contours visible in it? Do I need to go further in making the contour spacing narrower if there is only a small range in the FWHM?
Two parts of the same contour are just visible at the top left and bottom right but are not annotated. By analogy with other images I am pretty sure that it would be the FWHM 3.25 contour....hardly any greater than the 3.19 marked near the centre point of the eliipse
The tool certainly was useful when used in a general way and after a bit of practice. The exact pattern and apparent centre point shifted somewhat with shits of the middle slider of the histogram and alterations in the number of stars detected so I learned to make comparisons at roughly equivalent values. Not helped by ever changing sky conditions (Oh for one steady night to do it all properly)
Not sure that you need to go any further with fine contouring because even with the current settings it is contouring at 0.025 differences in FWHM which is really pretty small. I though that once I had reached the point of any apparent pattern recognition being being weak with no or few contours visible then the field must by then be pretty flat?
I think that the tool is good to go as it stands.
Ideally, and as previously discussed it is not so good on individual frames (too variable - at least currently) but pretty good used on autostacks of say half a dozen or more frames accumulated over 5 minutes or so (in my particular case). So the ideal would be to have a way of comparing the last autostack with the current one in order to be able to use it in real time to drive adjustments.
Have you tried setting the 'Averaging' settings to 'Average over 10 frames' on the left hand side. That's not quite the same as stacking then contouring the stack - instead it is collecting star data from each of the 10 frames and then contouring from all of that data. It should keep things more stable, but be aware that it is a rolling average, so if you change something then it will take 10 frames for all of the old (pre change) data to be flushed out and the new data to take full effect.
Thanks for that tip. I hadn't spotted the average button at all and that could potentially help a lot for using the tool to better make adjustments in real time. So far I have been autostacking for 4 min - then opening a second instance of SC to analyze - then making adjustments and trying another autostack etc. which is a bit clumsy.
