Thanks and a question about color separation in stars and how to fix.

Jeff

I am down to the final workflows in the Fundamentals videos and I have to say that I learned a lot although I have been using PixInsight for six years. I am especially thankful for your explanation of the color mask script.  I have been struggling with that one for years.  Now it makes sense and I can effectively use it. 

I do have a question about how to improve stars that start out as yellow on one side and are blue on the other.  I am not sure if I can post a picture here, but here is a link to one of my full size images: https://astrob.in/full/seeob9/D/?real=&nc=jeff2011

I think this has to do with some kind of tilt but CCDInspector shows a very flat field for my camera and the scope is an FSQ106EDX-IV.   I am still working out the root cause but until then, I was wondering if there was some kind of processing I could do.  I currently do my stretches by using a slight ArcSinH followed by masked stretch and finally HT.  I have tried HSV repair but that just exacerbates the issue.

BlockHead

Hi Jeff,

Yeah, I don't have an easy answer. That looks like the kind of optical distortion that I suspect is difficult to get right when you register the images. I assume you did play with turning on the distortion parameter in StarAlignment with the thin plates algorithm. 

The direction of the color offset changes as you go around the field. 

Another FSQ user had a similar problem. A tilted image plane I think is a good guess. 

There was a thread on the official forum about this. (This wasn't you was it?)

All I can say, is that is looks like an optical aberration. I wonder if, and this could be goofy, slight changes in focus (due to say non-parfocal filters) could cause this? I really don't know. 

You can attack various colors with masks... but it isn't a good solution. Much easier would be to approach from the equipment side I think. 

Did you ever measure the flatness of the field in each color?? Lol. 

I

-the Blockhead

Jeff

Thanks for your response Adam.  I forgot to mention this is an OSC camera but I have seen this on my mono camera also but I know that the mono has some tilt. I did try the distortion parameter and various interpolation methods to no avail. It is an optical problem because I can see it in a single debayered sub.

By official forum do you mean Cloudy Nights?  No that was not me. But I did see a few other threads there to no resolution that worked for me. 

I am using the stock Tak focuser and I have heard of issues with that and the camera angle adjuster.  It could also be focus. I do refocus on a 0.6 degree Celsius temperature change but perhaps may be a problem somewhere with it or with the telescope itself.  I did buy it brand new less than two years ago and picked it up directly from Takahashi America which is not far from me and they supposedly checked it all out.

I measured the flatness of my field while still in a bayered state.  I did not think about debayering and then measuring the flatness and also of each color channel.

BlockHead

No..the official PixInsight Forum. https://pixinsight.com/forum/index.php

No... definitely do not bother to measure for each channel- my concern would only be valid for a filter wheel situation. Unfortunately... this is probably pointing towards optics. 

-the Blockhead

Jeff

I have since looked at a lot of images on AstroBin and found out that many images taken with the FSQ106 show the same thing to varying degrees.  Even Images of the Day and Top Pics.   A guy on CN  put a $3K night crawler focuser on one and he flattened it with a tilt adjuster and was still getting it to a lesser degree.  I am beginning to think this is just part of the Petzval design of the Tak and I will need to keep from color  saturating my stars too much and I can't use HSV repair as that makes it really stand out. 

BlockHead

Hi Jeff,

I recently found something you might want to try. 

If you take your images and use ImageSolver... and then use the AlignbyCoordinates to undistort the images... and THEN use SA to align... maybe something magical happens?

It is worth an experiment I think.

-the Blockhead