RGB alignment

Ken82

Well-known member
Im hoping someone could help me out with this issue-

Basically after RGB combination using channel combination i always get strangely coloured stars. What should be a red star has a purple edge (this is always to the right hand side). I know it could possibly be CA but this is an fsq 106 so i'd expect it to be free of this issue and and that its likely something im doing at this stage.

My process after calibration is star alignment, Dynamic crop, DBE, Linear fit

Ive been told by the vendor that it could be alignment of the images ?

Red https://drive.google.com/file/d/1At2LiW0ml_VhA-rLwFrJw3IreV92XisQ/view?usp=sharing

Green https://drive.google.com/file/d/1T--4-xk3c09GXQEz6ooFtUF-f3pg-TFF/view?usp=sharing

Blue https://drive.google.com/file/d/1vx8qYX-e-0f1keJzMgk4cXplnjpw-3zI/view?usp=sharing

STAR ISSUE.jpg
 
what are your observing conditions like? did you cull out bad subexposures (high fwhm, high eccentricity, etc.)?

rob
 
Conditions were normal I’d say with guiding around 0.5” to 0.7” max. Target was also at zenith.

looking through subs FWHM is fairly stable as it auto focuses every 45mins.
 
In order to assess this, you need to look at the position angles of the fringes around the field. There is a general "rotation" of color from what it looks like to me when I inspected your RGB image. So... I disagree with vendor...this appears like optical aberrations (chromatic) to me. It is very small however... and I wouldn't be unhappy with this (wow..I have seen much much worse!). This is not an alignment issue.
-adam
 
these scopes are generally excellent but i have heard some horror stories about them. all the images i ever took thru an fsq106 were absolutely razor sharp (admittedly from dark and steady skies) which is why i asked about the observing conditions. if someone asked me to guess what kind of OTA took those images, i would not guess fsq. the stars seem quite bloated. but again, that could be caused by forces external to the optical system.

there is a script in PI that mike schuster wrote called WavefrontEstimator that will test the on-axis performance of your scope. i'd expect a strehl ratio of .99 from a properly collimated fsq106. if you get something different from that script, it is at least data that you can feed back to takahashi as evidence that something is wrong with the telescope.

rob
 
I don't know Rob... the data looks about right to me. This is an F/5 system with *tiny* 3.75um pixels? I am guessing this is around 1-1.5 arcsecond pixels. If the seeing was 2"... 2-3 pixel stars. According to DynamicPSF he has 2.4 FWHM stars... so things seem to make sense? This aberration is very tiny... the image above is zoomed in like 800% or something. You certainly could not see this by eye. It is difficult for me to see the effect until I zoom in around 200-300% .

-adam
 
Thanks for the feedback I’ll try that out later Rob.

It is 3.75um pixels.

I got the same issue with M106. It becomes a lot more noticeable after processing meaning I cannot amend curves.

 
Hi Ken.

I looked through both sets of images that you have posted and what sticks out most prominently is that each of the channel images exhibit some degree of tilt with star elongation in different amounts and in different directions in each of the channels, which I suspect is the main issue when the individual channels are combined.

I also ran both sets through CD inspector and that pretty much confirmed that there is variable tilt present between the individual filter channels.
The tilt is not that great but it is present in all the images to a greater or lesser degree.

I have never owned or seen a TAK 106 but having read the forums over many years a common theme crops up that the camera couplers/image rotator on the TAK106 is prone to some degree of looseness allowing the camera to sag and introduce tilt into the image path, I stress though that this is just what I have read, no idea if it is the fact.

I would begin your hunt there and check for movement in the camera as the scope angle changes during a session. It doesn't take much physical movement of the camera to introduce a fairly large tilt at the image plane.

Comparing the best stars in the blue channel images there does seem to be a little fuzziness on the right side of the star so to eliminate the TAK as the source take a series of blue channel images and rotate the camera and filter wheel 180 degrees half way through, say half a dozen subs with normal orientation and half a dozen with the camera rotated 180 degree.

Then look at the individual images and assess the star profile.

If the fuzziness stays on the same side of the star in both camera orientations then look at the filter wheel and make sure the blue filter is sitting square in the wheel, a little tilt in the filter will distort the star profile.
If the fuzziness is on one side of the star in the normal camera orientation and on the opposite side of the star in the rotated camera position then the optics may be to blame.

HTH

William.
 
I don't know Rob... the data looks about right to me. This is an F/5 system with *tiny* 3.75um pixels? I am guessing this is around 1-1.5 arcsecond pixels. If the seeing was 2"... 2-3 pixel stars. According to DynamicPSF he has 2.4 FWHM stars... so things seem to make sense? This aberration is very tiny... the image above is zoomed in like 800% or something. You certainly could not see this by eye. It is difficult for me to see the effect until I zoom in around 200-300% .

-adam

i don't disagree, i did the same pixel scale calc and that's why i asked about the seeing. but the aberrations are very unusual for a scope of this quality.

i downloaded the images and blinked them and i could definitely see the stars shifting across the channels so it just seems to me that something is off. that's why i asked about the bad frames - i figured the problem might be caused by high-eccentricity subframes as well. anyway it seems like an investigation of the optics is not a bad idea as i have heard a few people complain about the fsq - when they go wrong they go pretty badly wrong.

i think william's suggestions are good - tilt is a good avenue to explore as well.
 
i don't disagree, i did the same pixel scale calc and that's why i asked about the seeing. but the aberrations are very unusual for a scope of this quality.

i downloaded the images and blinked them and i could definitely see the stars shifting across the channels so it just seems to me that something is off. that's why i asked about the bad frames - i figured the problem might be caused by high-eccentricity subframes as well. anyway it seems like an investigation of the optics is not a bad idea as i have heard a few people complain about the fsq - when they go wrong they go pretty badly wrong.

i think william's suggestions are good - tilt is a good avenue to explore as well.

Yea thanks both for the feedback. The vendor is sending it back to takahashi Europe for evaluation. Hopefully that can then resolve the issue.
 
Just to add this is a crop of my M106 image data above. The issue is much more prominent here
5D41AE3D-6A08-4D68-9BFB-C72354F2D1B8.jpeg
 
A very interesting point from a friend. I did apply distortion when aligning the RGB but is there a method to increase distortion with scaling or better align -

I think this is alignment issue. In fact, I think it is very specific kind of issue that I'll try to explain.
First let's see if we can see what is happening in each corner. I took three corners to analyze. In first corner, we have this:
Stack-1.gif.fcc6db29b65f36444a6c3d4925b9d43f.gif
This is animation of R, G and B frames, Star positions look like pretty good match, there is no much shift between them, and if we measure centroid on a single star - this is what we get:
Screenshot_4.jpg.6c105fa4886bbe5568392220a29733de.jpg
Star center is the same to first decimal place (roughly - there is difference in 0.1 pixel at most between star centers), so it is very good match.
Now let's see what happens in opposite corner:
Stack-2.gif.1058ce566e9fe1a16edaea4cb9c25a48.gif
Don't know if you can see this but here there is a bit more "wobble" in star positions between R, G and B. To actually measure it - let's do the same and select star and do centroid:
Screenshot_5.jpg.3cdd412629c6132ecd05aeef2deb68a9.jpg
Ok, now we start to see that error in position is no longer 0.1 - it is larger, in fact, between R and G it is about 0.3 in X and almost the same in Y direction - total position being offset by almost 0.5px. Red and Green are closer and Blue is far from Red.
Now we need to see the third corner and see what the situation is there like, again, animation:
Stack-3.gif.6de52b659b971fed4ddf6ce47b5190b7.gif
Here I see quite a bit of wobble in vertical direction. Let's again check actual numbers:
Screenshot_6.jpg.65743068f038e9afead56b298f5ef62e.jpg
Interestingly - error in X is again 0.1 but error in Y is again being 0.3.
Do we see a pattern here? First corner both axis 0.1 error, second, diagonal corner both errors 0.3, third corner Y error 0.3 and X error 0.1. I wonder if fourth corner will show X error to be 0.3 and Y error to be around 0.1?
image.png.2f65ec5066f63f175480c088accc84ff.png
Interesting - this time error is not 0.3, it is 0.7 and it is in X axis. Y axis has same 0.1 error.
This means that Blue channel is more zoomed in, but how can this be?
We are dealing here with large sensor and relatively short focal length - wide field image.
Two things happen when we have wide field image. Projection distortion starts to creep in.
Extreme example of this comes from wide angle lens:
image.png.cd60d897860f04354ff62c3c78bfc4b5.png
Straight lines are no longer straight lines in the image because of this type of distortion. Since FOV is only a few degrees - it is not really visible by eye but can become a problem if there is slight misalignment between images and you try to align them without first correcting for this distortion.
Second thing that happens to produce different level of magnification in wider field image is atmospheric distortion.
Atmosphere bends light. It particularly bends blue light (shorter wavelengths). This effect is more evident close to horizon than up high towards zenith.
If blue channel was shot when target was lower in the sky - then "bottom" part of the frame could be influenced more by this bending of the light - thus creating "zoom" effect for blue channel.
I guess that this could be fixed by applying different alignment model - one that also allows scaling rather than pure rigid transform. If wide angle distortion is dominant - it should be corrected (like lens distortion correction).
 
I think this is alignment issue.

I disagree. There are no alignment issues with these images. To demonstrate this statement, here are PSF measurements performed at the four corners of the green and registered blue channel images with your M106 data (click the links to open full-size screenshots):

Top-right corner
Bottom-right corner
Bottom-left corner
Top-left corner

Image registration has been applied with distortion corrections enabled and a thin-plate based registration model; see a screenshot of the tool with relevant parameters attached to this post. As you can see, registration is accurate up to about 0.05 px for the best sampled stars, and about 0.1 px for the dimmest ones. These are good results, considering the scale of the image and the deformed star profiles.

Besides the chromatic aberration issue that you are describing here, there are other problems in these frames that are causing important deformations of star images at the corners, as can be seen in the screenshots above, or by inspecting the frames. I have computed an astrometric solution for the green channel image using our ImageSolver script, with the following result:

Code:
Image Plate Solver script version 5.4.5
===============================================================================
Referentiation matrix (world[ra,dec] = matrix * image[x,y]):
 +7.22648545e-06  +4.05954574e-04  -1.33116528e+00
 -4.05915869e-04  +7.21030722e-06  +1.92055027e+00
WCS transformation ....... Thin plate spline
Control points ........... 2182
Spline lengths ........... l:226 b:72 X:74 Y:224
Projection ............... Gnomonic
Projection origin ........ [4788.132626 3193.864505] px -> [RA: 12 18 49.497  Dec: +47 16 01.25]
Resolution ............... 1.462 arcsec/px
Rotation ................. -91.021 deg
Observation start time ... 2020-03-25 23:33:42 UTC
Geodetic coordinates .....   2 09 10 W  53 03 36 N
Focal distance ........... 530.62 mm
Pixel size ............... 3.76 um
Field of view ............ 3d 53' 16.3" x 2d 35' 36.7"
Image center ............. RA: 12 18 49.496  Dec: +47 16 01.26  ex: +0.000416 px  ey: -0.002328 px
Image bounds:
   top-left .............. RA: 12 10 41.257  Dec: +49 10 06.65  ex: -0.002782 px  ey: +0.006721 px
   top-right ............. RA: 12 11 39.423  Dec: +45 17 16.87  ex: +0.014623 px  ey: -0.023343 px
   bottom-left ........... RA: 12 26 32.885  Dec: +49 12 59.70  ex: -0.003614 px  ey: +0.010470 px
   bottom-right .......... RA: 12 26 23.614  Dec: +45 19 55.25  ex: +0.002168 px  ey: -0.010806 px

As can be deduced from the reprojection errors, the image is mostly flat. Only the upper right corner has significant field distortion. I have attached the generated distortion map, where this can be evaluated graphically. So geometric distortions caused by the telescope optics are not causing deformed star images.

You'll have to find the origin of these problems with this telescope, but definitely our image registration tools are not causing them.
 

Attachments

  • SA.png
    SA.png
    140.7 KB · Views: 61
  • dmap.png
    675.4 KB · Views: 73
Back
Top