AUS-2 and SPA-1 Takahashi FSQ-106 Telecopes

[Mark Mayer]

The AUS-2 and SPA-1 Takahashi FSQ-106 telescopes both produce images with distortion on the left and right sides of the image. Do both of these telescopes have field flatteners? 

The screen shots below are for the AUS-2 IC4812 mosaic panel 1 and generated using the PixInsight AbberationInspector script for a single 300 s light, and the master light, both with the luminancefilter. 

The before and after images show that BlurXTerminator partially corrects for the distortion, but that especially for the master, the stars have halos that impact image quality.

Is this an intrinsic limitation of the Takahashi FSQ-106 combined with the larger image sensor of the QHY600M, or is their performance sub optimal?

[ERNESTO GUIDO]

Hi Mark, 

What you see is an expected aberration, which was less visible with the CCD because the resolution of the images was about 3 times lower that the resolution we have now with CMOS.

These aberrations are seen only in the corners as we push beyond the optical performance of the Takahashi, which has an image circle of less than 40mm (smaller than the sensor diagonal).. There is also some residual tilt due to variable flexures of the optical train that causes variable distortions. The operators in Australia & Spain did everything possible to minimize it. These distortions have always there but were not visible earlier due to the larger pixel size of the CCDs. They are also only limited to the corners or sides.

We could crop the image, but we think leaving this decision to the user is better.

Clear Skies,
Ernesto

[Mark Mayer]

Hi Ernesto,

Thanks for explaining - makes sense. What puzzles me is the specs that Takahashi publish for the FSQ-106EDX4 

https://takahashiamerica.com/products/takahashi-fsq-106edx4-quadruplet-refractor-telescope

Where they state that the telescope has "A flat field without the need for an additional flattener, across a large 88 mm image circle"

Mark

[Marco Rocchetto]

Hi @Mark Mayer, note that we also use a 0.73x focal reducer, which increases the field of view but decreases the image circle in which the image remains flat.

[Mark Mayer]

Hi Marco,

I just checked my plate solving results for recent TL data sets I've processed for SPA1 and AUS2 and found the following: SPA1 plate solve focal length 385 mm; AUS2 plate solve focal length 531 mm. So it seems that only SPA1 has a reducer and AUS2 has the native focal length. The example pictures I posted at the start of this topic were from the recent AUS-2 IC4812 mosaic panel 1 generated using the PixInsight AbberationInspector script. I do understand that its not possible to easily access the equipment in a remote observatory, but currently I'm wondering if the AUS2 Tak, which does not have a focal reducer, is not performing to spec? 

Mark

[Craig Schultz]

Mark, are you finding it difficult to merge these mosaics due to the star abberitions in the corners and frame edges?  I was thinking of trying some mosaics, but this could prove extremely difficult if the stars don't align in the merge regions.

 

[Alexander Curry]

Adam covers dealing with the overlap in this webinar:

https://telescope.live/tutorials/mosaic-mastery-delving-deep-pixinsights-photometric-tool

[Mark Mayer]

On 11/17/2023 at 11:23 AM, Craig Schultz said:

Mark, are you finding it difficult to merge these mosaics due to the star abberitions in the corners and frame edges?  I was thinking of trying some mosaics, but this could prove extremely difficult if the stars don't align in the merge regions.

 

Merging the panels is not a problem, so do not be put off, but the distortions do propagate to the overlap regions, instead of being limited to the corner, as occurs in a single panel. Unless you look at high magnification, the images will look fine. The reason I've asked about this, and have yet to get an unambiguous answer from TL, is the simple question of whether the two TAKs performing as expected, or is their something wrong in alignment of the imaging train. The TAKs are $7k Petzval refractors which should have a flat field, but they are very sensitive to decentering/miscollimation of the four lenses.

[Juergen Beisser]

I am afraid that I have to agree to Mark that the FSQ-106 could perform better. Judging from the star shapes, I don't think that the telescope is out of collimation, but I strongly suspect tilt. This observation is supported by an ASTAP analysis, which clearly shows strong tilt. I don't know whether the QHY600 allows any tilt adjustments or whether a third party tilt plate should be installed in order to improve the image quality. The FSQ-106 ED as well as my own FSQ-85 ED of course show some distortion of stars at the corners (less with the 0.73x reducer!), but the amount of the AUS-2 camera or sensor tilt is inacceptable in my opinion. Please see the ASTAP output of one of my SMC images taken with AUS-2.

Juergen

 

[ERNESTO GUIDO]

Hi Juergen

As I wrote in this very same thread on October 20:

"There is also some residual tilt due to variable flexures of the optical train that causes variable distortions. The operators in Australia & Spain did everything possible to minimize it."

 

A couple of times a year we go to the observatories in person to monitor equipment more closely, bring improvements, etc. As soon as possible we will try to see firsthand if better results can be achieved and how. 

Clear Skies,

Ernesto

[Mark Mayer]

I just tried the newly released BlurXTerminator AI version 4 on these problematic images and the result is truly amazing.

Essentially all of the abberations are corrected. With this new tool, data from AUS-2 and SPA-1 with the QHY 600M cameras, is as good as it gets for earth based refractors. 

[ERNESTO GUIDO]

Thanks for this update Mark, It looks really interesting!

 

Clear SKies,

Ernesto