WavefrontEstimator: the defocused image diameter is too large...

sbasa

Member
Jan 13, 2016
11
0
Hello to all

In order to better characterize a 1.3m diameter telescope (focal=9398mm), we would like to use the WavefrontEstimator from Pixinsight. For this, we followed the recommendations by defocusing (intra and extra focale) and produced nice donuts (the test camera is a SBIG ST8300 with a pixel of 5.4 um). Despite our best efforts, we are unable to process the images and have the following error every time: "the defocused image diameter is too large.".

If we make a weaker defocus, it works very well, which makes us think that there is a parameter badly adapted to our telescope. Does anyone have an idea of the problem?

Thank you very much for your help.

Best regards
 

sbasa

Member
Jan 13, 2016
11
0
I did what you advised me in December: I changed the maximum diameter in the model.
 

sbasa

Member
Jan 13, 2016
11
0
It is about 350 pixels. I have also changed the max diameter (320 pixels) which is hardcoded. But it doesn't solve the issue.
 

sbasa

Member
Jan 13, 2016
11
0
250-300 pixels works correctly. However, 360-380 (the nominal diameter suggested ) fails.

I have changed the this.maximumDefocusDiameter in MainModel.js (a hardcoded limit), but it has changed nothing.
 

fredvanner

Well-known member
Apr 17, 2019
1,436
214
70
Wells, Somerset, UK
Unlike the aperture (which is just a parameter in the calculation), the diameter of the images determines the size and resolution of the image processing. I am not entirely surprised that the script can't handle changing this, but I haven't tried to work out why.
 

sbasa

Member
Jan 13, 2016
11
0
It's actually quite frustrating. There's no reason it shouldn't work. I will continue to investigate.
 

fredvanner

Well-known member
Apr 17, 2019
1,436
214
70
Wells, Somerset, UK
Since you have already modified it, you know that the WavefrontEstimator function is spread across 22 script files. I think it will be non-trivial to work out what is needed to support an increase in processed image size! :unsure:
 

fredvanner

Well-known member
Apr 17, 2019
1,436
214
70
Wells, Somerset, UK
At a first pass through the relevant code, I can't see why changing the built-in max-min radii shouldn't work (if you look in the change log you'll see that one of the most recent changes - back in 2016 - was an update to give the current 32 / 320 pix values). It might be worth instrumenting the script code with some console output of the actual calculated radii, just as a check.
 

sbasa

Member
Jan 13, 2016
11
0
I've come to the same analysis as you. I will do it on Monday. Now, a small weekend is welcome.