Image Inspection Show Distortion
Stacking program, astrometric (plate) solver, fits file viewer
Brought to you by:
han59
Menu
▾
▴
Hi. I'd like to obtain the numerical data behind the 'Show distortion' image.
1. I'm not seeing the 'Data to clipboard' box located to the right of the Extra stars box indicated in the Help.
2. No idea what the checkbox to the right of 'Show distortion' is supposed to do and I didn't see (or missed) reference to it in the help.
I've attached a screenshot that might help illustrate the above issues. I'm running ASTAP 2023.12.04b
Jesse
Hi Jesse,
Once you have clicked on 'Show distortion' button with the check-mark checked then you can just paste the data into a spreadsheet or word processor.
If it is useful, I doubt it but I made it on request of one of the users.
Han
That. Is. Exactly. What I want :)
I've been casually involved with the Jupiter deflection effort [https://skyandtelescope.org/get-involved/pro-am-collaboration/how-to-measure-jupiters-deflection-of-starlight/] and my thinking is that I can use this data to tweak (if necessary) the measured centroid position of off-axis reference stars which may be at different locations on the sensor on different nights. -Jesse
Hello Jesse,
That is an interesting experiment. Only I'm afraid the deflection of 0.0085″ is too small to measure optically. The accuracy of ASTAP is a fraction of a pixel. So lets say 0,25 pixel. You will need huge oversampling and superb seeing but likely will never achieve 0.0085″.
Han
I have tried to measure the difference in refraction for red and blue light of the same star for low altitudes but failed to do so. One user has achieved it but the effect is larger then the deflection by Jupiter.
Han
I'm thinking you might be right about the deflection but it's worth a go. Their forum is at: https://groups.io/g/JupiterDeflection
So what is the accuracy of ASTAP? The deflection spreadsheet has "measured" pixel values to the third decimal place. You're saying that's not real!? :) I think for the deflection measurement it's supposed to boil down to obtaining enough samples to get a really, really good estimate of the mean deflected position. (E.g. see:
https://en.wikipedia.org/wiki/Sample_size_determination#Estimation_of_a_mean )
Between us girls this is problematic in part because sigma isn't constant across a set of observations due to seeing, though maybe it's 'constant' enough for a large enough sample size, or you establish an upper bound?
Our group used 20s H-alpha images to reduce any wavelength dependent issues. That being said, regarding refraction, I've seen obvious color refraction issues in planetary images taken at 'low' altitudes and there's a real market for atmospheric dispersion correctors.
Jesse
If have compared the center position of solved image with other solvers like Nova.astrometry.net, local and online, Pinpoint, PlateSolve2/3. The solvers are in all reporting the same position within a fraction of a pixel. There is even difference between the local Astrometry.net and online version. Most likley due to the different star database used.
What I would suggest is to test your setup. Just measure position of one star in 20 or more images over time. What is the standard deviation of the test star position? The absolute accuracy is not so important only the stability of the reading. This would indicate what is achievable. The image should be pretty oversampled for maximum accuracy. But not too much otherwise the star detection will fail.
Next would be with Jupiter in the image. One of the problem is that Jupiter will remove the reading of one or more stars due to it brightness. Does this influence the stability of the test star?
Use the popup menu "Star info to clipboard" for the positions. See attached screenshot:
This is the readout. Six decimals in the position. Can you get a stable reading and what is the sigma?
Han
If you get promising reading we could consider to improve it and look to the factors influencing the readings.
Thanks. I've got data from before and during. I've been using MaxIm, PS3 and SExtractor for positions for the sequence from each night. I've got ASTAP but haven't really used it, though I'll try to become a more familiar as I'm trying to use multiple programs. The Group noted that distortion was something to consider and I vaguely recalled ASTAP measured it.
The other issue is that SEx and PS3 weren't happy measuring the target either with the bright Jupiter background or when I removed the background using a weird-ass morphological background removal scheme I got from the internet (that I thought worked rather well with maybe a few artifacts, but still...)