GNU Astronomy Utilities - Bugs: bug #65106, Out put of the custom-profile in...
You are not allowed to post comments on this tracker with your current authentication level.
bug #65106: Out put of the custom-profile in MakeProfile has distortion
Submitter: | Sepideh Eskandarlou <sepideh> | ||
Submitted: | Tue 02 Jan 2024 02:56:03 PM UTC | ||
Category: | MakeProfiles | Severity: | 3 - Normal |
Item Group: | Output not reasonable | Status: | Fixed |
Privacy: | Public | Assigned to: | makhlaghi |
Open/Closed: | Closed |
Tue 02 Jan 2024 04:20:10 PM UTC, comment #2: |
Mohammad Akhlaghi <makhlaghi> |
Tue 02 Jan 2024 02:57:25 PM UTC, comment #1: The output of the below command is not reasonable. Because when I am running the following command I was expecting the out put be a smooth profile. But in some parts it have distortion.
|
Sepideh Eskandarlou <sepideh> |
Tue 02 Jan 2024 02:56:03 PM UTC, original submission:
|
Sepideh Eskandarlou <sepideh> |
Depends on the following items: None found
Items that depend on this one: None found
There are 0 votes so far. Votes easily highlight which items people would like to see resolved in priority, independently of the priority of the item set by tracker managers.
Follow 6 latest changes.
Date | Changed by | Updated Field | Previous Value | => | Replaced by |
---|---|---|---|---|---|
2024-01-02 | makhlaghi | Attached File | - | Added oversampling-effect.png, #55509 | |
Status | None | Fixed | |||
Assigned to | None | makhlaghi | |||
Open/Closed | Open | Closed | |||
2024-01-02 | sepideh | Attached File | - | Added savannah.png, #55507 | |
Attached File | - | Added custom-tab.fits, #55508 |
Powered by Savane 3.13-caa5.
Corresponding source code
Thanks for reporting this Sepideh!
The problem is not in MakeProfile, but in the underlying custom table that was given to it: the resolution (interval covered in each row) of the custom table was 1 pixel! Therefore, MakeProfiles had to round each pixel's distance to the nearest pixel, which would cause these artifacts.
To fix this problem, if you used 'astscript-radial-profile', you should add an '--oversample=5' (for an oversampling of 5). In this way, the interval covered in each row of the radial profile script will be 1/5th of a pixel.
To make this more clear in the book, in commit df7cd772b, I added a full script that can help demonstrate the problem. Have a look at it and let me know if anything is unclear ;-).
Here is a screenshot of the output of that script (notice the edges of the 2D profiles with different oversampling):