View Issue Details

IDProjectCategoryView StatusLast Update
0002203CCdcielGeneralpublic19-10-12 13:54
Reporterhan Assigned ToPatrick Chevalley  
PrioritynormalSeverityminorReproducibilityhave not tried
Status resolvedResolutionfixed 
Target Version1.0 
Summary0002203: Histogram scale different compared with image pixel values
DescriptionWhile making flats for a local member, I noticed that the histogram scale is different compared with the values measured in the image (flat). It is possible to set the black value in the top left Tedit&Tupdown of the histogram and white at the top right Tedit&Tupdown of the histogram. Only these values have no relation with the image, so this could be confusing. Attached two sample flat.

The file "Flat_0_3s_-15.4C_20191003_104248" is made with CCDCiel and has for keyword datamax the value 31568. For this original file CCDCiel Tedit&Tupdown show a different scale. The other file is saved by SGP&ASTAP and has datamax value 65535. For this file the histogram Tedit&Tupdown values look correct.

I have the impression it is related to the datamin, datamax values.

TagsNo tags attached.

Activities

han

19-10-03 13:38

reporter  

ccdciel histogram scale.png (10,084 bytes)   
ccdciel histogram scale.png (10,084 bytes)   

Patrick Chevalley

19-10-03 18:24

administrator   ~0005986

Yes, the histogram values are for the internal display image where datamin-datamax is stretched to 0-65535.
This look useless at first, but this greatly simplify the image display when the fits file is not 16bit integer, for example a file saved by Pixinsight as floating point ranging from 0 to 1.

But I agree this two box must show the real value, as this is done at the other place a pixel value is show. This is the FimageMin+value/FimageC you can see at various place in the source code.
The only difficulty is to use a floating point input box when the fits file not use integer values.

han

19-10-03 21:29

reporter   ~0005987

For ASTAP, to read Pixel Insight/ APP floating point the range is changed from 0..1 to 0..65535 during loading. But my program should be more universal.

Anyhow it would nice if it could work correctly for native CCDciel files. My first idea would be to correct the Tupdown values with a datamax/65535 factor. It is not so important but it would be helpful if you cursor measured pixel values match with the histogram range.

P.s an other improvement could be if you could measure the histogram with the mouse using the Tonmousemove event. I did it for ASTAP.

Patrick Chevalley

19-10-08 23:01

administrator   ~0005998

Hello Han,

I do the change to show real values for integer and floating point files, also show the value in the status bar when moving the limit with the mouse.
https://github.com/pchev/ccdciel/commit/3f31b6061aa8b7de0e4481f0f9b11a133235e652

And this change to prevent to set min > max, even if this is already handled in the fits routines.
https://github.com/pchev/ccdciel/commit/39344fd258233e60d31d5bf10ccc75afbbe4ec65

I also add your photometry functions.
see https://www.ap-i.net/mantis/view.php?id=2181#c5993

Patrick Chevalley

19-10-12 13:54

administrator   ~0006001

The change is now include in new version 0.9.61.

Issue History

Date Modified Username Field Change
19-10-03 13:38 han New Issue
19-10-03 13:38 han File Added: test images.zip
19-10-03 13:38 han File Added: ccdciel histogram scale.png
19-10-03 18:24 Patrick Chevalley Assigned To => Patrick Chevalley
19-10-03 18:24 Patrick Chevalley Status new => assigned
19-10-03 18:24 Patrick Chevalley Target Version => 1.0
19-10-03 18:24 Patrick Chevalley Note Added: 0005986
19-10-03 21:29 han Note Added: 0005987
19-10-08 23:01 Patrick Chevalley Note Added: 0005998
19-10-12 13:54 Patrick Chevalley Status assigned => resolved
19-10-12 13:54 Patrick Chevalley Resolution open => fixed
19-10-12 13:54 Patrick Chevalley Note Added: 0006001