PhotometricColorCalibration wont work

[blinky]

I get the error below when I run the Photometric Colour Calibration routine - Have checked the RA/Dec and pixel size/focal length are correct, not sure what else to do?

Processing image C:/Users/Craig/AppData/Local/Temp/PCC_T_8VL7YQJLJV3U.xisf

Reading image:
C:/Users/Craig/AppData/Local/Temp/PCC_T_8VL7YQJLJV3U.xisf
Loading image: w=3040 h=2016 n=3 RGB UInt16
88 FITS keyword(s) extracted.
Seed parameters for plate solving:
   Image coordinates: RA=05 35 33.800, Dec=-05 16 12.40
   Resolution: 0
Starting StarAlignment iteration
Catalog Bright Stars already loaded
Catalog Bright Stars size: 9046 objects

*** Error: Unable to solve the image: Error: Invalid matrix inversion: Singular matrix.

** Warning: Process finished with errors:
  PCC_T_8VL7YQJLJV3U.xisf: The image could not be solved

*** Error: Failure to plate solve image: light_BINNING_1_integration
Reading swap files...
610.707 MiB/s
<* failed *>

[pfile]

can you post the xisf so someone can take a look? google drive, dropbox... post link here

rob

[blinky]

Here you go...  https://www.dropbox.com/s/f4aw6pt5sybwaa8/light_BINNING_1_integration_DBE.xisf?dl=0

I grab the co-ords from the image but manually enter 800mm focal length and 7.8 microns for my QHY8 pixel size

[msmythers]

The file would not work as you stated. I decided to resave the file without the FITs header keywords. I then opened that image and using the default settings the image solved. Something in the keywords is wrong or corrupted.


Mike

[Andres.Pozo]

The problem is that the key FOCALLEN has a 0 value. ImageSolver ignores it but PCC doesn't. I can not fix this since PCC was not written by me.

After clicking "Acquire from image" in PCC you can manually set the focal length and pixel size to the correct values. Another workaround would be to first solve the image using ImageSolver and then apply PCC. In any case I would recommend to always use this method since it lets you try PCC with different parameters without needing to solve the image again.

Also, for this image I would activate the distortion correction in ImageSolver. There is a noticeable geometric distortion at the corners. The solution using distortion correction has a RMS error of only 0.61 pixels versus 2.4 without it.

[blinky]

Odd - I entered the values that are on the screenshot and it still wont solve it!

[msmythers]

You still have the resolution = 0 issue with the file. I saved the image without the keywords and than solved that saved image. Andres method is easier. You click on the "Acquired from Image" button and then manually enter your Pixel Size and Focal Length.


Mike

[blinky]

You still have the resolution = 0 issue with the file. I saved the image without the keywords and than solved that saved image. Andres method is easier. You click on the "Acquired from Image" button and then manually enter your Pixel Size and Focal Length.


Mike

I thought I did Andres method - I you look at the screenshot you will see I have the same focal length as Andres screenshot.  Where is the resolution field? I dont see anything set to 0

[msmythers]

The resolution is not a setting. It is parameter showing Process Console window. Try these parameters, 'Force late solving' and 'Ignore existing metadata'.


Mike

[blinky]

Tried that on an image tonight -  same problem, even if I select force plate solve and ignore existing metdata

[blinky]

Finally figured it out -  you cant drag the little triangle to execute, you need to click the square!

[msmythers]

 


Mike

[Juan Conejero]

Finally figured it out -  you cant drag the little triangle to execute, you need to click the square!

Weird. If you drag the triangle on the active image view, both triangle and square icons perform exactly the same task. I guess it's something else...