Solver error? Resolution wrong calculated

[tommy_nawratil]

hello friends,

in applying Photometric Color Correction the solver fails, and it seems the reason is that the calculated resolution is not right.
It should be 0.78 arcsec/pixel not 9.221
or, what is my mistake? Cant get it. Look here:


thanks, Tommy

[tommy_nawratil]

I probably found the source of the problem. When PCC is called up it showed up with 85mm focal length written in the respective box.
That together with 3.8µ pixelpitch yields 9.21
Of course I changed to 1000mm, and also clicked into another field so it should became active.
But it did keep the old value internally. Even after closing and reopening with the new focal length.

After solving with a script PCC took the solution and worked normally.

Running PI newest version 1.8.8-5

Tommy

[Juan Conejero]

This is not a bug, but the expected behavior. The coordinates, observation date, focal length and pixel size parameters of PCC are just default values, which are only used if the target image has no valid metadata for the corresponding items. When the target image has valid metadata, it always takes precedence over user-defined values in PCC's interface. If the image metadata is wrong, then obviously the process may fail. Always click the 'Acquire from image' button with the target image active to check the actual metadata that will be used.

Admittedly, the fact that some PCC process parameters are used as fallback values should be made more obvious. I'll try to improve this time permitting.

[tommy_nawratil]

thank you Juan,

the thing is, there were no metadata in the original image. WCS header empty.
I went back to initial state, pressed "acquire from image" and the values that were inserted before (1000mm  3.8µm) remained.
Then tried to solve via PCC, again it came up with Resolution: 9.221 as/px
The image solver script was doing it correctly, again.

Its a little hickup and maybe just machine-related, do not sacrifice your valuable time.

best, Tommy

[niteman1946]

Hi Tommy,

hope you don't mind me jumping in here.
If this is not relevant, let me know.

When doing PCC, this has always been my procedure (my only way to make it work):
1.  type in the optical train's focal length and the camera's pixel size
2.  then, go to Search Coordinates and type in the targets name (i.e. NGC1579) for coordinates

Reading Juan's comments, I went back to a recent image and tried again the Acquire from Image button, and still no love.
However, using a single sub did produce the correct parameters.  Also using an integrated image provided the
parameters.  However, once the integrated image is cropped, then the data is no longer available.

I've noticed also, and it makes sense, that the coordinates from the "named" target are slightly different than those "acquired".
 But apparently not so much to make a difference.

From here on I'll be bringing at least an individual sub along to define the parameters.

Hope this helps,

Mark

[tommy_nawratil]

hi Mark,

thanks for the input! Does it mean you also have a problem here, and its not only me?
My workaround is simple: Let the image solver script do the solving, and PCC will take these values automatically.
Sometimes I also use Hartmut Bornemanns FluXX script for color calibration, it does the solving right, too.

Tommy