Author Topic: incompatible image geometry  (Read 2789 times)

Offline JGMoreau

  • PixInsight Addict
  • ***
  • Posts: 131
incompatible image geometry
« on: 2016 June 16 11:14:40 »
Hi, i just installed a Win 7 pro machine(LAN line) in my observatory as part of my home network.
I am just too far away for long usb with too many devices.
So far so good, but i run into an incompatibility geometry problem when i want to use batch preprocessing
on the saved images on the remote computer, that is i copy/paste the image file from remote desktop to my Win10
computer in house.
I use Nebulosity4 for capture in 16 bits Fits.

Fits header says 3896 X 2612 pixels captured with Win7
and 3894 X 2610 pixels captured with Win10 and long usb active cable, and no such problem.

What is the best strategy to deal with this ?

Thanks in advance,

JG Moreau
Eastman, Quebec

Offline pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4729
Re: incompatible image geometry
« Reply #1 on: 2016 June 16 15:20:39 »
what kind of camera is it? there must be some slight difference between the driver or how the capture program is configured. this happens sometimes with DSLRs. easiest fix is to throw away the old frames but if you can't do that you'll have to pixel peep the images to try to figure out where to add rows/columns to the small images (or delete rows/columns from the big images)

rob

Offline JGMoreau

  • PixInsight Addict
  • ***
  • Posts: 131
Re: incompatible image geometry
« Reply #2 on: 2016 June 16 15:35:22 »
Its a QHY10 OSC.

Sounds not an easy solution.

Resampling is not an option ?

Or make a new set of calibration frames with the Win7 machine ?

I still have the 35 M long powered USB cable to my Win10 machine as a back up...

Thanks nevertheless,

Jean Guy Moreau

Offline pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4729
Re: incompatible image geometry
« Reply #3 on: 2016 June 16 17:38:01 »
especially with an OSC you have to tread very carefully because whatever modifications you do have to be lined up right with the bayer matrix. for instance you'd have to take off 2 columns and 2 rows, probably at the right and bottom edges, of the larger frames to make sure the bayer matrices are still correct. taking one column of of each side for instance would mess things up pretty badly.

resampling is not an option because during calibration, you need the raw data exactly as it came off the sensor. the problem cropped up because either the driver for the QHY10 is of a slightly different version between the two machines, or nebulosity is a different version and is interpreting what comes from the camera slightly differently.

i think by far the easiest thing to do is get new calibration frames. of course that means flats as well, which is more work than just darks and bias.

rob

Offline JGMoreau

  • PixInsight Addict
  • ***
  • Posts: 131
Re: incompatible image geometry
« Reply #4 on: 2016 June 17 05:14:00 »
Thanks very much Rob, i will go the new calibration frames way.

JG Moreau