›

Problem combining data from scopes with 2 different focal lengths

Greg Erianne
in PixInsight
This is the first time I've tried to combine data from scopes with two different focal lengths, and WBPP is having a hard time first with registration, and second with plate solving.

Scope 1 fL 540mm, camera (ASI585MC Pro) with 2.9 micron pixels, 3840x2160; image scale ~1.1"/px; Ha/OII and SII/Hb dual narrowband filters
Scope 2 fL 749mm, camera (ASI2600MM) with 3.76 micron pixels, 6248x4176; image scale ~ 1.1"/px; S, H, O narrowband filters

In setting up the data, I had the data from each scope in separate folders so the appropriate darks/flats would be applied, and in WBPP I used the keyword SCOPE since that's the way I had the folders labeled.

Although I didn't intend this, the images from the 749mm scope were inverted from those of the 540mm scope.  This seems to be a problem in terms of plate solving.  In order to check this, I ran WBPP separately on the data for each scope, brought the master lights into PI and tried to merge them together by doing a star alignment.  However, the only way that was successful is if I inverted the image from the 749mm scope and cropped that image to approximately the same size as the smaller image (from the smaller sensor camera).  

Anyway, I was wondering if someone could get me started about how to set up the data in case I'm doing it incorrectly, or in checking certain parameters in WBPP.  (I tried changing the Registration parameters in Image registration by checking Distortion error and Use triangle similarity, but that actually made things worse.)

I'd appreciate any help getting started!  Thanks so much.

Greg

Comments

  • BlockHeadBlockHead
    Inverted or flipped? (mirrored)
    Are you certain this isn't an origin issue? 
    Were the images in the same orientation prior to any processing? 
    The flipped image I am guessing is the key here. 

    -the Blockhead
  • Greg Erianne
    Thanks, Adam!  I'm fairly certain that the orientation is/was the problem  (180 deg difference between images on the two different scopes) since I was able to rectify this manually using the master lights in PI and inverting the master light from one one of the scopes. So, the images were in the opposite orientation prior to WBPP.  But I was a little surprised, honestly, that the orientation would stump WBPP.  I didn't think the orientation mattered.  

    Yesterday, I changed a few settings, e.g., Sensitivity, and got more of them to register.  I also manually selected one of the lights with the the smaller FOV as the reference (it was previously using one of the lights with the larger FOV), and it worked better.  Still failed to register some frames, though.  I haven't been able to find what settings (if any) to change to get a clean registration in WBPP.

    Do the rejected frames get thrown out during integration, and excluded from the final master light?

    Thanks again for the reply!

    Greg
  • BlockHeadBlockHead
    Well, a 180 degree rotation should not affect registration. You are correct.
    The issue I was concerned about is having different origin axis in the image (which would flip or mirror the image...this is not a rotation).

    It could be a data quality issue really... without looking at the images it is hard to say.

    Yes, non-aligned frames will not be used in the integration.

    -the Blockhead
  • Greg Erianne
    Thanks again, Adam.  Both of the scopes are refractors, so in terms of flipping differently that shouldn't have been a problem.  

    I generated crosshairs in PI on the images a while ago and they are, within about 1-2" at the same origin point. I also checked my software (I'm using an ASIAir) and I had the same RA/DEC coordinates for both scopes.  I know you can't see the image below the smaller one that's at a different focal length but the center point looks almost identical when I blink the images (see attached file).

    It could very well be the image quality, since this was only the first session on this target with a total of about 3.5-4.0 hrs on target from each scope.

    I'm planning to do another run tomorrow night since it's forecast to be clear.  I flipped the camera orientation 180 deg on the 749mm fL scope, so I'm hoping that the images are at least correct in terms of vertical flipping.  

    Thanks! 

    Best,
    Greg
    CenterPoint.jpg
    1246 x 838 - 433K
  • BlockHeadBlockHead
    OK... but again, just so you understand where I am coming from. The origin of the image at top-left or bottom-left is the issue I was making certain was not in play. This has only to do with how the image is read... nothing about the coordinates of RA/Dec or anything like that.

    -the Blockhead
  • Greg Erianne
    Ah, okay.   I misunderstood, Adam.  

    Best,
    Greg
Sign In or Register to comment.