Mosaic issue--Star Alignment

[alanrock]

I have a six panel Mosaic that I am trying to assemble. The first three panels aligned properly, but for some reason I cannot get the other panels to align. It's a very busy part of the sky--lots of stars to choose from (maybe too many?) Any suggestions? Here's the tail end of the log from the process console:


* Reference image: Limiting to 8000 brightest stars.
Matching stars: done
2132 putative star pair matches.
Performing RANSAC: done
* Previous attempt failed - this is try #24
useScaleDifferences=true
* Reference image: Limiting to 8000 brightest stars.
Matching stars: done
1642 putative star pair matches.
Performing RANSAC: done
*** Error: Unable to find an initial set of putative star pair matches
<* failed *>

[pfile]

you can make previews over the parts of the mosaic and the part of the pane that you expect to overlap to help with this issue.

rob

[chris.bailey]

Once you get past four panels building the mosaic with star alignment does get more and more hit and miss, especially with wider fields of view. The MosaicByCordinates and GradienMergeMosaic is more certain of success.

[pfile]

yep i agree with that, i only use MBC and GMM now for these purposes...

rob

[alanrock]

I've never heard of either of those--any tips on use?

[alanrock]

When I try using ImageSolver, I get this:
Downloading Vizier data:
http://www.ukirt.hawaii.edu/viz-bin/asu-tsv?-source=I/259/tyc2&-c=314.179921 44.031400&-c.r=1.679454&-c.u=deg&-out.form=|&-out.max=200000&-out=TYC1&-out=TYC2&-out=TYC3&-out=RAmdeg&-out=DEmdeg&-out=pmRA&-out=pmDE&-out=VTmag&-out=BTmag&-out=HIP&VTmag=<13.65
 *** Error: response code = 302

[pfile]

try choosing a different server... (the orange cylinder next to the catalog chooser). i have mine set to the default which seems to be the server in france.

rob

[alanrock]

It fails with the error "This projection is undefined"

[pfile]

upload the image

[alanrock]

https://drive.google.com/file/d/1Z54Sz8StvbY48URhd8JHPSpb-ElNf6Ab/view?usp=sharing

[Andres.Pozo]

The image solves without problems using the default parameters. I only had to estimate the pixel size of your camera (2.4um) since it is not in the FITS header.

Code: [Select]

Image Plate Solver script version 5.1
===============================================================================
Referentiation Matrix (Gnomonic projection = Matrix * Coords[x,y]):
           +5.15681e-06         +0.00025507           -0.482524
           -0.000255033        +5.16292e-06           +0.691295
                     +0                  +0                  +1
Projection origin.. [2747.847349 1836.196702]px -> [RA:20 56 35.500 Dec:+44 47 08.86]
Spline order ...... 2
Num. ControlPoints. 89
Resolution ........ 0.918 arcsec/px
Rotation .......... -91.163 deg
Focal ............. 539.04 mm
Pixel size ........ 2.40 um
Field of view ..... 1d 24' 7.4" x 56' 12.3"
Image center ...... RA: 20 56 35.484  Dec: +44 47 08.72
Image bounds:
   top-left ....... RA: 20 53 50.454  Dec: +45 28 28.72
   top-right ...... RA: 20 54 03.909  Dec: +44 04 26.53
   bottom-left .... RA: 20 59 10.897  Dec: +45 29 37.88
   bottom-right ... RA: 20 59 16.656  Dec: +44 05 31.63
===============================================================================


[alanrock]

I still get the same issue, and I have left all the parameters at the defaults.

[msmythers]

You might try reseting ImageSolver first. It solved for me also using the same 2.4 pixel size as Andres.


Mike

[alanrock]

That worked! I certainly don't remember having changed anything in there, but it's possible. Thanks guys!

[msmythers]

Glad things are working again.



Mike