›

Mosaic Help

Rex Alldredge
in PixInsight
I'm trying to do a six panel mosaic.  Following your video, my idea was to produce two columns, of 3 images each.  The first two went together no problem.  When I attempted to do mosaicbycoordinates on the mosaic I had formed and the next panel, I received the error message "no astrometric solution".  A little research showed the two panel mosaic had no fits header at all??  I went through my old two panel mosaics and none of them had a fits header either.  So nothing works, not Image Solver, and certainly not Mosaicbycoordinates.  Your video seems to show you have no problem with this.  What am I doing wrong?

Comments

  • BlockHeadBlockHead
    The astrometric solution and other metadata is now found under the VIew Explorer.
    This is a change with the newer versions of PI.

    It sounds like the third panel is the issue? It needs the astrometric solution. The other two must be OK since they went together. 

    -the Blockhead
  • Rex Alldredge
    I'm trying to put the mosaic from 1&3 with 5.   5 definately has an astrometric solution.  It is the 1&3 mosaic which down not and I can;t seem to add one through Image Solver.
  • BlockHeadBlockHead
    Maybe you do need a solution for each created panel...I can't remember right now.
    I didn't think so because you will need to figure out the center of the frame.
    -the Blockhead
  • Rex Alldredge
    I guess I've pretty much given up on mosaics with more than two panels.  I believe Photometric Mosaic should output an astrometric solution for the Mosaic.  I've done may two panel mosaics.  None has a astrometric solution acceptable to MosaicbyCooridinates.  Image solver is very iffy in providing a solution.  I've had to fool around a lot to get it to work.  I was doing a 9 panel mosaic of the Pleiades.  I finally got the rows which looked pretty good.  Trying to get AS  for a row was very very hard.  The system finds many stars but cant align them with the catalog.  Eventually I copied the file header from the central panel to the row.  I'm not sure what I am doing even makes sense.  The final mosaic was awful, the row joins were very obvious and the were absolutely dark regions in the near the join.

    I'd like to think it isn't my computer since lots of stuff seems to work fine.  I suspect PIXINSIGHT has a problem with large image sizes.  This data comes from Telescope Live AUS-2, which gives 60 megapixel images.  I couldn't get those to work at all.  I had better luck with 15 megapixel images from CHI-1, but in the end, the result was very poor.

    By the way, computer is Ryzen 9 5950 with 32 Gigs for memory.
  • BlockHeadBlockHead
    I did not have these kinds of issues... I just looked at the 2x2s that put together to make the 4x4 and everything had solutions and coordinates.

    (please confirm your images are linear)

    If you would like to make the color frames available and I can take a look. 
    I don't think it should be that hard- and getting good seems with PM is pretty easy...there is enough adjustment to get what you want.
    I am using Processor 13th Gen Intel(R) Core(TM) i9-13900K, 3000 Mhz, 24 Core(s), 32 Logical Processor(s)
    with 64Gb of RAM. 

    -the Blockhead
  • Rex Alldredge
    Images are linear. Maybe I should go to 64 GB. Have you tried the big telescope live images?
  • Rex Alldredge
    I’d love to make some of my images available, if I only knew how. 
  • BlockHeadBlockHead
    You would need to upload them to a cloud server (Google Drive, One Drive, DropBox, WeTransfer...etc) and set the permission to share with anyone that has the link. That is pretty much it. 
    I have not tried the TL images. However the 4 panel I did had drizzled frames  (6000x6000). That is still smaller than the large sensors...but it isn't too bad. I am guessing the machine I have now would be OK. I think it is considered a reasonable one.

    -the Blockhead
Sign In or Register to comment.