›

Image Plate Solver Script Not Working

Kelly Hopson
in PixInsight

Brand new to Pixinisight and have been using Image Solver in Processes and all was working just fine until yesterday when it just stopped working and issuing Fail warning. In many pink lines it says "Unable to find a valid set of star pair matches" Then in red at the end it says "Error: Unable to find an initial linear transformation" and then also in real it says " The Image could not be aligned with the reference star field"


I sent the file I was using to a friend and they solved it using the same input values as I had entered.


Any ideas what I am doing wrong. I'm dead in the water with this.


Thank you

Comments

  • BlockHeadBlockHead
    Hi Kelly,

    Almost 99% of the time it means your plate scale, focal length or pixel size is incorrect- especially if someone else solved it and you did not. The software will behave consistently for everyone. You need to compare your screens. I can do this exercise with you as well. Just share your screen and make the file available. (See my video on this. 

    One more thing..I am assuming you have loaded ALL of the GAIA files and all of that is working. 

    -the Blockhead
  • Kelly Hopson
    BlockHead,

    Thank you for your response. I actually did share screens with the person that solved the image on her computer and we had the same numbers. Ran through 3 separate images with same failing results for me where her's would solve. 

    I have loaded all of the Gaia files the DR3 and DR3/SP Complete set. 
    I have also tried uninstalling Pixinsight and reinstalling it with no luck..

    Not sure what else to do.

    Headed to an appointment but can share one of these files that fail in a few hours.
  • Kelly Hopson
    Hello Sir,

    So I downloaded a file after watching the video you sent, 1 tif to my Google Drive and made it shareable but not sure where to send it now. I'm a little fuzzy on what you meant by share my screen and make the file shareable. I did make the file shareable so that's done.

    Forgive my computer skills are not very good. I'm pretty proficient in Photoshop and ACR but that's about it. 

    Thank you,

    Kelly
  • BlockHeadBlockHead
    You basically need to do what you did with your friend again with me.
    Show me a screenshot of your settings (that fail) and then upload the file somewhere for me to download (watch my video). 

    It is likely I to will be able to solve the image- but if you take a screenshot of your image and the ImageSOlver settings- I might be able to figure it out.

    Let me give you an example. If you sent the file you have saved on disk to your friend... and it works- but within PixInsight it does not work for you- perhaps the image you are working on it stretched in PI (but not the file you sent). You see..this would explain things. I am just using it as an example. This example is actually the common one that people trip up on. You cannot use ImageSolver on a stretched (non-linear) image.

    -the Blockhead
  • Kelly Hopson
    Hello,

    Here is a screenshot of the image I was trying to solve with the Image Solver settings I was using and also part of the long error message I keep getting when entering those values in Image Solver.

    I'm so new at this that I only have a couple of files (tiffs) Im working with so can't try new observations until the skies clear again. Bad time to live in Oregon! (rarely clear in the winter)
    Rosette2.jpg
    3792 x 2142 - 950K
  • Kelly Hopson
    I have downloaded this file to my google drive but not sure how to include you. This is what I copied and pasted from the image on Google Drive


  • BlockHeadBlockHead
    edited February 9
    Kelly,

    You gave me a TIF file. If this is what you are using...it will certainly fail to plate solve because it has no information about coordinates, pixel size or anything. You need to platesolve FITS or XISF files.

    Can you please confirm you and I are on the same wavelength? The fact you sent me a TIF is a bit concerning.

    I probably can solve this by figuring out the values...but you shouldn't have to do this with a proper file.

    -the Blockhead
  • Kelly Hopson
    So a file will fail to solve if it's a Tiff because there are no imbedded coordinates? I was manually putting in coordinates and pixel size etc and it was solving when I first learned about the Image Solver process. No idea how I did it but it would go through the process and solve at the end. Now it's not. Ive recently tried converting the tiff to xisf and won't solve. ? My friend solved an image I sent her but maybe it was converted to XISF? Obviously it had to be converted.  Im using a smart telescope and have zero idea how to work with the fits files so had the Vespera telescope generate tiff files to edit. For now.

    As I said I'm so new to all of this and don't know up from down. I've been doing landscape photography for years and have a general workflow I go through and would like to develop this workflow for these images as well but stuck at the start.

    I can tell you that when I first opened up images in PI and learned I need to "solve "the image, I would watch videos on how to use Image Solver and follow along and it was working every time. I'm pretty sure I would immediately save the tiff as XISF file then work with that. It's almost as if I inadvertently turned something on or off and now nothing will solve. I don't know.

    I don't mean to waste your time and appreciate your help but just not sure what to do. 
  • Scott Badger
    edited February 15
    FWIW, ImageSolver works for me on both non-linear and TIF files if I set the coordinates (using the search function) and image scale info manually.

    I believe the problem here is that the coordinates (in the screenshot) are wrong. ImageSolver failed for me too at those coordinates, but using the search function for the horsehead nebula, the coordinates that come up are 5h 40m 59s, -2deg 27' 30s and IS works using those.

    Cheers,
    Scott
  • BlockHeadBlockHead
    edited February 15
    Catalog coordinates for extended nebula should not be used in general. The "center" of a nebula is not well-defined. Coordinates of the actual center of the field should be used. 
    This is a Quickstart video on this topic:

    -the Blockhead
  • Scott Badger
    edited February 15
    FWIW, I just noticed that the file name and identifier (Rosette) in the screenshot don't match the file on the OP's Google drive.

    Adam, I didn't do anything with the advanced settings to get IS to work (not error) on a non-linear image, but if the solution isn't necessarily accurate, that's good to know. Thanks.

    Cheers,
    Scott

    Edit: The image identifier (Rosette) does match the coordinates in the screenshot, but the file on GD is the horsehead nebula and surroundings.
    Edit2: While IS didn't error on a LRGB version of a non-linear image, it did error on the HaLRGB version even though the stars are the same for both.


Sign In or Register to comment.