I did notice that a couple of NGC thumbnails are there. I have also tested a manual method of adding them to kstars by copying ngcX.png and thumb_ngcX.png to the $HOME/.local/share/kstars directory with the other PNG files. This seems to automatically assign NGC object X the thumnail image desired. If someone has a complete thumbnail set of NGC objects and can reformat them and zip them back up and maybe they can be included in the distribution if the licensing is appropriate.
So I have worked out a process that I am going through now. This process does the following:
1 - Pulls down the images from the skyview nasa site for each NGC object and combines them into an RGB color image.
2 - Transforms the image from a jpg to png format picture
3 - Renames the objects to the thumb-ngcXXX.png name format they expect.
I am in the middle of pulling the first 1000 NGC objects now. I will delete the duplicates that are already included in the kstars images. I will also have to re-grab several of the objects with a larger pixel size to fit the full image into the resulting frame. I am only going to do that for images that don't comfortably fit in the 100x100 pixel frames I pulled during the initial run. If this all goes well, I am pretty sure I can do this same process for the other catalogs as well. If knro is interested, I will upload the files for inclusion in the distribution.
If you want to do this process for yourself, you can download a few packages and do this yourself.
1 - Download Skyview in a Jar from the NASA website
2 - Install mogrify and mmv packages in linux
3 - use the following syntax on each NGC object you wish to grab the file from the skyview website for:
java -jar skyview.jar position='NGC 1' survey=dss,dss2b,dss2r rgb scaling=linear pixels=100 output=ngc1
NOTE: You will need to make adjustments for the NGC name and the pixel size of the image.
4 - Transform the image from jpg to png with mogrify
5 - Rename the files with a batch rename in mmv or individually if you want to lose your mind.
6 - When you have the objects you want to include thumbnails for, you can copy the png files to $HOME/.local/share/kstars and it should show up if the objects are named properly and are of the correct format.
I changed the order of the surveys so that when they combine, they will be in the proper channels R G and B. Overall, the images are heavy in the green and I haven't figured out how to make that go away in the pull but and probably do a batch color correct in photoshop after the fact. You will still need to adjust the size parameter or the pixels parameter for larger objects, but otherwise this should pull data directly for anyone who is interested.
It will take me several days to get things finalized but I will get it to you after that. It takes time to pull each image down and compile them into an RGB image. I should probably do an at least rudimentary color correction in photoshop for each image as well. I will also need to re-pull several of the larger objects with different settings.
When I have them all, I will archive them up and work with you to get them to you. This very same process should work for the IC and Messier objects as well and probably any other catalogs that NASA has tied to skyview.
Upon further consideration, I have decided to only pull the red channel and leave these images and black and white. That will save significant time in color correction. All I will need to do once I complete the pulls is re-pull the objects that are of larger size than the current settings account for. That should enable me to get something to you before the end of the week.
I decided to only pull the DSS2r (red channel) images for these. It appears the original DSS images are at a lower resolution and they don't have a DSS2g channel run. I made the assumption that the DSS run was the green channel. Anyway, I figure black and white thumbnails of the red channel would be adequate. I have pulled all 7840 objects and the files are in jpg format in 500x500 pixel files. I need to convert them to PNG files next. The initial challenge is that the png files are significantly larger than the jpg files on conversion. I need to figure out what compression syntax to use in mogrify to get the size back down to jpg levels which makes the images about 25k each. The current png conversions are coming out over 120k each.
I have a request to help me get the sizing of the objects correct in the frame. I believe you have the apparent size listed in all of the NGC objects. Is there any way you can send me a comma separated output with the NGC number and the apparent size in arcmin for each object? What I can do from there is use that data as input for the script to pull down the image with the proper sizing straight away instead of just picking a set size and correcting from there.