Trouble installing platesolve3

[Stargeezer1]

I have a 16 inch Alt Az with a focal length of 1779 mm. I'm useing a ZWO ASI678mc camera and am trying to set it up for plate solving. The chip is only 7.68mm x 4.32mm so it may or may not be able to work. As I understand it, my best chance is to use Platesolve3. My issue is that when I install it, it produces no exe file. In the plate solve setings tab, selecting Platesolve3 requires a path and an exe file. The only thing that installs on my Windows 11 machine is a Kepler directory and sub directories but no exe except an uninstall file. What am I missing?

[admin]

Hi,

interesting - I have always previously used the download link listed on the NINA documentation (https://nighttime-imaging.eu/docs/maste ... latesolve3) to download PlateSolve3 in a zip file that included both the executable and the index files. I just tried that a moment ago and the zip file now does not contain the application, just the index files.

The download file has obviously been changed - previously it was about 5.5Gb, now it is 3.8Gb. There may be copies of the older version of the download available elsewhere online (be cautious - this is a download of copyrighted software - other online copies may not be authorised or may have been tampered with). At the moment, I can't seem to locate a legitimate download of the actual software online.

It may be worth having a look at the NINA support Discord forum (https://discord.com/invite/nighttime-imaging) to see if they have discussed this change and/or have a new download link.

cheers,

Robin

[Stargeezer1]

I found this post on the Cloudy Nights forum.

Posted 27 May 2025 - 06:59 PM

Hi folks. The research group I am a part of (headed in part by Dave Rowe) has access to the PlateSolve3 source code. Our group has made some fixes and updates to it. Nothing major - some bug squashing, making it work correctly on PCs that use a comma as the decimal separator, and packaging it in an installer. I am currently working with Dave and PlaneWave to work out a solution to make PlateSolve3 available in a more formal venue that would also allow future versions to be published. I have no timeline to offer as discussions have just started, but I hope it won't be far from now.

[admin]

Hi,

thanks for sharing that info - good to know. Shame that in the meantime the download that worked is now broken

cheers,

Robin

[kaymann]

Nagging question - is platesolve3 better with native SCT's (2350mm) because of a deeper catalog to inquiry or is it in the numerical magic only they and you know how to comprehend and code

If it is in the catalog I have downloaded the big one you allow, but could you allow an even larger optional one for SharpSolve?

[admin]

Hi,

I must admit that I haven't tried creating an even larger star database for SharpSolve - right now there is the 0.5 degree default one that contains about 10 million stars and the 0.25 degree 'extended' one that has about 40 million stars. A 'super extended' one to get down to maybe 0.125 degrees would need ~160 million stars and be somewhere over 0.5Gb in size. It would also need some extensive testing against images at that scale (actually cutting appropriately sized ROIs out of bigger images - say 1x1 degree) works nicely for testing lots of different fields of view from one image...

So in theory, it might be possible - one thing that concerns me though is that for the current 'extended' database there are some areas that have sparse data in the database because there simply aren't enough stars in that region to meet the required number - this is usually for areas looking straight out of the galaxy away from the milky way or any dense star fields. I'm already using the Gaia database as the source of star data for the catalog, so I don't think there is anything with more stars in that I can switch to.

cheers,

Robin

[kaymann]

PS3 works with sct's native, just trying to do the same with SharpSolve...
Thank you for considering it.

[admin]

Hi,

what camera do you have combined with your SCT? I have had SharpSolve working with an IMX571 camera at 3600mm focal length - that's an FOV of 0.37x0.25 degrees. It ought to work with smaller sensors quite well at 2350mm, say an IMX533 (11.3x11.3mm sensor), but probably not anything much smaller like the 585 (11.2x6.3mm)

cheers,

Robin

[kaymann]

...
what camera do you have combined with your SCT? I have had SharpSolve working with an IMX571 camera at 3600mm focal length - that's an FOV of 0.37x0.25 degrees. It ought to work with smaller sensors quite well at 2350mm, say an IMX533 (11.3x11.3mm sensor), but probably not anything much smaller like the 585 (11.2x6.3mm)
...

I have a ASI533MC with a 9.25 Native 2350mm F/10. Adding the 2.5 Barlow is F/25 5875 (I know this is impossible to PS).

[admin]

Hi,

OK, so with the 533 and the 9.25" SCT, you should have an FOV (without barlow) of 0.28x0.28 degrees. That should work fine with SharpSolve already (providing you have downloaded the extended index from the plate solve settings). Based on my testing I would expect a success rate of >99%, providing you have enough clear stars in the frame.

If you add the x2 barlow then you are at 0.14x0.14 degrees - I would expect that to fail in SharpSolve more than 50% of the time, even when there is a good selection of stars to work on in the image.

As an interesting experiment, I created a 'super' plate solving database for SharpCap - in theory designed to work down to a field of view of about 0.125 degrees, with four times as many stars in it as the normal 'extended' database. The data file for this is 795Mb, and will contain data on nearly 200 million stars. Using this data file on my existing test images, I can get successful plate solves at 0.14x0.14 about 85% of the time. The way I test is to cut ROI regions from larger images (0.75x0.5 degree FOV from an IMX571 camera) and then see how many of those solve. Some of those ROI areas are pretty poor in decent stars, and I suspect that this is a significant contributor to the failure rate of 15% at these scales - to check this I would probably need to go and capture some longer exposure images at the same FOV to get better stars relative to the image noise.

Do you have any sample images (single frames that were captured at reasonable settings for plate solving - ie 10s exposure or similar) that you could share - particularly with the 2x barlow in place? It would be intersting to see if those solve with my new 'super' index.

cheers,

Robin