PlateSolve2 incorrect initialization

All localization issues have been addressed. We have people from all over the world successfully using PS2 now (with 2.4.2.10). We need to see the SGPro dialog that sends hints to PS2 and your logs or we will not be able to provide much help.

You are absolutely right. I was in the lab and reproduced it.

sg_logfile_20150906171144.txt (27.4 KB)

I hope it helps.
Thomas

Can you also post the image you are trying to solve?

No it is not allowed in this forum.

I stored it in my dropbox

Thx. I will see what I can find a bit later tonight.

Also, I do not think the issue is regional at this point, but can you let me know what region your machines are set to?

Well, I confirmed that there is no regional problem, but I’m afraid I cannot be of much help to you with your image solving issues. Taking SGPro out of the picture entirely, I just start up PS2, load your image, enter the exact location, scale and size as hints (I confirmed these hints with a Pinpoint solve). Even with hints having extreme precision, PS2 does not solve the image you made available (I tried both APM and UCAC3 catalogs).

Just a couple of comments. The image was captured with Nebulosity and the FITS headers that were inserted are formatted very differently than headers inserted by SGP. Not saying they are wrong, just different. SGP is not able to get the hint from the headers that PlateSolve2 needs to be successful but even when an accurate hint is given, PS2 still fails to solve.

I was successful in getting a solve from Astromentry.net. My stand-alone copy of PinPoint also fails to solve even with an accurate hint. PP reports this image contains 84,000+ stars and PS2 shows the image size is 222’ x 148’ and Astrometry.net shows the plate scale is 3.33 arc seconds / pixel.

I suspect this image is simply overwhelming both PS2 and PP because it captures a much bigger field than either of those programs were designed to process.

Personally, I have found PS2 to be very reliable with my SGP images (KAF-8300) but my images have 100X less area than yours!

It would be interesting to get some feed back from SGP users that have STL-11000 cameras.

cm

For whatever it’s worth, Pinpoint was successful on this image when I was playing around with it (GSC).

Hello
I just purchased SGPro. Nothing but clouds and rain here in S. FL.
SO I have been learning the program and have been following this thread.
I have been using Dave’s PlateSolve for about 3 plus years now with a few different OTA and camera set up. All thou SiTech.EXE. What I have found on large FOV images is that some time I have to go into the edit parameters and lower the Max Star Size Sigma. I down loaded the posted image and after seeing how LARGE the FOV was, I set the Max Star Sigma to 2 and the plate size set to 1 X 1 degrees, it solved in 10 regions. I have found that watching the image and seeing how PlateSolve picks the stars, shows the stars in the catalogue and the ones that matched, one can get a better feel for what is happening. Load a easily solved image and play with the Max Star & Detection Threshold Sigma setting and see how it responds. Also change the Plate size, smaller sizes in the field seems to have the matched stars in a smaller region. It asol lets PlateSolve run thou the regions to find a match much faster. With my Canon 7D, I need to change the min & max star size fields when doing projection imaging in order for PlateSolve to work well. One need to play with it in order for it to work best with there imaging set up.
Not sure if it’s of any help, I thought I would share what I have found to work for me.

Steve B, new guy

@Steve_Barton

Thanks for the tips. Not too many folks (here) are super-familiar with the intricacies PS2 settings and anything helps. The unfortunate part of this “fix” is that I don’t think the interface between SGPro and PS2 is flexible enough to tweak these parameters at run-time. This means that if you switch back and forth between gear that significantly alters FOV, you will need to remember to set this manually each time. Better than failing though… worst case is that there is a minor inconvenience.

Hello Ken, Hello Steve,
thanks for your replys. I must excuse my poor english, I learned it at school.

@Ken: I do not know, if it is worth mentioning. Have you recognised that my ps2 window shows “degrees” instead of minutes? Is this controlled by SGP?

I do not have a observatory, if it is clear (rare here) I work in
the field, so I need a foolproof solution. If I select “use blind
solver” astrometry.net solves the image right. But I do not have
reliable internet in the field.

I have telescopes from 550 to 1135mm focal lenght and a STL and a
Moravian G2-8300. What should I do, install elbrus or a local version of
astrometry.net? Is the local install of astrometry.net official
supported way? I use a german version of windows7

Thomas

Hi Thomas,
If you click on the “degrees” button it will change to “arcminutes”

Nigel

Hi Thomas

For me, Elbrus works very fine and is also very fast, taking a few seconds to solve an image.

The only thing to keep in mind is that it needs to have the image scale hint. You can set it up in the profile you are using.

Best regards,

Alfredo
Bogotá, Colombia

Thanks Nigel, I noticed, but if platesolving was started by SGP I cannot change this parameter intitial.
Thomas

Thanks Alfredo,
is your equipment wide field like mine or has it a small field?
Thomas

Hi Thomas

It is wide field. My setup is a C925 EdgeHD with hyperstar at f2.3 (540 mm). I use it with a modified Canon T1i, which gives a FOV of 142 x 95 arcseconds (2.4 x 1.6 degrees).

Alfredo

Hello Alfredo,
a fine equipment. I tried elbrus but had no success, I probably run into the same error as described here:

sg_logfile_20150908142447.txt (31.3 KB)

Finally I installed a local version of astrometry.net. This works much better, no chrashes no errors. I can sucessfully platesolve images made with the same equipment, but not the problem image I mentioned in this thread. All this is very time consuming.

Thomas

Hi guys,

more than 3 years later I’m facing exactly the same initialization problem. The tread ends without any hint of solution. Does it continue somewhere else?
Cu ggtux

I suggest you start a new thread with the details of the issue you are having rather than trying to use this old thread which may or may not have anything to do with your problem. You need to give specific examples of what you are seeing with appropriate screen images or other details so that the issue can be identified. To my knowledge, there is no general PlateSolve2 initialization issue so you need to give enough details to help isolate your particular problem.