Page 1 of 3

Alternative Plate Solver

Posted: Fri Sep 27, 2019 2:13 pm
by nexusjeep
Hi Robin,
Are there any plans to include the ability to use the ASTAP plate solver in Sharpcap as it appears to be getting a lot of good reviews in terms of the speed of solving, I have tried it via NINA and it is quick.

https://www.hnsky.org/astap.htm#command_line

Cheers
Nick

Re: Alternative Plate Solver

Posted: Fri Sep 27, 2019 6:59 pm
by admin
Hi Nick,

Thanks for the heads up on that software – I wasn't aware of its existence, so I will have a closer look and see how easy it would be to support it in a future version. It looks like it shouldn't be too hard as there is a good commandline interface.

Cheers, Robin

Re: Alternative Plate Solver

Posted: Tue Oct 01, 2019 8:29 pm
by admin
Hi,

Just had a bit more time to look into this and unless I'm missing something the Astap solver isn't a true blind plate solving application – it seems to require the field of view to be provided either from the image headers or as a commandline parameter. Unfortunately SharpCap doesn't have this information, since it doesn't know anything about the focal length of the telescope or any barlows/reducers in use.

On the flipside I guess this is why the plates solving in the application is so fast – because it is only trying one set of data because it knows the size of the image. AstroTortilla is slower because it has to try whole range of different image sizes to find the right one.

Cheers, Robin

Re: Alternative Plate Solver

Posted: Tue Oct 01, 2019 8:52 pm
by nexusjeep
Hi Robin,
Had not thought about that as NINA does know about the telescope and pixel size of the camera so would have an idea of the field of view + location from the telescope. Were as I assume SharpCap only effectively knows roughly were the mount thinks it is pointing.

Cheers
Nick

Re: Alternative Plate Solver

Posted: Wed Oct 02, 2019 5:24 pm
by admin
Hi,

SharpCap has an idea of the pointing if you're using an Ascom mount – that's how the plates solving from the mount control is faster because it gives the pointing hint to Astro tortilla and tells it to only check within a limited range of that point. If you aren't using a mount then you have to use the blind plate solve button from the menu and that has to try all parts of the sky at all fields of view.

Cheers, Robin

Re: Alternative Plate Solver

Posted: Wed Apr 01, 2020 11:43 am
by han59
admin wrote:
Tue Oct 01, 2019 8:29 pm
Hi,

Just had a bit more time to look into this and unless I'm missing something the Astap solver isn't a true blind plate solving application – it seems to require the field of view to be provided either from the image headers or as a commandline parameter. Unfortunately SharpCap doesn't have this information, since it doesn't know anything about the focal length of the telescope or any barlows/reducers in use.

On the flipside I guess this is why the plates solving in the application is so fast – because it is only trying one set of data because it knows the size of the image. AstroTortilla is slower because it has to try whole range of different image sizes to find the right one.

Cheers, Robin
For solving with ASTAP it would be sufficient if SharpCap could provide the mount position in the command line. If the position is within 20 or 30 degrees accuracy know, it will solve in no time. A full 360 degrees search will take a few minutes. Command line example:

astap.exe -f c:\images\image.png -ra 23.000 -spd 179.000

Pre-condition, set in ASTAP graphic user interface the field of view correct within 10% accuracy. This will be saved and re-used if none is specified next time. If it is unknown. ASTAP has also an "auto field of view" function function. I will try all field of views and find a solution. The found field of view will be remembered next time. This learn function can also be executed using the command line with the -fov 0 parameter. E.g.

learn step:
astap.exe -f c:\images\image.png -ra 23.000 -spd 179.000 -fov 0 {solve & find the fov by specifying -fov 0}

next step(s):
astap.exe -f c:\images\image.png -ra 23.000 -spd 179.000

Documentation:
http://www.hnsky.org/astap.htm#astap_command_line

Cheers, Han
author of ASTAP

Re: Alternative Plate Solver

Posted: Wed Apr 01, 2020 5:17 pm
by admin
Hi Han,

Thanks for that information – it looks like I now have all the information I should need to get this working.

Cheers, Robin

Re: Alternative Plate Solver

Posted: Sun Apr 05, 2020 1:31 pm
by han59
Hello Robin,

Noticed you have been active.

I played a little with the latest version from 2 days ago. Hooked up mount simulator and sky simulator. Tried to solve an simulated image but got an error "only 0 stars detected". See screenshot. Could not bypass it. The image written C:\Users\jkl01\AppData\Local\Temp contains stars. 1) How can I bypass this error message 0 stars detected?

I could only initiate a solve from "Test Camera 1" Noticed there that the initial position send to ASTAP is RA, DEC:= 000h,000d. See screenshot 2) The mount was connected but the position was not send to ASTAP.

I'm not familiar with Sharpcap so maybe it is something simple.

Regards, Han
Untitled.png
Untitled.png (137.64 KiB) Viewed 750 times
Untitled2.png
Untitled2.png (51.13 KiB) Viewed 750 times

Re: Alternative Plate Solver

Posted: Sun Apr 05, 2020 2:47 pm
by admin
Hi,

The plate solve item in the menu does a whole sky blind solve and takes no further action (just displaying the solution as an alert in SharpCap). This will run the plate solver with the centre position being RA=0,Dec=0 and the radius being 180°. In the case of Astap it will not send any parameters to control field-of-view, so the FOV will be used from the configuration or the last run if I understand correctly.

The other place to kick off the plate solving in SharpCap is from the telescope mount control – if you have connected a mount via its Ascom driver then alongside the directional controls is a 'GPS location' button. Using this button will perform a plate solve with a radius of 15° around the current mount coordinates. If the plate solver succeeds then SharpCap will try to re-centre the target in the field of view either by synchronising and then going to the target coordinates again or by an offset go to. Using this option will potentially run Astap twice - first with no FOV option and if that fails with '-fov 0' to do a more thorough search.

The number of stars detected is extracted from the solver log file – the contents of the log file and the WCS file if it is created should end up in the SharpCap log, so be interesting to see if there's any useful information there.

Cheers, Robin

Re: Alternative Plate Solver

Posted: Sun Apr 05, 2020 4:54 pm
by han59
Hello Robin,

Your setup looks good to me.

I found the solve button at the telescope menu and looked a little deeper why no stars are detected by ASTAP. The problem using the ASCOM simulator is that the image fed to ASTAP "frame.png" has only two levels, 0 and 1 of the range 0..255. It will see the 0..1 values as background noise and find no stars. For ASTAP to work the orginal grey levels should be preserved. ASTAP will use them to detect the brightest stars and separate them from the faintest and noise.

Attached a frame.png. It looks black but it contains an 1 bit image of M13 with values 0 and 1. Is there a way to preserve the original gray levels or is the effect I'm seeing caused by using the ASCOM simulator (8 bit image) instead of a real camera?

What happens if Sharpcap receives a 16 bit image. Does Sharpcap writes a 16 bit frame.png or a 8 bit frame.png for solving? A 16 bit file would be much better.

Regards, Han