The 64 bit version for Raspberry Pi (arm64) had a bug which was related the optimisation factor of the compiler. This is now fixed (prevented). ASTAP is now at 0.9.289 and should work on all platforms.

Han

I'm glad it works for you. I suspect that with 3x3 binning setting the number of tetrahedron matches will be less, so the reliability will be less. But if it works for you keep it for the moment. In case of solve failures go to bin 1x1 or 2x2. If the -update is not required ,yes then remove it. It will speed up the process with one or two seconds.

There is no need to change command line parameters but just in case they are specified here:

www.hnsky.org/astap.htm#astap_command_line

Han

Hello Jasem,

I had a look EKOS using ASTAP. Got it working in simulation and it looks all good! A few remarks:


1) De tolerance doesn't need to be in the settings. The default setting will work for all.

2) Update FITS. This option is default checked. For the moment, I would leave it unchecked to speed up solving.

3)You could add an option "Force large search windows" Default unchecked. Unchecked is "-speed auto", Checked is "-speed slow" . The hint would be "Check-mark this to improve solving reliability in some cases. Search window will be always large with overlap but will slow down solving."

4) I would add a remark: For noisy cmos cameras, image calibration prior to solving can be set in the ASTAP program.

For the remark in option 4), I'm in doubt if it's required to add. More then 99% of the users are happy without. But the possibility exist.

Han

In addition to previous message. The GUI remembers the last setting, so it is better to specify the following options always:

- Search radius [0-180 degrees] default 30
- Down sample [0..4] default 0 equals auto
- Speed [auto/slow] Default unchecked is "-speed auto" checked is "-speed slow"

So two check marks could be removed. Resulting in this following settings menu:

Han

Thank you for the feedback, Hans. It's not implemented as you suggested!

Tested it out a couple of nights this week. I admit I forgot I had it configured. It "just worked".

And wow - very fast solutions. Love it. Nice work Jasem & Hans.

It solves super fast for me too!
Agreed, great work by Jasem and Hans!

I ran into an issue with resolving a Ha image with 20 minutes exposure, which had a lot of stars - ASTAP was not able to solve.
It worked when I ran the solver from command line and added -s 30 parameter. I believe it reduces number of stars to be used for plate solving.

Hans,
Do you think we should include configuration for -s value in Kstars?

Alex

Thanks for the feedback.

Alex, 20 minutes exposure is pretty long. If used for syncing the mount, a exposure of 10 seconds using H-alpha filter would be enough and maybe 5 seconds using no filter. So 20 minutes is an outlier. ASTAP doesn't like heavily saturated images. It needs to be able to separate the brightest stars from faint stars. It will work on 5, 10 or 200 or 100x200 seconds images, but less good on a single 20 minutes exposure. This is different then Astrometry.net. Astrometry.net doesn't require the star intensity. You can solve a 1 bit image with Astrometry.net. ASTAP can't.

Sometimes in long exposures of the Milky-way containing many stars of the same magnitude it helps to increase the number of star setting to 1000, but it is better to reduce the exposure time. But in general you should have a success rate of 99.9% or better.

Since Jasem provided a possibility to add additional command, i would prefer to keep it as it is. You can also set it in ASTAP and save settings. But remember any setting change in ASTAP will be remembered (saved) after a solve.

Checklist for successful solving with ASTAP:

- images should be preferable raw and unstretched.
- Severe Image saturation should be avoided.
- Stars should be pretty round. (if not is they will be skipped as double stars or galaxies)
- Minimum of about 20 to 30 stars should be visible (this could be just above noise level. Exposure time of 1 or 2 seconds can work)
- Image height should be at least 1000 pixels.
- Image height in degrees should be defined within maybe 30% accuracy or better. (so in Ekos focal length, sensor size should be set correctly)


For the last point, if you open in ASTAP viewer an unsolved FITS image saved by Ekos, the image dimensions in degrees as shown in the ASTAP status bar should be correct. After a solve in the viewer the image dimensions will be updated and should be the same. If they change a lot after a solve in the ASTAP viewer, you better update the focal length, sensor size settings in EKOS.

Han

Thank you Han!
My 20 minutes image was not a sync image, it was an image from my previous session, which I used for setting the go to target.
I will try additional command and see how it goes.
Alex

Wow this is brilliant! Thanks Han and Jasem. Easy to set up and very fast!
The bushfires have put paid to any testing outside but it has just solved and slewed a loaded image (10 sec) in 3 secs! Using all default settings. I will definitely be using this in preference to astrometry.net.
Does anyone know if the polar alignment assistant uses ASTAP if this is enabled in EKOS or if it defaults back to astrometry.net. I will find this out when I can get outside but thought I would ask if anyone has used the polar alignment assistant while they have ASTAP enabled?

Cheers,

Mike

I wanted to share my findings using the new ASTAP, when it works it is super fast. But I can only get it to work about 40% of the time, the rest of the time it will not solve and just times out...when it works it’s done in a few seconds...but the rest of the time I just get a box come up in the top left corner of the screen and it just times out...well there is a number in that box that is counting up, so I assume that’s what it is...

What I have noticed is that if you point to random parts of the sky, it seems to work, but when I slew to a specific object such as a galaxy, cluster or Nebula, it will not work, and have to switch to Astrometry.net to get the solve...
Also I have tried with the simulators in Ekos, and it’s the same story with those too....

I have nearly 0 problems from ASTAP when solving anywhere/type IF you have the right settings for the Sky quality, Exposure length,noise etc on any types - my initial testing was with M31 and Polar stars all done fast any problems have been mine except for the odd glitch which Ham solved. I use DSLR so ASTAP has a few more hurdles to do ASTAP. When it fails I find Astrometry fails (even online version).. When it solves I say on the whole it solve in 1-5secs - on the same images all other Astrometry based Plate solving takes 100% longer. However Blind solving is slower (i.e. no RA/Dec) but then I for one will accept that as I dont do many blind solves. PLUS the big point it saves over 1GB of space compared to Index'd based solvers. I would say well done Han.