12 replies to this topic

[Jackey]

Hello everyone! 
I developed an app capable of performing alignment the same way as the StarSense Autoalign accessory, just from the main camera and with a much more flexible distribution of alignment points. 

I believe the big benefit is that it uses the main optical axis and not the secondary guider scope. 
Below is a screenshot:

The way it works is that you connect your Camera and CPWI-compatible telescope, then below you define the alignment point distribution and click auto-align.
Using the aligning options requires you to get to the alignment menu in CPWI first. Then you don't select the alignment targets in CPWI but switch to my application and run the automated alignment procedure.

The "Solve and add current location" plate solves and adds the current point to the alignment model.

Auto Align function will automatically go through all the defined points, plate solve them, and add to the model if solving was successful. 
During the auto-alignment process, the dots will change color- the current dot turns yellow when the processing starts, and then turns red or green depending if plate-solving was successful.
The circle presenting point distribution is pointing south by default (as our telescopes do when we start alignment with CPWI) but the view can be flipped with the "Flip" checkbox if someone needs it oriented north. 

The azimuth shift option changes the azimuth degree, which can be useful if you have a mount on a wedge and a long optical train - in my case when the telescope is pointing exactly north on high azimuth the camera could hit the mount during slewing, so I decrease the azimuth by a few degrees so that my camera is safe. 
For non-ASCOM cameras, you can use the  "From Photo" tab to provide the image file rather than downloading it from the ASCOM camera. 

 

 

If you already have plate-solving results from other software, you can manually add the coordinates of the current point at which the telescope is pointing at that moment.

 

 

Plate solving is done using ASTAP, so you need to provide its path and parameters like search radius, exposure time, and FOV in the settings window. 

 

 

 

 

The application requires CPWI to be up-to-date, as it uses ASCOM SlewToAltAz command which was only recently added. I can work on support for older versions if needed, it would probably only require to calculate Ra/Dec coordinates out of alt/az, which should not be a big problem.  

I tested that recently on my Evolution mount with C8 and ASI1600MM on the back and it managed to align my telescope without any issues. I am able to align my telescope super precisely now when using it on a balcony with a very limited FoV.

As this is an early development version, please keep a close eye on your telescope when testing and be ready to push the emergency stop button in case it starts behaving the unexpected way. 
If someone would be interested giving this app a try and giving me some feedback, I would be very thankful. 
Please contact me in PM to obtain the executable.  

Keep up the good job guys!
Clear skies!
 

Attached Thumbnails

• 
• 
• 
• 

Edited by Jackey, 08 February 2024 - 09:02 PM.

[PeterAB]

Where do I go download this software?   It looks interesting.

 

Peter

[Jackey]

I think it will be the best to add a google drive link here:
https://drive.google...?usp=drive_link
I am accepting all the requests to access as soon as I notice
Good luck with your alignment!

[Mitchell Duke]

Count me in! I will test it on the CGXL.

[Jackey]

Thank you all for the interest so far! I'm looking forward to hear about your experience

[Mitchell Duke]

Thank you all for the interest so far! I'm looking forward to hear about your experience

I tried to download it just now but its still asking for your aporoval.

[Mitchell Duke]

Got it! I will try it when the skies clear.

[Jackey]

Thank you everyone for the interest, I hope you have better weather than I do;) 

[ultramaddie]

Hi Jackey! This looks great, I can’t wait to give it a shot with my AVX. Can you upload the code to GitHub as well?

[ButterFly]

Very similar to PlateAlign (latest version here).

 

Add some kind of a star chart to help users see where the selected points are on the sky.  It helps to see where the obstructions are, so users don't waste time trying to solve in a tree!

[Jackey]

Thanks guys for all the feedback so far, I can consider adding some Sky Chart to see where the points are located on the sky  
The work on this project is currently paused, but I will probably come back to it during the summer.  
Would you consider it better to display the aligment points location directly within the app on top on the sky chart, or rather send them to the planetarium like SkySafari/Stellarium/Whatever else?
The last version published seems to be fully functional, worked for me every time last summer. I was using that mostly from a balcony with a very limited view (like 90deg) and I could use the software to have a dense distribution of points just in that small visible part of the sky, resulting in quite a good and efortless  alignment while I had time to enjoy my coffee  
I think a good improvement would be to use ZWO ASI SDK rather than ASCOM drivers to get the image from the camera, actually the biggest time waste right now is transmitting the image thorugh the COM port.  If anyone of you have some ready code samples to share, that could increase the alignment speed significantly. 

[Jackey]

@ultramaddie 
I host my private gitlab instance, I'd consider making it more public after some more code cleanup  
 

[ButterFly]

Would you consider it better to display the aligment points location directly within the app on top on the sky chart, or rather send them to the planetarium like SkySafari/Stellarium/Whatever else?

Up to you really.  Not all software can accept a list of points like that.  For SkySafari, for example, you can't even make an observing list, which the user would then have to display themselves.  If an object isn't in SS's catalog, you can't display it.  It's unlikely that you will find something at a random RA/Dec.  You could try to get close, but it's not really worth it in the end.  What will the user do with that info?

 

 

I think a good improvement would be to use ZWO ASI SDK rather than ASCOM drivers to get the image from the camera, actually the biggest time waste right now is transmitting the image thorugh the COM port.  If anyone of you have some ready code samples to share, that could increase the alignment speed significantly. 

For focusing, yes.  For everything else, it's a marginal gain.

 

A much better gain can be had by not saving the image to disk in the first place.  With astap, one needs to save the image to disk (which involves bit flipping in windows for a fits file).  Then, astap reads into memory that same image (which again involves bit flipping in widows for a fits file).  That's a horrid waste of time.  A flip plus write, then a read plus flip, just to get where you started is ... .

 

PlateSolve3 lets one just work directly with the image in memory.  However, the licensing for that is non-existent, so you can't really publish something that makes use of that (or use it yourself even).

 

You can save much more time if you simply work with image from memory.  Astap is open, so you can use the parts of astap that solve, reworking them for your language of choice.  The plus side is that if the user has astap installed, it's the same database, so no need for yet another several gb source for solving.  Another option is to use tetra based solvers, with the down side of the standard database being insufficient for most scope/chip combos, so not only do there need to be yet another database, but a custom one that user has to create.

 

My latest project, building off PlateAlign, is more of a refinement aid than anything else.  Rather than merely sending the current RA/Dec to the model, it sends another goto to the intended target after the sync, and iterates to within my set tolerance, just like imaging software ordinarily does.  It sends an audio cue to let me know to back off the eyepiece.  On most nights, I simply let the first goto be off by however much it's off, let it iterate, then carry on with observing.  After the third or fourth goto, it rarely needs to iterate.  To integrate with outside planetarium software, I send out a spoofed alpaca device containing one telescope.  When the planetarium software sends a goto to the spoof, it catches the coordinates, then iterates if needed after the first goto.

 

It's essentially building the model on the fly that way.  For CPWI, it's adding alignment points.  For my other mounts (synscan and servocat), it adds local align points.  The model just gets better the more things I go to, without my having to wait.
 

The reason I would never release this is because it can move a scope without the user specifically sending that particular goto (it iterates, after all).  I simply don't trust people on a ladder to read or heed a giant warning saying - risk of death - or the like.  I know I'm not an idiot, so I'll keep using it just fine without dying.  I don't trust other idiots to not die, and it's not worth my time to dismiss vexatious suits from people who don't read warnings.