×

INDI Library v1.9.8 Released (29 Sep 2022)

Bi-monthly INDI Library released with new drivers and bug fixes.

10Micron Mount Modelling

  • Posts: 86
  • Thank you received: 10

Replied by Brent on topic 10Micron Mount Modelling

so does the Ekos mount model feature now build a 10micron model or not ?

thanks
2 years 5 months ago #55892

Please Log in or Create an account to join the conversation.

  • Posts: 83
  • Thank you received: 11
I’d love to know how to model inside Ekos. It looks like most of the features required are in the internal alignment Subsystem. When it builds points I can’t get it to honor my horizon masks. I also can’t figure out the driver workings… there’s a model building section in the alignment tab.

My main issue is Astap/astrometry aren’t as robust as the internal solver. The result is I need excessive exposures (30s) before I’m polar aligned and the mount won’t track well enough before that to allow for the exposures required. But, the internal solver works great. If I could do a small model I could swap over to MW4 to polar align using this model.

My guess is this feature died with MW4 supporting INDI.
6 months 4 days ago #83237

Please Log in or Create an account to join the conversation.

  • Posts: 83
  • Thank you received: 35

Replied by Hans on topic 10Micron Mount Modelling

> I’d love to know how to model inside Ekos.

I still propose you use MW4 for that.

> It looks like most of the features required are in the internal alignment Subsystem. When it builds points I can’t get it to honor my horizon masks. I also can’t figure out the driver workings… there’s a model building section in the alignment tab.
>
> My main issue is Astap/astrometry aren’t as robust as the internal solver.

Oh ? Maybe I can help there. For model building a good signal to noise ratio can be easily reached with strong binning, like 3x or 4x. And downscaling the image as well, I've tried up to 6x downscale and found 4x donwscale works well for my setups (F/12 and an F/5.4 systems). With this exposures of about 4 seconds suffice.

> The result is I need excessive exposures (30s) before I’m polar aligned and the mount won’t track well enough before that to allow for the exposures required. But, the internal solver works great. If I could do a small model I could swap over to MW4 to polar align using this model.
>
> My guess is this feature died with MW4 supporting INDI.

This confuses me. Do you think something got removed from MW4 or from Ekos ?
Last edit: 6 months 4 days ago by Hans. Reason: quoting system is completely broken
6 months 4 days ago #83239

Please Log in or Create an account to join the conversation.

  • Posts: 83
  • Thank you received: 35

Replied by Hans on topic 10Micron Mount Modelling

> so does the Ekos mount model feature now build a 10micron model or not ?

Not afaik.
6 months 4 days ago #83240

Please Log in or Create an account to join the conversation.

  • Posts: 83
  • Thank you received: 11
I was reading the original threads before MW3 existed. I think you were working on it weren’t you? There’s the alignment tab in the driver and it was supposed to use the IAS to build a model.. but I can’t find any directions.

With narrow fields of view like my C11, the internal solver is significantly faster than ASTAP. I can plate solve in 10s with Ekos… ASTAP requires 20-30s to generate enough stars and is super sensitive to focus.

But, if that tab should be deprecated… that’s fine too.
6 months 4 days ago #83244

Please Log in or Create an account to join the conversation.

  • Posts: 124
  • Thank you received: 28

Replied by Michael on topic 10Micron Mount Modelling

Hi,
my two cents to sort some of the questions out:
- MW3 used a interface to astrometry.net online for solving
- MW4 removed this, because of the long solving times and added the support of the local astrometry.net solver provided by ekos. This support is still there, but on some platforms (MAC) the binaries were removed (I think when stellar solver started).
- MW4 added the support of ASTAP as this solver is available on all platforms.
- I would love to support the internal solver, if there is any binary to call from external on all platforms. So far Rob had a lot to do to make the internal happen. Perhaps there is the opportunity to make this happen in future. (DBUS calls are unfortunately not the solution)


For the solving topic:
My experience is that all solvers work fast if they get a reasonable starting point to search from. The internal one is very efficient as it could solve on multiple threads (so it is somehow a brute force attack :-))

Michel
The following user(s) said Thank You: Chris Madson
6 months 4 days ago #83251

Please Log in or Create an account to join the conversation.

  • Posts: 83
  • Thank you received: 11
Hey Michael,

I'm definitely not saying anything bad about MW4. I'm frustrated with ASTAP; For instance, last night I had a run of 66 stars where 52 were successful. I started over and did another 66 star run (same points) and only 19 solved the second time around. Same settings as the first time. I don't remember having this issue with my widefield scopes so I'm assuming it's related to my field of view (.46').

I had this discussion with Rob when I swapped over and he didn't want to open it up. Apparently it's integrated into the modules and that's what makes it faster. You'd have to write MW4 into Ekos in order to make it work. I don't understand the programming aspect; it seems like you could 'ask' Ekos to run it for you similar to how you 'ask' ASTAP/Astrometry to do it. Shoot, you could even just make a 'ask' that tells it, "I'm looking for this from you, I'll wait".
6 months 3 days ago #83267

Please Log in or Create an account to join the conversation.

  • Posts: 124
  • Thank you received: 28

Replied by Michael on topic 10Micron Mount Modelling

Chris,
yes, the smaller the FOV there more challenging it might be. just for interest, could you post an image which took that long ?
Michel
6 months 3 days ago #83275

Please Log in or Create an account to join the conversation.

  • Posts: 83
  • Thank you received: 11
Yea, I'll send you the entire run on Github when I get home. I've started learning how to do the DSO track and it gave me good results last night in order to save the huge 100 point runs. I'll give it another go tonight and see if I can get a decent full run done.
6 months 3 days ago #83277

Please Log in or Create an account to join the conversation.

Time to create page: 0.645 seconds