I would like some suggestions on the best way to build a model using plate solving.

I always coldstart because my mount is not in a fixed observatory.

My understanding is that the first alignment behaves like a synchronisation on the mount and subsequent alignments contribute to building a model.

When I do this to go to an object there are often 4 or more alignment points added that are very close to each other as the mount homes in on the object.

I may be wrong but I believe this can introduce issues for the mount model which may on occasion cause it to go off somewhere completely random.

My thought is that it may be useful to have an option to only update the alignment model for the first alignment point after a go to and to ignore any of the fine alignment movements.

Does anyone have an suggestions or information on this?

Paul

Paul,

I am in a similar situation as you. My setup isn't fixed. I set it up every night I go imaging and I take it inside once the imaging session is done. The only difference is that I use SkyWatcher mounts (most often an HEQ5 mount and occasionally an EQ6-R mount).

I never worry about the mount model. Plate solving is all I need. I always start with a polar alignment, which uses plate solving as well and each solution contributes to the mount model. Once done I either make the mount slew to an object by right clicking it in KStars and select <My Mount> -> Goto or a program a sequence of jobs in the sequence tab. Occasionally the mount has problems pointing at the object within the specified 30 arcsec but this happens very rarely. If and when it happens, I click the Clear Model button in the Mount tab and clear the pointing solutions on the Align tab by clicking the red button at the bottom of the tab. With the sequencer I typically enable the "Reset Mount Model Before Starting Each Job" option in the Ekos settings in the KStars configuration options. So far, my mount has never missed a target.

I hope this answers your questions.


Clear skies, Wouter

Just to clarify, all Ekos does is SYNC. What the mount does with this information is up to the mount internal logic. The only exception for this is EQMod and any mount drivers that utilize INDI Alignment Subsystem . When this system is used, then it actually builds and stores the mount model in INDI.

Wouter,
I have automated my set up so I am often not in a position to manually do anything, usually I am asleep
I have not used the option to reset the mount model before starting each job, partly because of uncertainty about what it does.
I guess my major issue is not being clear on what the mount model is doing. Is is building a model in the mount (if this is supported) or just within Ekos?
Paul

Jasem,
With the Gemini 2 there are two possible actions, a Synchronisation which aligns the mount with the sky and an Alignment to build a model in the mount.
If INDI is only sending a SYNC then it will not be building any model in the mount but keep telling the mount where it is pointed.
When you say it is up to mount internal logic, as there a two different commands that the mount can obey is there any control over what the mount receives?
Does it do the same thing if I build a model?
When the Gemini 2 has a model it can be very accurate with Go To's so what is the preferred method to update the mount model?
Paul

So Ekos does not know about that. Maybe some sort of switch can be toggled in the INDI driver that makes it interpret the SYNC as an alignment point? Once done, you can toggle this switch off again and it would use SYNCs as normal after the model is built. That would be one way to do it.

Jasem,

Sorry I haven't got back to you earlier, the notifications for this do not seem to be working.

At least you have cleared up in my mind how Ekos works.

Each time you do a solve it sends a Synch command to the mount, nothing more and nothing less.
The mount will just update its location but not build any model.

It all seems to work but I have been having occasional issues with the mount getting confused so thought it was corrupting the model with too many points close together.
It appears this cannot be the case.

Paul

I have a further thoughts on this.
As the INDI alignment module only does a Synch with the Losmandy Gemin, Does it do a Synch when 'Slew to Target' is used?
I assume it does if a synch has not already been done.
If the build a mount model is used, does it or can it be configured to send an 'alignment' command or 'synch', if not the first time, to actually update the mount model?
Paul

Hi!

I don't have an answer for you question, but antother question I also have a G11 with Gemini 2, and I never care about a mount model. I don't see what model would help me with, when running Ekos.

You mention some cases when it seems to be confused, or lost - can you describe them more?

Magnus

Magnus,

By building a model I was hoping the goto's would be more accurate, even though it works reasonably well.
Some of the mount abilities are also more reliable with an accurate model.

I have not really tracked down the mount random pointing issues as they happen only rarely (and it has been a long run of cloudy weather lately), but I don't get them when I do everything through the hand controller and build a model.
My belief was that the mount was becoming confused but now I am thinking the INDI model is getting confused.
But then again I rarely do visual so maybe I just don't spend enough time doing things with the hand controller.

Paul

Hi!

Yes, well, I am curious about the potential value of a model. On the Losmandy mailing list, there are numerous issues with the Gemini2 modelling and pointing behaviors, and a lot of people seem to just skip modelling and rely on plate solving instead. As do I - but maybe I am missing something I also experience some weird positioning at times (like plate solving the NGC4565, finding it is within 1 degree of the target, then slewing to the south celestial pole almost... or at lest to a mount limit), but so far I have attributes them to quite rare odd behaviors in ASTAP or Astrometry.net, depending on which I use for plate solving. Not been able to produce logs either.

So if you have some input on the value of a model, I'd be very interested. As well as in some more info about those rare weird slews

My experience of just plate solving is that when the mount is aligned (doing a sync) with a good PA, it slews to within 1-2 degress of targets anywhere. Does that resonate with your experience, or is it worse?

Magnus

Magnus,

I find the initial goto varies a lot depending on various factors but following a synch it is generally much better, probably as accurate as you find.
I also find the alignment usually takes 5 or more steps to get to a final position.
If I build a good model I find it is much more accurate with goto but as I usually cold start I would need to do it frequently.
It is interesting to hear about it deciding to go somewhere odd after the plate solve, it maybe something similar I am experiencing.
There are two things I do when something like that happens, park the mount and cold start, but maybe it is something other than the mount getting corrupted, like you I am unsure and generally do not want to spend time trying to understand it properly - clear nights are a rare treat lately.
Paul