Exciting! One thing I love about opensource development is meeting authors along the way. Thank you for your contribution!
Do you know if building a mount model will effect ratios and therefore improve tracking rates in those mounts that have it? Or is it only going to improve slews?

cheers!
Ray

I've got a skywatcher az eq 6 gt mount. When i buid a mount model can i store it in Ekos to use the next day? Or is it stored in de mount?

Eqmod model work as expected and really improve the pointing accuracy.

I have a mobile HEQ5 and until last year I always run an automated script at the beginning of the night to add about 15 points in the model.
This work fine but this is still insufficient to reach the less than 1 arcmin precision necessary to realign a frame with those of the previous night. So plate solving is still necessary to precisely center the target.
And I find that plate solving and sync work the same if you are at ten arcmin or at one degree of the target.
So now I not do any model, "Sync Mode" is set to "Standard Sync" and "Alignment Mode" to "Nearest Point", the first plate solving correct the home error and then it work fine.

Probably a better mount on a fixed pier can get more benefit of the mount model.

You can store the model data in your computer.
In the Eqmod driver Align tab you have a "Write file" and "Load file" button to save and reload the model.
The file name is set in the field "Data file", the default is ~/.indi/AlignData.xml

A small comment of this excellent feature for which I need your opinion...
In terms of robustness of a permanent setup, adding sync points is a weakness: it is better to start with a deterministic mount model, and use it without modification.
So that when something goes wrong during the alignment procedure, the ability of the setup to abandon the current target and slew to the next is preserved. We perhaps all got in the situation where the alignment enters a weird loop because of a wrong sync point.
Thus the method is, once the model is loaded, to always use a differential goto during alignment, and disable the model reset on alignment failures or job startup.

-Eric

Hi!

I have a Losmandy G11, and I thought that this mount could not benefit from the Mount model in Ekos. Is there a way to upload the model to the mount, or is it jus saved in Ekos?

My understanding is that if it is just preserved in Ekos, then why would the mount type matter?

Yes, as you can hear, I don't really understand what the "Mount model" does. Even now....:/

Magnus

Hi Magnus,

So I believe you misunderstand. I wrote the Mount Model feature in the align tab specifically because I have a Losmandy G11 and it really benefits from multiple align points. I'm not sure how many the G11 can take, but the more points you get the better the pointing model. I don't always use it, but I like that i have the feature. Ekos does not save any alignment data when it does any plate solve in the alignment module, it simply does plate solves and tells the mount where it is located. What the mount or driver does with that information is up to it.

For the Losmandy G11, it does all the alignment points and modeling internally. There are two commands that the G11 takes, the sync command and the align command. You need to make sure that the Gemini system is getting any additional alginment points from plate solving as an "align" command so that it updates the pointing model instead of shifting the model to do a sync. There is a menu option that you need to change in the hand paddle to make sure it is getting the right command. Once it builds a pointing model, it can retain that model from night to night, or you can replace the model with a new one. You cannot export the model to a computer or transfer a model to the mount as far as I know.

For EQMod, it can also build very complex models, but the model is built in the driver, not in the mount. But again this is not handled in Ekos, it is handled in the INDI driver. Since here the model is built on your computer, you can save and load the model using the INDI control panel. As I have said before, I don't know everything about the EQMod driver so I can't speak more to that.

Thanks,

Rob

I believe I have encountered what TallFurryMan just referred to as a 'weird loop'. While plate solving, occasionally it struggles to actually make the necessary corrections between iterations of slew to target. It is as if it is cluttered with alignment points in the vicinity. I have to clear the alignment model to recover from that.
I have run the mount model routine before. But I didn't see much benefit, especially if it later struggled to slew to target while plate solving.

My recommendation would be to not use "slew to target" but instead to just "sync" when building a model. As I mentioned before, each time I plate solve in a sky location it improves the model because it adds another alignment point. All I do is just plate solve at several locations around the sky and it gets better with each point. I believe you are correct that it is a bad idea for the model to have a bunch of alignment points all clustered closely together. They should be more spread out. You can avoid this by not trying to "slew to target" when building the model. It really doesn't need to get the target in the frame, it just needs a bunch of plate solved points spread around the sky for the alignment. I stopped trying to "slew to target" about 2 years ago, and ever since I think it has worked much better.

Thanks,

Rob

Rob, do you have a permanently mounted telescope? Or do you set up and take down your telescope every night? I do the latter and need to do a polar alignment every time I set up my telescope. Setting it up and taking it down means I cannot use a mount model like you do, or I have to rebuild it every time. Do you concur?

Thanks, Wouter

rlancaste: Sorry for the confusion. I did use sync for the mount model. Slew to target is only used while getting to my imaging subject.

This is a very good question! Obviously building a really good pointing model that you can use from night to night would only work with a permanent setup. Alas, I do not have an observatory or a pier. So yes, I have to build my pointing model fresh every time I observe. I always try to get as many alignment points in there as I can, but of course there is a limited time in an evening. So usually what I try to do is add as many alignment points as I need to until it gets reasonably accurate slews, and then I call it good. Then later I might add more points if i'm still not happy with it. Usually that ends up being a minimum of 3 points, but often somewhere between 5 and 10 points before the night is done.

Thanks,

Rob