Just what is a Dec Arc model?


ap@CaptivePhotons.com
 

On Mon, Oct 31, 2022 at 12:13 AM, Ray Gralak wrote:
Is the DEC Arc tracking box needed, doesn't it just KNOW if there's a dec arc model (conversely shouldn't it use DEC Arc
automagically if in the arc where there's data even if it's a full sky model?)
No. That's not the way it works. I've said this before there is no concept of a "Dec Arc Model" in the point map files that APPM creates. They are just data points which can be used to independently and simultaneously create All-Sky and Dec Arc Tracking models. Let me say that again... the same set of sky points can be used for either model, and you have a choice to use one or the other. And, BTW, models only exist at runtime in APCC.
Ray, 

I hope you don't mind I removed this from the "initialize APPM" thread and moved it to the ap-gto forum to ask for a bit of insight here. 

I do know you have said the above before, but I remain a bit confused, and wonder if you can help provide insight into three specific scenarios that have come up for me.  I think it might help others as well. Let me set the stage: 

From your description above, it sounds like the DEC Arc modeling (note ending) is basically a different algorithm than the All Sky modeling operating on data points.

So I might collect a partial set of points across a DEC Arc in the sky, and not the whole sky.  Let's call those data points "A" (I'm avoiding the term model at this point). 

Or I might build a full set of points covering the whole sky.  Let's call that data points "W". 

Now I load data points "W", and check pointing and tracking corrections but not check DEC Arc.  That results in calculations using the (let's call it) classic modeling algorithm that builds a whole sky model.  Right so far? 

But if on data points "W" I also check the DEC Arc box, this results in it using a different calculation, the new-coke type calculation for DEC Arc using the same data.  Right? 

If I have a whole sky set of data, how do I decide whether it is a good idea or not to check that DEC Arc box?   Is it purely a try it and see, or are there criteria I should use to decide? 

Inverse question - I go back and use the points "A" which cover only the DEC Arc of the sky.  I check pointing and tracking.  I don't check DEC Arc (there's a reason - more in a moment).  What happens?   Is this bad, or is it smart enough to either not use the data, or to use it only where relevant? 

I ask this last question because of automation -- if you tell NINA to build a DEC Arc model, and no model is currently active in APPC at all, when APPM is done collecting that arc of the sky, something (I think APPM) loads it in APPC and activates ONLY the Pointing and tracking check boxes.  In terms of automation, there does not appear to be a way to check the DEC Arc box, but it does check the other two.  You need to check it by hand.

If one forgets that manual step (as I suspect people might), what happens?  Nothing, something bad, or something good while on the Arc?  Is there a cooperative way between you and Dale that we could have his "Build Dec Arc" instruction result in that check box being activated so one doesn't have to manually intervene? 

Final related question: I have points "A" loaded with just the DEC Arc in the sky, I have all three check boxes checked. Because I'm sleep deprived or not paying attention, I start imaging in a completely different area of the sky way off the ARC leaving these loaded and checked.  I'm sure nothing good happens (i.e. it can't help me track), but does anything bad happen, or does it "notice" that it has no data in that area and just turn off tracking/pointing corrections.  My vague recollection is you fixed this to ignore the points when it does not have data, is that correct?  So nothing good happens but nothing bad happens? 

Sorry to be OCD about this, but still trying to get my head around how the check boxes operate with respect to the data points that may or may not be loaded. 

Linwood 


Ray Gralak
 

Hi Linwood,

Technically there is no such thing as a "Dec Arc model" in APCC, but that name has taken on a life of its own. Without going into too much detail, the Dec Arc Tracking algorithm is different than the all-sky model that other mounts use even when they are doing what may be considered a "Dec Arc model".

In regards to NINA, the user needs to manually enable the Dec Arc Tracking checkbox if it is available to them.

NOTE: If a user has an older license, then the Dec Arc Tracking algorithm may not be available unless they renew their subscription. And before anyone asks, that feature will be perpetually available with no further need to purchase a subscription the following year unless they want the new features available at that future time.

That said, if you don't mind , I'll try to address your questions as much as I can in a private email later when I have some time. This goes into the "theory of operation" that I don't think should be shared publicly because I believe that APCC is the only software using a different algorithm like this.

-Ray

-----Original Message-----
From: main@ap-gto.groups.io [mailto:main@ap-gto.groups.io] On Behalf Of ap@...
Sent: Monday, October 31, 2022 6:03 AM
To: main@ap-gto.groups.io
Subject: [ap-gto] Just what is a Dec Arc model?

On Mon, Oct 31, 2022 at 12:13 AM, Ray Gralak wrote:


Is the DEC Arc tracking box needed, doesn't it just KNOW if there's a dec arc model (conversely
shouldn't it use DEC Arc
automagically if in the arc where there's data even if it's a full sky model?)

No. That's not the way it works. I've said this before there is no concept of a "Dec Arc Model" in the point
map files that APPM creates. They are just data points which can be used to independently and simultaneously
create All-Sky and Dec Arc Tracking models. Let me say that again... the same set of sky points can be used for
either model, and you have a choice to use one or the other. And, BTW, models only exist at runtime in APCC.

Ray,

I hope you don't mind I removed this from the "initialize APPM" thread and moved it to the ap-gto forum to ask for a
bit of insight here.

I do know you have said the above before, but I remain a bit confused, and wonder if you can help provide insight
into three specific scenarios that have come up for me. I think it might help others as well. Let me set the stage:

From your description above, it sounds like the DEC Arc modeling (note ending) is basically a different algorithm
than the All Sky modeling operating on data points.

So I might collect a partial set of points across a DEC Arc in the sky, and not the whole sky. Let's call those data
points "A" (I'm avoiding the term model at this point).

Or I might build a full set of points covering the whole sky. Let's call that data points "W".

Now I load data points "W", and check pointing and tracking corrections but not check DEC Arc. That results in
calculations using the (let's call it) classic modeling algorithm that builds a whole sky model. Right so far?

But if on data points "W" I also check the DEC Arc box, this results in it using a different calculation, the new-coke
type calculation for DEC Arc using the same data. Right?

If I have a whole sky set of data, how do I decide whether it is a good idea or not to check that DEC Arc box? Is it
purely a try it and see, or are there criteria I should use to decide?

Inverse question - I go back and use the points "A" which cover only the DEC Arc of the sky. I check pointing and
tracking. I don't check DEC Arc (there's a reason - more in a moment). What happens? Is this bad, or is it smart
enough to either not use the data, or to use it only where relevant?

I ask this last question because of automation -- if you tell NINA to build a DEC Arc model, and no model is
currently active in APPC at all, when APPM is done collecting that arc of the sky, something (I think APPM) loads
it in APPC and activates ONLY the Pointing and tracking check boxes. In terms of automation, there does not
appear to be a way to check the DEC Arc box, but it does check the other two. You need to check it by hand.

If one forgets that manual step (as I suspect people might), what happens? Nothing, something bad, or
something good while on the Arc? Is there a cooperative way between you and Dale that we could have his
"Build Dec Arc" instruction result in that check box being activated so one doesn't have to manually intervene?

Final related question: I have points "A" loaded with just the DEC Arc in the sky, I have all three check boxes
checked. Because I'm sleep deprived or not paying attention, I start imaging in a completely different area of the
sky way off the ARC leaving these loaded and checked. I'm sure nothing good happens (i.e. it can't help me
track), but does anything bad happen, or does it "notice" that it has no data in that area and just turn off
tracking/pointing corrections. My vague recollection is you fixed this to ignore the points when it does not have
data, is that correct? So nothing good happens but nothing bad happens?

Sorry to be OCD about this, but still trying to get my head around how the check boxes operate with respect to the
data points that may or may not be loaded.

Linwood


ap@CaptivePhotons.com
 

Thanks, Ray... 


On Mon, Oct 31, 2022 at 10:50 AM, Ray Gralak wrote:
In regards to NINA, the user needs to manually enable the Dec Arc Tracking checkbox if it is available to them.

If you and Dale ever work again on the interface between NINA and APCC/APPM, it would be very useful to have a function to enable/disable that (and probably the other two also). 

From a NINA perspective, there are two separate instructions - build an all-sky model, and build a dec-arc model. I think people are using the latter in particular for portable imaging where they kick it off early then it falls into a sequence to image with that model loaded.  It may be less than obvious that they need to manually intervene, though clearly this discussion will help get the word out.

And of course if they ALWAYS do a DEC Arc model from NINA and the box is always checked, it's moot as they change targets. 

Thanks again,

Linwood


Dale Ghent
 

Just as a general reminder to everyone, full instructions on how to use the Astro-Physics Tools plugin in NINA for automated modeling, including APCC and APPM configuration topics, are on my website:

https://daleghent.com/astro-physics-tools

On Oct 31, 2022, at 10:50, Ray Gralak <iogroups@...> wrote:

Hi Linwood,

Technically there is no such thing as a "Dec Arc model" in APCC, but that name has taken on a life of its own. Without going into too much detail, the Dec Arc Tracking algorithm is different than the all-sky model that other mounts use even when they are doing what may be considered a "Dec Arc model".

In regards to NINA, the user needs to manually enable the Dec Arc Tracking checkbox if it is available to them.

NOTE: If a user has an older license, then the Dec Arc Tracking algorithm may not be available unless they renew their subscription. And before anyone asks, that feature will be perpetually available with no further need to purchase a subscription the following year unless they want the new features available at that future time.

That said, if you don't mind , I'll try to address your questions as much as I can in a private email later when I have some time. This goes into the "theory of operation" that I don't think should be shared publicly because I believe that APCC is the only software using a different algorithm like this.

-Ray

-----Original Message-----
From: main@ap-gto.groups.io [mailto:main@ap-gto.groups.io] On Behalf Of ap@...
Sent: Monday, October 31, 2022 6:03 AM
To: main@ap-gto.groups.io
Subject: [ap-gto] Just what is a Dec Arc model?

On Mon, Oct 31, 2022 at 12:13 AM, Ray Gralak wrote:


Is the DEC Arc tracking box needed, doesn't it just KNOW if there's a dec arc model (conversely
shouldn't it use DEC Arc
automagically if in the arc where there's data even if it's a full sky model?)

No. That's not the way it works. I've said this before there is no concept of a "Dec Arc Model" in the point
map files that APPM creates. They are just data points which can be used to independently and simultaneously
create All-Sky and Dec Arc Tracking models. Let me say that again... the same set of sky points can be used for
either model, and you have a choice to use one or the other. And, BTW, models only exist at runtime in APCC.

Ray,

I hope you don't mind I removed this from the "initialize APPM" thread and moved it to the ap-gto forum to ask for a
bit of insight here.

I do know you have said the above before, but I remain a bit confused, and wonder if you can help provide insight
into three specific scenarios that have come up for me. I think it might help others as well. Let me set the stage:

From your description above, it sounds like the DEC Arc modeling (note ending) is basically a different algorithm
than the All Sky modeling operating on data points.

So I might collect a partial set of points across a DEC Arc in the sky, and not the whole sky. Let's call those data
points "A" (I'm avoiding the term model at this point).

Or I might build a full set of points covering the whole sky. Let's call that data points "W".

Now I load data points "W", and check pointing and tracking corrections but not check DEC Arc. That results in
calculations using the (let's call it) classic modeling algorithm that builds a whole sky model. Right so far?

But if on data points "W" I also check the DEC Arc box, this results in it using a different calculation, the new-coke
type calculation for DEC Arc using the same data. Right?

If I have a whole sky set of data, how do I decide whether it is a good idea or not to check that DEC Arc box? Is it
purely a try it and see, or are there criteria I should use to decide?

Inverse question - I go back and use the points "A" which cover only the DEC Arc of the sky. I check pointing and
tracking. I don't check DEC Arc (there's a reason - more in a moment). What happens? Is this bad, or is it smart
enough to either not use the data, or to use it only where relevant?

I ask this last question because of automation -- if you tell NINA to build a DEC Arc model, and no model is
currently active in APPC at all, when APPM is done collecting that arc of the sky, something (I think APPM) loads
it in APPC and activates ONLY the Pointing and tracking check boxes. In terms of automation, there does not
appear to be a way to check the DEC Arc box, but it does check the other two. You need to check it by hand.

If one forgets that manual step (as I suspect people might), what happens? Nothing, something bad, or
something good while on the Arc? Is there a cooperative way between you and Dale that we could have his
"Build Dec Arc" instruction result in that check box being activated so one doesn't have to manually intervene?

Final related question: I have points "A" loaded with just the DEC Arc in the sky, I have all three check boxes
checked. Because I'm sleep deprived or not paying attention, I start imaging in a completely different area of the
sky way off the ARC leaving these loaded and checked. I'm sure nothing good happens (i.e. it can't help me
track), but does anything bad happen, or does it "notice" that it has no data in that area and just turn off
tracking/pointing corrections. My vague recollection is you fixed this to ignore the points when it does not have
data, is that correct? So nothing good happens but nothing bad happens?

Sorry to be OCD about this, but still trying to get my head around how the check boxes operate with respect to the
data points that may or may not be loaded.

Linwood





Howard Hedlund
 

This looks like a great thread for some clarification and clean-up of our language.  Consistent  language is vital to avoiding confusion.
  • The theory behind the Dec-arc approach is that it concentrates the data used to model along the natural arc in declination that the object appears to trace across the sky.
  • The Dec-arc approach suffers the same naming pitfall as the all-sky approach. The biggest offender is the word *model*.
  • A model is a predictive set of algorithms that take existing real data, and apply that data to predict future outcomes.  Numerous algorithms are employed to optimize results.
  • To acquire that data, we employ a technique we call *mapping*.  The better the data we map, the better the predictive value.
  • We, as users, never see the model.  We don't create the model.  We do, however, collect the data that the model will employ, and we do this by *mapping* points in the sky.
  • APPM stands for Astro-Physics Point Mapper.  APPM is simply the data collector for APCC.  APPM allows you to strategize your data collection to best serve your individual needs. 
  • Our observing and imaging goals determine the best approach to mapping and using the data.
    • Are you worried about pointing?  Create an all-sky mapping in APPM and then go ahead and bounce from target to target.  
    • Are you tracking just one or two objects over the course of a night?  Then use APPM to map the relevant individual arcs in declination that you will be following as you track.
  • All we do is to map.  What APCC does is to take our mapped data and build a model.
HKH


Mark Tenboer
 

Regarding the Dec-Arc check box in APCC, I saw a post that said that it's purpose is to correct for refraction.  However, there is already a check box that says "Correct for Refraction in the Point Model box that contains all the statistics.  Can someone clarify this ?
Also:
1.  Should the Dec-Arc box be checked only if there is a model created in APCC -  AND -  only if the targets to be tracked that night will travel through the mapped area?
2.  What happens if the box is checked (and potentially the other Pointing and Tracking boxes are checked) but the target will NOT travel through the mapped area ?  Does that cause worse tracking?
2.  Does the Dec-Arc check box only apply to a Dec-Arc map, or does it's algorithm also benefit an All Sky map?

I might not be using the word Map correctly here (vs model), but my questions are only intended to understand when the Dec-Arc check box should be checked and what happens if is checked when it shouldn't be.

Thanks
Mark


 

HI Mark

If you are referring to this checkbox?
image.png

That checkbox is for enabling the Dec Arc Tracking algorithm.  It's active when you have a pointing model loaded (the model would be created in APPM). it can be used with any type of sky map, but it's better when you have a denser map of points. 

But as you pointed out there is a separate "correct for refraction" checkbox that is used in any model and will correct for refraction when enabled. 

image.png


Also:
1.  Should the Dec-Arc box be checked only if there is a model created in APCC -  AND -  only if the targets to be tracked that night will travel through the mapped area?

Dec Arc box operates on a model that is created by a map. So yes, you should have a model created and active. 

If you do have an all sky model, your targets will of course be within the model and dec arc tracking will operate on it

If you have a limited dec arc map (one that is specific to a target) and you are imaging outside of that map, I don't know for sure what happens in this case. I think the model is not used, but Ray or AP would know for sure

2.  What happens if the box is checked (and potentially the other Pointing and Tracking boxes are checked) but the target will NOT travel through the mapped area ?  Does that cause worse tracking?

see above re: model may not be used. But generally you don't want to be imaging outside of your mapped area. If you think that may happen, it's probably better to do an all sky model

2.  Does the Dec-Arc check box only apply to a Dec-Arc map, or does it's algorithm also benefit an All Sky map?

It can be used with either all sky or dec arc map. That's a great question because the similarity of names can sometimes be confusing here. 

hth

Brian





On Wed, Nov 2, 2022 at 5:34 PM Mark Tenboer <mtenboer@...> wrote:
Regarding the Dec-Arc check box in APCC, I saw a post that said that it's purpose is to correct for refraction.  However, there is already a check box that says "Correct for Refraction in the Point Model box that contains all the statistics.  Can someone clarify this ?
Also:
1.  Should the Dec-Arc box be checked only if there is a model created in APCC -  AND -  only if the targets to be tracked that night will travel through the mapped area?
2.  What happens if the box is checked (and potentially the other Pointing and Tracking boxes are checked) but the target will NOT travel through the mapped area ?  Does that cause worse tracking?
2.  Does the Dec-Arc check box only apply to a Dec-Arc map, or does it's algorithm also benefit an All Sky map?

I might not be using the word Map correctly here (vs model), but my questions are only intended to understand when the Dec-Arc check box should be checked and what happens if is checked when it shouldn't be.

Thanks
Mark




Chris White
 

On Mon, Oct 31, 2022 at 10:50 AM, Ray Gralak wrote:
Technically there is no such thing as a "Dec Arc model" in APCC, but that name has taken on a life of its own.
If the semantics are important then the nomenclature within the software should be amended.

When I implement a "Dec Arc Tracking Algorithm" in APCC I do so in the Pointing Model tab.  When I load a "Dec Arc Tracking Algorithm" for APCC to use I do so by clicking the Model button.  The APCC documentation PDF when describing how to create and load a point map refers to the creation of a Model and provides instructions on verifying the Model.

The only place where the word Model is left out is where the check box is for enabling Dec arc tracking.  I'm guessing that the way that Dec Arc Tracking is implemented within APCC utilizing the Model that is in place, is where the confusion arises.  Maybe the inclusion of a separate Tab for Dec Arc Tracking would make this clearer for those of us communicating our questions and issues.  I have seen you correct people multiple times about referring to it as a Dec Arc Model, so I assume that this is important for the understanding and usage of the software.

Thank you Ray.


DFisch
 

Chris, well said and thanks for identifying the likely root of my misunderstanding. Tom Fischer

On Thu, Nov 3, 2022 at 09:56 Chris White <chris.white@...> wrote:
On Mon, Oct 31, 2022 at 10:50 AM, Ray Gralak wrote:
Technically there is no such thing as a "Dec Arc model" in APCC, but that name has taken on a life of its own.
If the semantics are important then the nomenclature within the software should be amended.

When I implement a "Dec Arc Tracking Algorithm" in APCC I do so in the Pointing Model tab.  When I load a "Dec Arc Tracking Algorithm" for APCC to use I do so by clicking the Model button.  The APCC documentation PDF when describing how to create and load a point map refers to the creation of a Model and provides instructions on verifying the Model.

The only place where the word Model is left out is where the check box is for enabling Dec arc tracking.  I'm guessing that the way that Dec Arc Tracking is implemented within APCC utilizing the Model that is in place, is where the confusion arises.  Maybe the inclusion of a separate Tab for Dec Arc Tracking would make this clearer for those of us communicating our questions and issues.  I have seen you correct people multiple times about referring to it as a Dec Arc Model, so I assume that this is important for the understanding and usage of the software.

Thank you Ray.

--
TJF MOBILE


Mark Tenboer
 

Thanks Brian.  I'm hoping Ray or Howard can pipe in and answer #2
2.  What happens if the Dec-Arc box is checked in the Tracking Correction Status section (and also the other Pointing and Tracking boxes on that screen are checked) but the target will NOT travel through the mapped area ?  Does that cause worse tracking for that target, or does APCC suspend the model adjustments?

The reason I ask is that once you make a Dec-Arc map/model for a target, the model stays in place in APCC indefinitely until it is replaced with another.  For the next target that night, I might decide there isn't enough time left to make a new Dec Arc map and so I will track without it.  However, the model for target 1 is still active in APCC and the check boxes are checked.  Does APCC disregard the model in place if and when the target travels through an area outside the map?  

Mark


 

Hi Mark

>>>2.  What happens if the Dec-Arc box is checked in the Tracking Correction Status section (and also the other Pointing and Tracking boxes on that screen are checked) but the target will NOT travel through the mapped area ?  Does that cause worse tracking for that target, or does APCC suspend the model adjustments?

I can offer a couple clarifications

First, the pointing corrections are always done using the all sky "algorithm" (which is there and active, but you don't see it). Tracking outside of the points would also revert to the all sky model. However - as you may point out and rightly so - there are no points outside of your dec arc model, so if you anticipate imaging outside of the dec arc, you may want to use an all sky model, or at least have one ready and available (even something simple like a 20 point model)

 I can't speak if it produces indifferent, better or worse results. I think some experimentation would probably give you a good answer here



On Thu, Nov 3, 2022 at 1:56 PM Mark Tenboer <mtenboer@...> wrote:
Thanks Brian.  I'm hoping Ray or Howard can pipe in and answer #2
2.  What happens if the Dec-Arc box is checked in the Tracking Correction Status section (and also the other Pointing and Tracking boxes on that screen are checked) but the target will NOT travel through the mapped area ?  Does that cause worse tracking for that target, or does APCC suspend the model adjustments?

The reason I ask is that once you make a Dec-Arc map/model for a target, the model stays in place in APCC indefinitely until it is replaced with another.  For the next target that night, I might decide there isn't enough time left to make a new Dec Arc map and so I will track without it.  However, the model for target 1 is still active in APCC and the check boxes are checked.  Does APCC disregard the model in place if and when the target travels through an area outside the map?  

Mark




Ray Gralak
 

Hi Chris,

If the semantics are important then the nomenclature within the software should be amended.
A model only exists inside APCC at runtime, not in the PNT file, which is just a bunch of data points. Loading the file "loads" (creates) the model.

When I implement a "Dec Arc Tracking Algorithm" in APCC I do so in the Pointing Model tab.
In order to view APCC's "Pointing Model" tab, you have to have APCC running, so a model exists then and is semantically accurate.

BTW, there is no mention of a "Pointing Model" in APPM. The act of collecting data with APPM is just gathering data. As stated before, once a mapping file (.PNT) is loaded, then a model is created inside APCC and goes away when another set of data points is loaded or APCC is closed.

The only place where the word Model is left out is where the check box is for enabling Dec arc tracking. I'm
guessing that the way that Dec Arc Tracking is implemented within APCC utilizing the Model that is in place, is
where the confusion arises.
The Dec Arc tracking algorithm is independent of the All-Sky model. They actually coexist with each other. The All-Sky model is used for pointing correction even when Dec Arc Tracking is enabled. And, the All-Sky model has a tracking rate correction algorithm of its own. Enabling Dec Arc Tracking simply tells APCC to use that tracking rate correction algorithm instead of the All-Sky tracking rate algorithm.

-Ray


Ray Gralak
 

Hi Mark,

2. What happens if the Dec-Arc box is checked in the Tracking Correction Status section (and also the other
Pointing and Tracking boxes on that screen are checked) but the target will NOT travel through the mapped area ?
Does that cause worse tracking for that target, or does APCC suspend the model adjustments?
To maintain the highest tracking accuracy of the algorithm, the Dec Arc algorithm is only active within the area of sky captured by APPM.

APCC will fall back to the All-Sky tracking rate algorithm when the scope points outside the mapped sky area.

-Ray


Greg Salyer
 

Ray, I got to ask a dumb question. Why would you not check the Dec Arc box?

On Nov 4, 2022, at 9:17 AM, Ray Gralak <iogroups@...> wrote:

Hi Mark,

2. What happens if the Dec-Arc box is checked in the Tracking Correction Status section (and also the other
Pointing and Tracking boxes on that screen are checked) but the target will NOT travel through the mapped area ?
Does that cause worse tracking for that target, or does APCC suspend the model adjustments?
To maintain the highest tracking accuracy of the algorithm, the Dec Arc algorithm is only active within the area of sky captured by APPM.

APCC will fall back to the All-Sky tracking rate algorithm when the scope points outside the mapped sky area.

-Ray







Ray Gralak
 

Ray, I got to ask a dumb question. Why would you not check the Dec Arc box?
To compare against the All-Sky model or if you pass a custom set of points or random points which may not be the best data for the algorithm.

-Ray


Mark Tenboer
 

On Fri, Nov 4, 2022 at 06:17 AM, Ray Gralak wrote:
APCC will fall back to the All-Sky tracking rate algorithm when the scope points outside the mapped sky area.
Ray - it seems the term "all sky" is used for two different purposes. 
1.   The user creates a point map of the entire sky (the sky within the horizon limits) and APCC uses that map to create an "all sky model".  This model is replaced when the user creates a new map of any kind (all sky, or Dec-Arc)   
2.   The name of a permanent tracking algorithm in APCC - the "All Sky tracking rate algorithm" -  that is used by default in APCC when ANY point map is loaded (regardless of whether is covers the entire sky)  AND either the "Enable Dec-Arc Tracking" box is not checked, or the scope is pointing outside the Dec-Arc map. 

Is this correct ?  

And finally - is there any reason NOT to keep the "Correct for Refraction" box checked for all models ?    I realize that refraction changes on elevation, and the benefit might be immaterial at times, but shouldn't we just keep it checked ?    


Ray Gralak
 

Hi Mark,

APCC will fall back to the All-Sky tracking rate algorithm when the scope points outside the mapped sky
area.

Ray - it seems the term "all sky" is used for two different purposes.
1. The user creates a point map of the entire sky (the sky within the horizon limits) and APCC uses that map to
create an "all sky model". This model is replaced when the user creates a new map of any kind (all sky, or Dec-
Arc)
2. The name of a permanent tracking algorithm in APCC - the "All Sky tracking rate algorithm" - that is used by
default in APCC when ANY point map is loaded (regardless of whether is covers the entire sky) AND either the
"Enable Dec-Arc Tracking" box is not checked, or the scope is pointing outside the Dec-Arc map.

Is this correct ?
Yes, that is correct, except the part about a "Dec-Arc Map", because it seemingly associates the points with the Dec Arc tracking algorithm.

I think a better name might be a "sky-area map" or "quick map".

And finally - is there any reason NOT to keep the "Correct for Refraction" box checked for all models ? I realize
that refraction changes on elevation, and the benefit might be immaterial at times, but shouldn't we just keep it
checked ?
Yes, if you are using the mount's King rate, or you are tracking a comet or object with supplied coordinates that already embed a refraction calculation.

-Ray


Nathan Myhrvold
 

OK, just so I keep this straight, here is my understanding, Ray, can you tell me if it is correct?

Note that the OTA for my AP1100 AE is not here yet, so I am presently do not carry the burden by any actual experience 😊. Sorry about that!

1. You can sample the entire sky with data points (i.e. slew to a point in the sky, take a photo, then platesolve) which the mount will turn into a model used for pointing the telescope, or "all sky" tracking.
(a). How many points is a reasonable number? Like integration time, I suspect more is better but there must be some rule of thumb.
(b). I presume that you should cover the portion of the sky that care about - i.e. if you don't intend to point the telescope below say 40 degrees altitude, then you don't need to sample there, right?
(c). Or, perhaps having data points all over the sky more or less symmetrically helps refine the curve fits?

2. You can also sample points along a constant declination arc in anticipation of photographing an object at that declination for a period of time.
(a). How many points is a good number for sampling?
(b). Presumably, this depends on the length of the arc (in RA degrees or hour angle range), but also on the declination. An arc near the celestial equator would need more points than an arc near the celestial pole.
(c). I presume that the samples should fully cover the anticipated range of the imaging.
(d). Dec Arc tracking is a specific and different tracking algorithm, which you must select with a check box in APCC.
(e). I think what you are saying below is that Dec Arc tracking is not specifically tied to the extra samples along a constant declination per se, is that correct?

Basically, in the post below, and some other email replies, you seem talk in terms of Dec-Arc Tracking as an algorithm which is enabled or assisted by data along the arc, but it is not directly tied to that data. But maybe I am not understanding it.

3. I have experience with unguided pointing and tracking models with 10micron mounts. There appear to be several big differences in how "model building" is done with APCC versus 10micron.
(a). For 10micron mounts, third party software (like ModelWizard, ModelCreator) is used to gather the datapoints, actual modeling is done inside the mount.
(b). ModelWizard and similar don't actually make the model. The actual modeling is done with software in the 10micron mount - what the software does is help automate the process and provide a user interface.
(c). APCC performs both data point gathering, and modeling, for AP mounts. One can also use Dale Ghent's NINA plug ins to further automate gathering data points, such as collecting Dec-Arc samples.
(d). For 10micron, the usual recommendation is to build a small (12 -15 data point) model first, then make a larger 50-100 point model for the whole sky. Nobody in the AP thread talks about that, so it does not seem to be standard practice.
(e). The 10micron mount will return error information about the data points - basically whether the point is well fit by the model, or is an outlier for some reason. This is essentially like labeling outliers in statistical curve fitting.
(f). So, most users go through an iterative process where you can delete the worst data points and then re-take them. Nobody in the AP thread talks about this, so I am presuming that this isn't normally done, or an issue.
(g). Overall modeling error also comes from the 10micron mount. This is used to diagnose whether there is something the model can't handle - such as flexure or a loose screw someplace on the OTA.
(h). So, 10micron owners obsess over the modeling error, and brag what their RMS error is, but it is modeling error, not actual tracking performance. Again, nobody on the AP thread seems to talk about that.

Nathan

-----Original Message-----
From: main@ap-gto.groups.io <main@ap-gto.groups.io> On Behalf Of Ray Gralak via groups.io
Sent: Friday, November 4, 2022 11:29 AM
To: main@ap-gto.groups.io
Subject: [EXTERNAL] Re: [ap-gto] Just what is a Dec Arc model?

Hi Mark,

APCC will fall back to the All-Sky tracking rate algorithm when the
scope points outside the mapped sky area.

Ray - it seems the term "all sky" is used for two different purposes.
1. The user creates a point map of the entire sky (the sky within the horizon limits) and APCC uses that map to
create an "all sky model". This model is replaced when the user
creates a new map of any kind (all sky, or Dec-
Arc)
2. The name of a permanent tracking algorithm in APCC - the "All Sky tracking rate algorithm" - that is used by
default in APCC when ANY point map is loaded (regardless of whether is
covers the entire sky) AND either the "Enable Dec-Arc Tracking" box is not checked, or the scope is pointing outside the Dec-Arc map.

Is this correct ?
Yes, that is correct, except the part about a "Dec-Arc Map", because it seemingly associates the points with the Dec Arc tracking algorithm.

I think a better name might be a "sky-area map" or "quick map".

And finally - is there any reason NOT to keep the "Correct for Refraction" box checked for all models ? I realize
that refraction changes on elevation, and the benefit might be
immaterial at times, but shouldn't we just keep it checked ?
Yes, if you are using the mount's King rate, or you are tracking a comet or object with supplied coordinates that already embed a refraction calculation.

-Ray


Chris White
 

On Thu, Nov 3, 2022 at 10:06 PM, Ray Gralak wrote:
Hi Chris,

If the semantics are important then the nomenclature within the software should be amended.
A model only exists inside APCC at runtime, not in the PNT file, which is just a bunch of data points. Loading the file "loads" (creates) the model.

When I implement a "Dec Arc Tracking Algorithm" in APCC I do so in the Pointing Model tab.
In order to view APCC's "Pointing Model" tab, you have to have APCC running, so a model exists then and is semantically accurate.

BTW, there is no mention of a "Pointing Model" in APPM. The act of collecting data with APPM is just gathering data. As stated before, once a mapping file (.PNT) is loaded, then a model is created inside APCC and goes away when another set of data points is loaded or APCC is closed.

The only place where the word Model is left out is where the check box is for enabling Dec arc tracking. I'm
guessing that the way that Dec Arc Tracking is implemented within APCC utilizing the Model that is in place, is
where the confusion arises.
The Dec Arc tracking algorithm is independent of the All-Sky model. They actually coexist with each other. The All-Sky model is used for pointing correction even when Dec Arc Tracking is enabled. And, the All-Sky model has a tracking rate correction algorithm of its own. Enabling Dec Arc Tracking simply tells APCC to use that tracking rate correction algorithm instead of the All-Sky tracking rate algorithm.

-Ray
Ray,

You missed my point.  I'm not arguing about the semantics, I'm trying to communicate that there is a TON OF CONFUSION surrounding the topic of Dec Arc "whatever".... 

Maybe I should have just said that perhaps a future release of APCC could utilize a UI update that made this less confusing. 

I am a regular user of APCC modeling and dec arc tracking.  I still don't know if it would be smarter for me to burn an entire night and create a 1000 point model so that all declinations are just covered for the purpose of using dec arc tracking anywhere in my sky.  I like to image multiple targets in a single night and I need to load new models for each target when my script changes.  I'd happily burn a night if you said a 1000 point model would last me for an entire year without needing updates, across multiple scopes.  But as it is, there is very little official information about this entire subject other than how to create and enable. 

Thank you,


Ray Gralak
 

Hi Chris,

You missed my point. I'm not arguing about the semantics, I'm trying to communicate that there is a TON OF
CONFUSION surrounding the topic of Dec Arc "whatever"....
Others have run away with using "Dec Arc xxx" phrases, but that was well after APCC and documentation was released with Dec Arc Tracking. So although I don't believe the confusion you mention is our fault, we are trying to fix this in the next release of APCC.

I am a regular user of APCC modeling and dec arc tracking. I still don't know if it would be smarter for me to burn
an entire night and create a 1000 point model so that all declinations are just covered for the purpose of using dec
arc tracking anywhere in my sky. I like to image multiple targets in a single night and I need to load new models
for each target when my script changes. I'd happily burn a night if you said a 1000 point model would last me for
an entire year without needing updates, across multiple scopes.
If you change scopes, you will likely need to do a mapping run again. If the temperature changes dramatically, you may need another mapping run. This has nothing to do with refraction, but with the changes in stiffness of the scope. Sometimes this has little effect. In other setups it may be measurable. Every setup is different so you may need to experiment to determine under what conditions you will need to do another mapping run.

-Ray