|
Re: Results I get from my Mach 2
Chris White
I was imaging last night, 300 second unguided images with Mach 2. I collected 86 subs and every single one had an eccentricity of below 0.4 and many below 0.35.
I was doing a mosaic so had a pretty large chunk of sky. I created a 110 point model with 1 degree spacing in Dec and 7 degree spacing in RA.
Performance from the mount was mind blowing. Not one single sub had one single issue. PERFECTION!
|
|
|
Re: Driver Error with J2000/JNow conversion
Chris I think the logs would be helpful to make available to Ray et al
Last night I was imaging a mosaic with Voyager and there was a strange issue with the capture. One of the Panes did not shift to the new sky coordinate and I ended up getting double data on one pane and nothing on the other. (Because of the glitch Voyager did not get an updated coordinate for the new pane and reverted to the previous value, i.e previous pane). I reached out to Leo at Voyager and sent his logs. He found that there was a glitch in the Mount Driver where there was a failed conversion from J2000 to JNow.
This error could easily happen when imaging a single target as you would never know, but doing the Mosaic it obviously created an issue.
Thought I would bring it up to see if Ray and the other Software wizards here could provide some insight. I can provide logs and I know what time (within a few minutes) the issue occurred.
Thanks in advance.
|
|
|
Driver Error with J2000/JNow conversion
Chris White
Last night I was imaging a mosaic with Voyager and there was a strange issue with the capture. One of the Panes did not shift to the new sky coordinate and I ended up getting double data on one pane and nothing on the other. (Because of the glitch Voyager did not get an updated coordinate for the new pane and reverted to the previous value, i.e previous pane). I reached out to Leo at Voyager and sent his logs. He found that there was a glitch in the Mount Driver where there was a failed conversion from J2000 to JNow.
This error could easily happen when imaging a single target as you would never know, but doing the Mosaic it obviously created an issue.
Thought I would bring it up to see if Ray and the other Software wizards here could provide some insight. I can provide logs and I know what time (within a few minutes) the issue occurred.
Thanks in advance.
|
|
|
Re: Dec-Arc model vs. All Sky
Karen Christen
Hey Astro-Nuts,
A side note on group etiquette: I’m going to ask that we don’t bash other companies’ products or other astronomers’ choices. If someone asks for advice,
of course it makes sense to guide them to what you feel is the best configuration or process. But we can be supportive and helpful without denigrating others’ preferences. Astro-Physics prides itself on our kind, knowledgeable customer interactions. We
know many of you have plenty of knowledge to share, as well. Please do so in the same manner.
Karen
AP
toggle quoted message
Show quoted text
From: main@ap-gto.groups.io <main@ap-gto.groups.io>
On Behalf Of Arvind
Sent: Wednesday, October 12, 2022 3:20 AM
To: main@ap-gto.groups.io
Subject: Re: [ap-gto] Dec-Arc model vs. All Sky
if not for ASIAIR I wouldn't be in this hobby, so yes to that.
Mind your language, "bro"
On Wed, Oct 12, 2022 at 1:16 AM Bill Long <bill@...> wrote:
About to share some results friend. ASI Air, best for AP? LOL no bro. GTFO with that.
ELI5 docs anywhere? :-) Dec-Arc, Dec-Arc-tracking, all-sky. Ofcourse one can do all everytime.. but what's a good rule of thumb for someone (like me) that needs to physically setup up and tear down each night?
Or should I continue reading the manual.. (I've been reading it already while I wait for the 1100ae to arrive).
On Tue, Oct 11, 2022 at 9:24 PM Bill Long <bill@...> wrote:
I think there is something really important to call out about the Dec Arc models, and that is that you CAN make very effective models with very few points. Tonight I am
imaging with a tiny 32 point model, that took 10 minutes to make, and I am seeing 2.1" FWHM and 0.35 eccentricity with 10 minute HA frames here in the PNW.
A lot of people can get caught up in needing hundreds of points to feel like they can lay the Linus Blanket of guiding down, and that just is not the case.
Roland was imaging with a couple of points (less than a handful) and got great performance.
>>> Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
That's true, and there's a separate Dec Arc Model i.e., a limited model around a specific declination with an eye towards imaging one specific target
It's easy to get the dec arc model and the dec arc tracking algorithm mixed up, it seems to be used interchangeably in some threads.
On Tue, Oct 11, 2022 at 7:41 PM Ray Gralak <iogroups@...> wrote:
> >>>What is surprising in this thread is folks complaining about scope pointing to the ground after a 30pt model.
> How does one manage to end up in such a state?
>
> just to be clear, i think we're talking about a dec arc pointing model, not an allsky model. So terms that are
> calculated are unique to a very narrow range in declination. I could see that causing issues for pointing outside of
> the model, particularly around the poles
Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
However, for pointing correction, APCC will fall back to an all-sky pointing model, if there is one. The problem with the "mini" Dec-Arc model is that all of the data points line up at about the same declination which when coupled with too few points could
sometimes produce a ridiculously incorrect pointing model on the West side (pier-side=East). Pointing was usually good enough on the East side because of the starting Zenith point, which usually "stabilizes" the solution.
APCC now does two things to help prevent bad pointing:
1) it throws away any ridiculous pointing terms.
2) APPM automatically inserts a zenith point in the West hemisphere to mirror the starting Zenith East point.
Even given those two remedies, it might be better to be safe and turn off pointing correction (but leave on tracking rate correction of course).
-Ray
--
-- Karen ChristenAstro-Physics
|
|
|
Re: Dec-Arc model vs. All Sky
Dean Jacobsen
|
|
|
Re: Dec-Arc model vs. All Sky
Hi Ray
Yes, I understand completely. it can just get confusing with some of the overlapping terms that are commonly used
On Wed, Oct 12, 2022 at 6:37 AM Ray Gralak < iogroups@...> wrote: Hi Brian,
> That's true, and there's a separate Dec Arc Model
> i.e., a limited model around a specific declination with an eye towards imaging one specific target
Okay, I see what you meant, but to be exact, models only exist inside APCC . People may call what you describe above a "Dec Arc Model", but the data collected by APPM is just a collection of sky points. From APCC's perspective, there are only pointing-term models (aka "all-sky" modeling) and Dec Arc Tracking rate models. Both can be created from the same data points. They are just different algorithms.
However, as you know, APCC can use a pointing- term model for pointing correction independently of its Dec Arc Tracking model. This can produce a very accurate all-sky combination of pointing and tracking rate correction when points are collected throughout the sky (i.e. a "true" all-sky model instead of what people call a "Dec Arc model"). Of course, tracking rate correction will be excellent in what people call a "Dec Arc model".
-Ray
|
|
|
Re: Unguided Tracking on a very still night: Mach2 with 110 GTX
"The mount is being tracked with a 4 point drift model that I did with the keypad."
Does that mean you have dual axis tracking turned off? Therefore you do not have a polynomial for the sky model, and therefore no Dec adjustments due to polar alignment error? So you are tracking sidereal in RA only, since your Polar alignment is that good?
|
|
|
Re: Dec-Arc model vs. All Sky
Hi Brian, That's true, and there's a separate Dec Arc Model
i.e., a limited model around a specific declination with an eye towards imaging one specific target Okay, I see what you meant, but to be exact, models only exist inside APCC . People may call what you describe above a "Dec Arc Model", but the data collected by APPM is just a collection of sky points. From APCC's perspective, there are only pointing-term models (aka "all-sky" modeling) and Dec Arc Tracking rate models. Both can be created from the same data points. They are just different algorithms. However, as you know, APCC can use a pointing- term model for pointing correction independently of its Dec Arc Tracking model. This can produce a very accurate all-sky combination of pointing and tracking rate correction when points are collected throughout the sky (i.e. a "true" all-sky model instead of what people call a "Dec Arc model"). Of course, tracking rate correction will be excellent in what people call a "Dec Arc model". -Ray
|
|
|
Re: Dec-Arc model vs. All Sky
Thank you Brian and Ray - that helped! Here’s a screen shot of the Enable Dec Arc Tracking box with the button checked so others can see what I needed to do. I’m observatory based, so I create a 300 point model every couple months. Thanks, Bob
Robert J. Enouen Cell 513-504-4410
On Oct 8, 2022, at 3:44 PM, Ray Gralak <iogroups@...> wrote:
Hi Bob,I just want to confirm what Brian stated. If you have a permanent setup you can create a dense all-sky model with hundreds of points and use DecArc-Tracking throughout the entire area of the sky that you mapped. In fact, the intended operation for Dec-Arc Tracking was to provide a better tracking model than the traditional all-sky models used by APCC, TPoint, 10_Micron, etc. However, if you are mobile and short on time you can use Dec-Arc tracking for a small swath of the sky and automate the collection process via NINA. -Ray-----Original Message-----
Thanks Brian!
Bob
Robert J. Enouen
Cell 513-504-4410
On Oct 8, 2022, at 1:21 PM, Brian Valente <bvalente@...> wrote:
Hi Bob
>>> Would a Dec-Arc model be an improvement over my 300 point APPM model?
you do a dec arc model with significantly more points, it may yield some improvement. There's a lot of
variables in there that doing a test may be the only way to know for sure about your specific setup. 300 points is a
lot of points, so I suspect any improvements may be limited.
And just in case it isn't clear, you can enable dec arc tracking with your existing model (sometimes people
use dec arc interchangeably to mean tracking and/or the modeling along a specific Declination range)
image.png
On Sat, Oct 8, 2022 at 10:11 AM Bob Enouen <renouen@...> wrote:
Thanks Marj!
I have a follow up for Ray - Nathan suggested the following re my DR350 focus. Would a Dec-Arc
model be an improvement over my 300 point APPM model?
‘You might want to try the Dec-Arc model for tracking. I would think that you could get smaller
FWHM on the stars. You can still make an overall model for pointing. I am NOT an expert on this having never
done it for real but that is what I have learned on this forum.”
Thanks, Bob
Robert J. Enouen
Cell 513-504-4410
On Oct 8, 2022, at 12:54 PM, Marj Christen <marj@...> wrote:
Hello everyone. Brian is finishing up the new YouTube video and is incorporating the images
that were posted here and also those he received privately. He won’t be able to add more so any new ones won’t
be incorporated. Just wanted you to know so you won’t be disappointed. However, please feel free to post your
photos any time.
The outpouring of photos has been fabulous and we have enjoyed seeing your equipment
systems and astro-images. Thank you so much for sharing them with all of us.
I have noticed that this thread had taken a bit of a turn. Perhaps a new thread with an
appropriate title should be started. Please remember that this group is centered on sharing experiences and
information about AP GTO mounts.
Clear Skies,
Marj Christen
Astro-Physics
11250 Forest Hills Road
Machesney Park, IL 61115
Phone: 815-282-1513
www.astro-physics.com
Video Tutorials: https://www.youtube.com/astrophysicscorp
From: main@ap-gto.groups.io <main@ap-gto.groups.io> On Behalf Of
ap@...
Sent: Saturday, October 8, 2022 11:38 AM
To: main@ap-gto.groups.io
Subject: Re: [ap-gto] We want your Astro and AP Equipment Images
On Sat, Oct 8, 2022 at 12:30 PM, Howard Ritter wrote:
Yes, but the air in the scope doesn’t contribute to its weight.
So a scope filled with helium would not be lighter than one filled with air? :)
A-ha... a new way to create more portable scopes when every ounce counts! Though I
wonder if the refractive index calculations change in the glass -> He boundary.
This reminds me of those physics problems, whether a bird flying free inside a rail car
contributes to its weight on the tracks. And does your answer change if the windows are open? You just never
know what these discussions will bring, now I have something new to puzzle over as the weather turns to crap
this week.
Linwood
--
Brian
Brian Valente
astro portfolio https://www.brianvalentephotography.com/astrophotography/
portfolio brianvalentephotography.com <http://brianvalentephotography.com>
astrobin https://www.astrobin.com/users/bvalente/
|
|
|
Re: Results I get from my Mach 2
Bill,
Thank you so much for sharing the specific tools/data to help me improve! I’d appreciate your (and anyone’s) advice based on facts below.
I used FWMHEccentricity on 3 Ha test images taken in succession one night ago during a full moon from my Observatory-based AP1600 with no AE, unguided, Delta Rho 350, ASI 6200MM Pro with a 300 point APPM model with Dec Arc Tracking enabled, using NINA. The pixel scale = .74”/px at 1x1 binning. I ran autofocus right before these images were taken.
Here are the results:
Time (sec) Median FWHM Median Eccentricity 180 sec 5.310 px .3622 360 sec 5.516 px .3585 600 sec 5.705 px .3404
I was amazed that the center of each star is at the exact same camera sensor coordinates! The best FWMH I’ve obtained with this system has been 3.3 on a red Chroma filter at 180 sec with no moon, so when I see your and Roland’s results it’s pretty clear that my system needs work.
My conclusions are: 1. I have a focus issue. 2. The AP1600 is doing a spectacular job of pointing and tracking. 3. Guiding/AE may help a bit, but with a fairly tight range of 5.3-5.7 FWMH from 3 minutes to 10 minutes that’s not my key issue.
Thoughts? Thanks, Bob
Robert J. Enouen Cell 513-504-4410
On Oct 12, 2022, at 4:32 AM, Bill Long <bill@...> wrote:
Zoom into the grade:
FWHM in pixels, scale is 0.89"/px. 1.689 * 0.89 = 1.50321
Get out! Thank you Roland!! 😄
From: main@ap-gto.groups.io <main@ap-gto.groups.io> on behalf of Bill Long <bill@...>
Sent: Wednesday, October 12, 2022 1:24 AM
To: main@ap-gto.groups.io <main@ap-gto.groups.io>
Subject: [ap-gto] Results I get from my Mach 2
This:
FWHM is in pixels here. The scale is 0.89"/px. That is a 1.5" FWHM image, with 0.33 eccentricity. The sharpness of those stars in this image, and the roundness of them is what some people here just dont get.
Show me the same, from something using ASI Air. Anyone?
|
|
|
Re: Dec-Arc model vs. All Sky
The discussion suggested me a question: How much important is to provide apcc precise temperature and humidity? It is important to collect data on field when we run the model or the weather website data is enough? Thank you Andrea
Il giorno mer 12 ott 2022 alle 04:41 Ray Gralak < iogroups@...> ha scritto: > >>>What is surprising in this thread is folks complaining about scope pointing to the ground after a 30pt model.
> How does one manage to end up in such a state?
>
> just to be clear, i think we're talking about a dec arc pointing model, not an allsky model. So terms that are
> calculated are unique to a very narrow range in declination. I could see that causing issues for pointing outside of
> the model, particularly around the poles
Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
However, for pointing correction, APCC will fall back to an all-sky pointing model, if there is one. The problem with the "mini" Dec-Arc model is that all of the data points line up at about the same declination which when coupled with too few points could sometimes produce a ridiculously incorrect pointing model on the West side (pier-side=East). Pointing was usually good enough on the East side because of the starting Zenith point, which usually "stabilizes" the solution.
APCC now does two things to help prevent bad pointing:
1) it throws away any ridiculous pointing terms.
2) APPM automatically inserts a zenith point in the West hemisphere to mirror the starting Zenith East point.
Even given those two remedies, it might be better to be safe and turn off pointing correction (but leave on tracking rate correction of course).
-Ray
|
|
|
Re: Results I get from my Mach 2
Zoom into the grade:
FWHM in pixels, scale is 0.89"/px. 1.689 * 0.89 = 1.50321
Get out! Thank you Roland!! 😄
toggle quoted message
Show quoted text
From: main@ap-gto.groups.io <main@ap-gto.groups.io> on behalf of Bill Long <bill@...>
Sent: Wednesday, October 12, 2022 1:24 AM
To: main@ap-gto.groups.io <main@ap-gto.groups.io>
Subject: [ap-gto] Results I get from my Mach 2
This:
FWHM is in pixels here. The scale is 0.89"/px. That is a 1.5" FWHM image, with 0.33 eccentricity. The sharpness of those stars in this image, and the roundness of them is what some people here just dont get.
Show me the same, from something using ASI Air. Anyone?
|
|
|
Results I get from my Mach 2
This:
FWHM is in pixels here. The scale is 0.89"/px. That is a 1.5" FWHM image, with 0.33 eccentricity. The sharpness of those stars in this image, and the roundness of them is what some people here just dont get.
Show me the same, from something using ASI Air. Anyone?
|
|
|
Re: Dec-Arc model vs. All Sky
if not for ASIAIR I wouldn't be in this hobby, so yes to that.
Mind your language, "bro"
On Wed, Oct 12, 2022 at 1:16 AM Bill Long < bill@...> wrote:
About to share some results friend. ASI Air, best for AP? LOL no bro. GTFO with that.
ELI5 docs anywhere? :-) Dec-Arc, Dec-Arc-tracking, all-sky. Ofcourse one can do all everytime.. but what's a good rule of thumb for someone (like me) that needs to physically setup up and tear down each night?
Or should I continue reading the manual.. (I've been reading it already while I wait for the 1100ae to arrive).
On Tue, Oct 11, 2022 at 9:24 PM Bill Long < bill@...> wrote:
I think there is something really important to call out about the Dec Arc models, and that is that you CAN make very effective models with very few points. Tonight I am imaging with a tiny 32 point model, that took 10 minutes to make, and I am seeing 2.1" FWHM
and 0.35 eccentricity with 10 minute HA frames here in the PNW.
A lot of people can get caught up in needing hundreds of points to feel like they can lay the Linus Blanket of guiding down, and that just is not the case.
Roland was imaging with a couple of points (less than a handful) and got great performance.
>>> Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
That's true, and there's a separate Dec Arc Model i.e., a limited model around a specific declination with an eye towards imaging one specific target
It's easy to get the dec arc model and the dec arc tracking algorithm mixed up, it seems to be used interchangeably in some threads.
On Tue, Oct 11, 2022 at 7:41 PM Ray Gralak < iogroups@...> wrote:
> >>>What is surprising in this thread is folks complaining about scope pointing to the ground after a 30pt model.
> How does one manage to end up in such a state?
>
> just to be clear, i think we're talking about a dec arc pointing model, not an allsky model. So terms that are
> calculated are unique to a very narrow range in declination. I could see that causing issues for pointing outside of
> the model, particularly around the poles
Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
However, for pointing correction, APCC will fall back to an all-sky pointing model, if there is one. The problem with the "mini" Dec-Arc model is that all of the data points line up at about the same declination which when coupled with too few points could
sometimes produce a ridiculously incorrect pointing model on the West side (pier-side=East). Pointing was usually good enough on the East side because of the starting Zenith point, which usually "stabilizes" the solution.
APCC now does two things to help prevent bad pointing:
1) it throws away any ridiculous pointing terms.
2) APPM automatically inserts a zenith point in the West hemisphere to mirror the starting Zenith East point.
Even given those two remedies, it might be better to be safe and turn off pointing correction (but leave on tracking rate correction of course).
-Ray
--
|
|
|
Re: Dec-Arc model vs. All Sky
About to share some results friend. ASI Air, best for AP? LOL no bro. GTFO with that.
toggle quoted message
Show quoted text
From: main@ap-gto.groups.io <main@ap-gto.groups.io> on behalf of Arvind <base16@...>
Sent: Tuesday, October 11, 2022 10:27 PM
To: main@ap-gto.groups.io <main@ap-gto.groups.io>
Subject: Re: [ap-gto] Dec-Arc model vs. All Sky
ELI5 docs anywhere? :-) Dec-Arc, Dec-Arc-tracking, all-sky. Ofcourse one can do all everytime.. but what's a good rule of thumb for someone (like me) that needs to physically setup up and tear down each night?
Or should I continue reading the manual.. (I've been reading it already while I wait for the 1100ae to arrive).
On Tue, Oct 11, 2022 at 9:24 PM Bill Long < bill@...> wrote:
I think there is something really important to call out about the Dec Arc models, and that is that you CAN make very effective models with very few points. Tonight I am imaging with a tiny 32 point model, that took 10 minutes to make, and I am seeing 2.1" FWHM
and 0.35 eccentricity with 10 minute HA frames here in the PNW.
A lot of people can get caught up in needing hundreds of points to feel like they can lay the Linus Blanket of guiding down, and that just is not the case.
Roland was imaging with a couple of points (less than a handful) and got great performance.
>>> Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
That's true, and there's a separate Dec Arc Model i.e., a limited model around a specific declination with an eye towards imaging one specific target
It's easy to get the dec arc model and the dec arc tracking algorithm mixed up, it seems to be used interchangeably in some threads.
On Tue, Oct 11, 2022 at 7:41 PM Ray Gralak < iogroups@...> wrote:
> >>>What is surprising in this thread is folks complaining about scope pointing to the ground after a 30pt model.
> How does one manage to end up in such a state?
>
> just to be clear, i think we're talking about a dec arc pointing model, not an allsky model. So terms that are
> calculated are unique to a very narrow range in declination. I could see that causing issues for pointing outside of
> the model, particularly around the poles
Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
However, for pointing correction, APCC will fall back to an all-sky pointing model, if there is one. The problem with the "mini" Dec-Arc model is that all of the data points line up at about the same declination which when coupled with too few points could
sometimes produce a ridiculously incorrect pointing model on the West side (pier-side=East). Pointing was usually good enough on the East side because of the starting Zenith point, which usually "stabilizes" the solution.
APCC now does two things to help prevent bad pointing:
1) it throws away any ridiculous pointing terms.
2) APPM automatically inserts a zenith point in the West hemisphere to mirror the starting Zenith East point.
Even given those two remedies, it might be better to be safe and turn off pointing correction (but leave on tracking rate correction of course).
-Ray
--
|
|
|
Re: Dec-Arc model vs. All Sky
Arvind I think it depends on some of your preferences and what kind of tolerances/precision you desire
If you like the notion of using sky modeling for improved tracking (with or without guiding, with or without encoders), you probably want to: - Create a dec arc model built around the declination of your target for the evening (you can repeat this for multiple targets through the night and even automate it in NINA using the plugin)
- enable tracking corrections
- enable dec arc tracking
If you prefer not to use sky modeling (for example, several users here report they are happy just guiding, or using encoders and unguided) then you don't need to do any of the above
I hope that helps
Brian
On Tue, Oct 11, 2022 at 10:27 PM Arvind < base16@...> wrote: ELI5 docs anywhere? :-) Dec-Arc, Dec-Arc-tracking, all-sky. Ofcourse one can do all everytime.. but what's a good rule of thumb for someone (like me) that needs to physically setup up and tear down each night?
Or should I continue reading the manual.. (I've been reading it already while I wait for the 1100ae to arrive).
On Tue, Oct 11, 2022 at 9:24 PM Bill Long < bill@...> wrote:
I think there is something really important to call out about the Dec Arc models, and that is that you CAN make very effective models with very few points. Tonight I am imaging with a tiny 32 point model, that took 10 minutes to make, and I am seeing 2.1" FWHM
and 0.35 eccentricity with 10 minute HA frames here in the PNW.
A lot of people can get caught up in needing hundreds of points to feel like they can lay the Linus Blanket of guiding down, and that just is not the case.
Roland was imaging with a couple of points (less than a handful) and got great performance.
>>> Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
That's true, and there's a separate Dec Arc Model i.e., a limited model around a specific declination with an eye towards imaging one specific target
It's easy to get the dec arc model and the dec arc tracking algorithm mixed up, it seems to be used interchangeably in some threads.
On Tue, Oct 11, 2022 at 7:41 PM Ray Gralak < iogroups@...> wrote:
> >>>What is surprising in this thread is folks complaining about scope pointing to the ground after a 30pt model.
> How does one manage to end up in such a state?
>
> just to be clear, i think we're talking about a dec arc pointing model, not an allsky model. So terms that are
> calculated are unique to a very narrow range in declination. I could see that causing issues for pointing outside of
> the model, particularly around the poles
Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
However, for pointing correction, APCC will fall back to an all-sky pointing model, if there is one. The problem with the "mini" Dec-Arc model is that all of the data points line up at about the same declination which when coupled with too few points could
sometimes produce a ridiculously incorrect pointing model on the West side (pier-side=East). Pointing was usually good enough on the East side because of the starting Zenith point, which usually "stabilizes" the solution.
APCC now does two things to help prevent bad pointing:
1) it throws away any ridiculous pointing terms.
2) APPM automatically inserts a zenith point in the West hemisphere to mirror the starting Zenith East point.
Even given those two remedies, it might be better to be safe and turn off pointing correction (but leave on tracking rate correction of course).
-Ray
--
|
|
|
Re: Dec-Arc model vs. All Sky
ELI5 docs anywhere? :-) Dec-Arc, Dec-Arc-tracking, all-sky. Ofcourse one can do all everytime.. but what's a good rule of thumb for someone (like me) that needs to physically setup up and tear down each night?
Or should I continue reading the manual.. (I've been reading it already while I wait for the 1100ae to arrive).
On Tue, Oct 11, 2022 at 9:24 PM Bill Long < bill@...> wrote:
I think there is something really important to call out about the Dec Arc models, and that is that you CAN make very effective models with very few points. Tonight I am imaging with a tiny 32 point model, that took 10 minutes to make, and I am seeing 2.1" FWHM
and 0.35 eccentricity with 10 minute HA frames here in the PNW.
A lot of people can get caught up in needing hundreds of points to feel like they can lay the Linus Blanket of guiding down, and that just is not the case.
Roland was imaging with a couple of points (less than a handful) and got great performance.
>>> Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
That's true, and there's a separate Dec Arc Model i.e., a limited model around a specific declination with an eye towards imaging one specific target
It's easy to get the dec arc model and the dec arc tracking algorithm mixed up, it seems to be used interchangeably in some threads.
On Tue, Oct 11, 2022 at 7:41 PM Ray Gralak < iogroups@...> wrote:
> >>>What is surprising in this thread is folks complaining about scope pointing to the ground after a 30pt model.
> How does one manage to end up in such a state?
>
> just to be clear, i think we're talking about a dec arc pointing model, not an allsky model. So terms that are
> calculated are unique to a very narrow range in declination. I could see that causing issues for pointing outside of
> the model, particularly around the poles
Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
However, for pointing correction, APCC will fall back to an all-sky pointing model, if there is one. The problem with the "mini" Dec-Arc model is that all of the data points line up at about the same declination which when coupled with too few points could
sometimes produce a ridiculously incorrect pointing model on the West side (pier-side=East). Pointing was usually good enough on the East side because of the starting Zenith point, which usually "stabilizes" the solution.
APCC now does two things to help prevent bad pointing:
1) it throws away any ridiculous pointing terms.
2) APPM automatically inserts a zenith point in the West hemisphere to mirror the starting Zenith East point.
Even given those two remedies, it might be better to be safe and turn off pointing correction (but leave on tracking rate correction of course).
-Ray
--
|
|
|
Whatever your seeing will offer you, the Mach 2 will deliver. No more, no less. Emphasis on no less.
|
|
|
Re: Dec-Arc model vs. All Sky
I think there is something really important to call out about the Dec Arc models, and that is that you CAN make very effective models with very few points. Tonight I am imaging with a tiny 32 point model, that took 10 minutes to make, and I am seeing 2.1" FWHM
and 0.35 eccentricity with 10 minute HA frames here in the PNW.
A lot of people can get caught up in needing hundreds of points to feel like they can lay the Linus Blanket of guiding down, and that just is not the case.
Roland was imaging with a couple of points (less than a handful) and got great performance.
toggle quoted message
Show quoted text
From: main@ap-gto.groups.io <main@ap-gto.groups.io> on behalf of Brian Valente <bvalente@...>
Sent: Tuesday, October 11, 2022 9:15 PM
To: main@ap-gto.groups.io <main@ap-gto.groups.io>
Subject: Re: [ap-gto] Dec-Arc model vs. All Sky
>>> Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
That's true, and there's a separate Dec Arc Model i.e., a limited model around a specific declination with an eye towards imaging one specific target
It's easy to get the dec arc model and the dec arc tracking algorithm mixed up, it seems to be used interchangeably in some threads.
On Tue, Oct 11, 2022 at 7:41 PM Ray Gralak < iogroups@...> wrote:
> >>>What is surprising in this thread is folks complaining about scope pointing to the ground after a 30pt model.
> How does one manage to end up in such a state?
>
> just to be clear, i think we're talking about a dec arc pointing model, not an allsky model. So terms that are
> calculated are unique to a very narrow range in declination. I could see that causing issues for pointing outside of
> the model, particularly around the poles
Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
However, for pointing correction, APCC will fall back to an all-sky pointing model, if there is one. The problem with the "mini" Dec-Arc model is that all of the data points line up at about the same declination which when coupled with too few points could
sometimes produce a ridiculously incorrect pointing model on the West side (pier-side=East). Pointing was usually good enough on the East side because of the starting Zenith point, which usually "stabilizes" the solution.
APCC now does two things to help prevent bad pointing:
1) it throws away any ridiculous pointing terms.
2) APPM automatically inserts a zenith point in the West hemisphere to mirror the starting Zenith East point.
Even given those two remedies, it might be better to be safe and turn off pointing correction (but leave on tracking rate correction of course).
-Ray
--
|
|
|
Re: Dec-Arc model vs. All Sky
>>>
Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
That's true, and there's a separate Dec Arc Model i.e., a limited model around a specific declination with an eye towards imaging one specific target
It's easy to get the dec arc model and the dec arc tracking algorithm mixed up, it seems to be used interchangeably in some threads.
On Tue, Oct 11, 2022 at 7:41 PM Ray Gralak < iogroups@...> wrote: > >>>What is surprising in this thread is folks complaining about scope pointing to the ground after a 30pt model.
> How does one manage to end up in such a state?
>
> just to be clear, i think we're talking about a dec arc pointing model, not an allsky model. So terms that are
> calculated are unique to a very narrow range in declination. I could see that causing issues for pointing outside of
> the model, particularly around the poles
Actually, it is called a Dec-Arc *Tracking* rate model. The algorithm is strictly about getting the best tracking performance. The algorithm does not correct pointing.
However, for pointing correction, APCC will fall back to an all-sky pointing model, if there is one. The problem with the "mini" Dec-Arc model is that all of the data points line up at about the same declination which when coupled with too few points could sometimes produce a ridiculously incorrect pointing model on the West side (pier-side=East). Pointing was usually good enough on the East side because of the starting Zenith point, which usually "stabilizes" the solution.
APCC now does two things to help prevent bad pointing:
1) it throws away any ridiculous pointing terms.
2) APPM automatically inserts a zenith point in the West hemisphere to mirror the starting Zenith East point.
Even given those two remedies, it might be better to be safe and turn off pointing correction (but leave on tracking rate correction of course).
-Ray
|
|
|
