APPM and ASTAP #APCC


Bill Conrad
 

I have APCC 1.9.0.11 and latest ASTAP. When I try to run a pointing model the ASTAP solve always fails. If I leave the scope at the run target location and press the Plate Solve button, the solve is always  successful in a few seconds. What am I missing?

Bill Conrad


Ray Gralak
 

Bill,

There was talk about this last week on this forum, so you may want to look back on those posts. The short answer is that APPM uses ASTAP's "automatic mode" for calculating the view/image scale field during a regular APPM run with the expectation that ASTAP will accurately calculate image scale. However, ASTAP's automatic mode is not very automatic.

So, you may have to configure the ASTAP's settings in APPM to get ASTAP to solve images. I am working on a fix to get around ASTAP's unexpected behavior, but it is not ready.

In the longer term, I am considering porting just ASTAP's plate solving logic to C#/.NET so that I can make plate solving work in a way that is most efficient for future APPM features.

-Ray

-----Original Message-----
From: main@ap-gto.groups.io [mailto:main@ap-gto.groups.io] On Behalf Of Bill Conrad
Sent: Monday, October 11, 2021 11:39 AM
To: main@ap-gto.groups.io
Subject: [ap-gto] APPM and ASTAP #APCC

I have APCC 1.9.0.11 and latest ASTAP. When I try to run a pointing model the ASTAP solve always fails. If I
leave the scope at the run target location and press the Plate Solve button, the solve is always successful in
a few seconds. What am I missing?

Bill Conrad


Brent Boshart
 

Hi Bill,

I had experienced the same as you are describing.  When I changed "Min. Star Size" under "Plate Solve Settings" from "0" to "1.5" it has worked reliably ever since.  Setting it "0" and it is supposed use the default setting in ASTAP but it did not seem to work for me. 

Brent


Bill Conrad
 

Thanks Brent. I'll try that next clear sky.

BC


Erkaslan Aygen
 

Hi Bill,

Did you have any success while changing the settings ? 

I am just asking since I am also experiencing the same issue: Plate solving is pretty much always failing while using ASTAP. However, when used "outside" APPM, it's working as a breeze and solving a couple of secs. 

I did increase the "min star" size to 1.5 (for example), but it didn't change my situation. 

As for my equipment :
  • ASI 2600 MM PRO used with a TAK FSQ 85 (2.73 resolution)
Cheers,
Aygen


Ray Gralak
 

Hi Aygen,

I am just asking since I am also experiencing the same issue: Plate solving is pretty much always failing while
using ASTAP. However, when used "outside" APPM, it's working as a breeze and solving a couple of secs.
Is this with version 1.9.0.11, or the new version 1.9.1.1 beta with a fix?

https://www.apastrosoftware.com/apcc_download/APCC_Pro_BETA_Setup_1.9.1.1.exe

-Ray

-----Original Message-----
From: main@ap-gto.groups.io [mailto:main@ap-gto.groups.io] On Behalf Of Erkaslan Aygen
Sent: Friday, October 22, 2021 1:23 AM
To: main@ap-gto.groups.io
Subject: Re: [ap-gto] APPM and ASTAP #APCC

Hi Bill,

Did you have any success while changing the settings ?

I am just asking since I am also experiencing the same issue: Plate solving is pretty much always failing while
using ASTAP. However, when used "outside" APPM, it's working as a breeze and solving a couple of secs.

I did increase the "min star" size to 1.5 (for example), but it didn't change my situation.

As for my equipment :


* ASI 2600 MM PRO used with a TAK FSQ 85 (2.73 resolution)

Cheers,
Aygen


Erkaslan Aygen
 

Hi Ray,

This was with the official version (not the beta). I could give it try tonight with the beta.

Aygen

Le ven. 22 oct. 2021 à 15:48, Ray Gralak <iogroups@...> a écrit :
Hi Aygen,

> I am just asking since I am also experiencing the same issue: Plate solving is pretty much always failing while
> using ASTAP. However, when used "outside" APPM, it's working as a breeze and solving a couple of secs.

Is this with version 1.9.0.11, or the new version 1.9.1.1 beta with a fix?

https://www.apastrosoftware.com/apcc_download/APCC_Pro_BETA_Setup_1.9.1.1.exe

-Ray

> -----Original Message-----
> From: main@ap-gto.groups.io [mailto:main@ap-gto.groups.io] On Behalf Of Erkaslan Aygen
> Sent: Friday, October 22, 2021 1:23 AM
> To: main@ap-gto.groups.io
> Subject: Re: [ap-gto] APPM and ASTAP #APCC
>
> Hi Bill,
>
> Did you have any success while changing the settings ?
>
> I am just asking since I am also experiencing the same issue: Plate solving is pretty much always failing while
> using ASTAP. However, when used "outside" APPM, it's working as a breeze and solving a couple of secs.
>
> I did increase the "min star" size to 1.5 (for example), but it didn't change my situation.
>
> As for my equipment :
>
>
> *     ASI 2600 MM PRO used with a TAK FSQ 85 (2.73 resolution)
>
> Cheers,
> Aygen
>







Bill Conrad
 

Aygen:

I have not had any luck adjusting the parameters.
The ImageLink Test also fails. When the ImageLink Test dialog pops up, ASTAP has computed a FOV twice what it should be. If I enter the correct FOV in the dialog then ImageLink Test is successful.
I too will try Ray's beta version, but I can't until the first week of Nov.

BC


Bill Conrad
 

Got a model built using ASTAP with Beta 1.9.1.3. Did not work at bin 2, but did work at bin 1 with min star at 1.0.
Thanks Ray.

BC


Worsel
 

Ray

 

Thanks for the new BETA 1.9.1.3.  A few comments follow.  Your insight IS appreciated.    I hope these are helpful to others.

·        Program installed without issue. (1100 GT non-AE)

·        Following using

o   FOV = 135’ x 90’

o   Image scale = 1.1”/pxl

o   70 MP fits file.   No need to bin.   APPM would not read the original file. Cropped to 1/4x1/4.  Cropped image also downloads faster and solves fast.  No brainer there.

·        Initially, astrometry.net (local install) would solve, but ASTAP and, hence, APPM would not.

·        Took significant fiddling with gain, exposure, etc to come up with images that will solve.  Some images that would solve in ASTAP directly would not solve in APPM. 

·        Finally got everything lined out sufficiently to build a 227 pt model (Large Map default) with only 3 Solve Fails and those point were on the horizon.   Using this model, I was able to capture up to 15 minute subs un-guided.  I did not test beyond 15 minutes.

·        I also solved images from following train

o   FOV = 27 x 18’

o   Image scale = 0.4’/pxl

o   22 MP fits.  Uncropped, no need to bin.

·        Tips (at least these helped me)

o   Getting the image close is key (Ray says this in the documentation as well.)

o   However, setting Image scale tolerance to 50% gave me more latitude. May increase solve time, but only by small amounts.

o   Solve button will work when Blind Solve does not on the same image.

o   The initial RA and Dec only needs to be accurate to 1 hour and 1 degree, e.g. 3 50 will solve an image with center at

§  3h 23m 15s

§  49d 43m 36s

One comment.  People mention ‘setting’ the FOV.   As near as I can determine, there is no place to set the FOV directly.  APPM can calculate FOV from the sensor parameters, which is supplied from FITS header, and image scale, which is set in APPM and sent to FITS header.  Is this correct?

 

Bryan


Ray Gralak
 

Bryan,

 

Regarding ASTAP plate solves, you can look at the APPM logs to see the command line arguments passed to APPM. If you set all of ASTAP values to zero in APPM’s plate solve tab and check the box to use the FITS header, then APPM will pass just the filename to APPM. If an image is not solving it’s likely a setting in ASTAP needs to be tweaked. APPM cannot predetermine which settings will best work with ASTAP, which is why it exposes controls to tweak command line parameters.

 

Regarding the 70 Mpix image not loading, I assume you are using the camera’s ASCOM Camera driver?  There are some VERY inefficient ASCOM drivers that require over 1 GB of ram for a 40 mpixel image. The maximum memory for APPM is 1GB so the image fails to load. If the manufacturers would create efficient ASCOM camera drivers then APPM would not have a problem. Binning or using a subframe will reduce memory size below 1GB and thus APPM can use the image.

 

Lastly, FOV is something ASTAP needs to reduce plate solve time. With the default values ASTAP will use the FITS header to calculate FOV.

 

-Ray

 

From: main@ap-gto.groups.io [mailto:main@ap-gto.groups.io] On Behalf Of Worsel via groups.io
Sent: Tuesday, November 2, 2021 7:41 AM
To: main@ap-gto.groups.io
Subject: Re: [ap-gto] APPM and ASTAP #APCC

 

Ray

 

Thanks for the new BETA 1.9.1.3.  A few comments follow.  Your insight IS appreciated.    I hope these are helpful to others.

·        Program installed without issue. (1100 GT non-AE)

·        Following using

o   FOV = 135’ x 90’

o   Image scale = 1.1”/pxl

o   70 MP fits file.   No need to bin.   APPM would not read the original file. Cropped to 1/4x1/4.  Cropped image also downloads faster and solves fast.  No brainer there.

·        Initially, astrometry.net (local install) would solve, but ASTAP and, hence, APPM would not.

·        Took significant fiddling with gain, exposure, etc to come up with images that will solve.  Some images that would solve in ASTAP directly would not solve in APPM. 

·        Finally got everything lined out sufficiently to build a 227 pt model (Large Map default) with only 3 Solve Fails and those point were on the horizon.   Using this model, I was able to capture up to 15 minute subs un-guided.  I did not test beyond 15 minutes.

·        I also solved images from following train

o   FOV = 27 x 18’

o   Image scale = 0.4’/pxl

o   22 MP fits.  Uncropped, no need to bin.

·        Tips (at least these helped me)

o   Getting the image close is key (Ray says this in the documentation as well.)

o   However, setting Image scale tolerance to 50% gave me more latitude. May increase solve time, but only by small amounts.

o   Solve button will work when Blind Solve does not on the same image.

o   The initial RA and Dec only needs to be accurate to 1 hour and 1 degree, e.g. 3 50 will solve an image with center at

§  3h 23m 15s

§  49d 43m 36s

One comment.  People mention ‘setting’ the FOV.   As near as I can determine, there is no place to set the FOV directly.  APPM can calculate FOV from the sensor parameters, which is supplied from FITS header, and image scale, which is set in APPM and sent to FITS header.  Is this correct?

 

Bryan


Sébastien Doré
 

Hi Bryan,

People mention ‘setting’ the FOV.   As near as I can determine, there is no place to set the FOV directly. APPM can calculate FOV from the sensor parameters, which is supplied from FITS header, and image scale, which is set in APPM and sent to FITS header.  Is this correct?


You have already figured the following out but for the sake of other people who wouldn't have yet the FOV for ASTAP is the field height (in degrees) of the image you're trying to solve. So it is related to your "vertical" sensor size (in pixels) and your image scale (arc-sec per pixel), through the following formula:


FOV_height = vertical_height * image_scale / 3600


and can be set directly in ASTAP in the Stack Menu -> Alignment Tab. (That answers in part your above question)


For example, my ASI183MC Pro sensor which has a 5496 x 3672 pixels and used with a 70mm refractor at 338 mm FL. Image scale for this setup is 1.4646 arc-sec/pixel so FOV height to set in ASTAP will be : 


3672 px x 1.4646 arc-sec/px / 3600 arc-sec/deg = about 1.49 degree. 


That's the value ASTAP needs to work the most efficiently, all others being equal. 




With the new Beta version of APCC (1.9.1.3), this FOV is automatically calculated from the FITS header information, as you said. And it is used by APPM when launching the ASTAP platesolving process if you have checked the new "Use FITS header for RA, Dec and Image scale" checkbox. So the need for you to provide the FOV manually to ASTAP is now gone as APPM correctly does the work for you. That assumes your imaging program includes all the required information in the FITS header in the first place (which NINA 1.11 does), of course. 


Prior to APCC v1.9.1.1, APPM forced the FOV setting to be determined by ASTAP through trial and error (i.e. not using the manually user-supplied value as seen above) which added some unnecessary time to solving each point of a mapping run. So it is a great bug-fix in the latest version in this regard, especially for those using somewhat old/slow imaging PC by today's standard.


Hope that helps understand it a bit further.


Sébastien


Worsel
 

Thanks, Ray!

You said

"Regarding the 70 Mpix image not loading, I assume you are using the camera’s ASCOM Camera driver?  There are some VERY inefficient ASCOM drivers that require over 1 GB of ram for a 40 mpixel image. The maximum memory for APPM is 1GB so the image fails to load. If the manufacturers would create efficient ASCOM camera drivers then APPM would not have a problem. Binning or using a subframe will reduce memory size below 1GB and thus APPM can use the image."

You are correct.  As far as I know, APPM can only use ASCOM drivers to connect a camera.  However, for me, point is moot, since I crop and get a smaller size file that APPM solves well!  The addition of ASTAP is a great time-saver for model builds.

Bryan


Ray Gralak
 

Hi Brian,

As far as I know, APPM can only use ASCOM drivers to connect a camera.
APPM can also use NINA, SGPro, SkyX, CCDSoftV5, and MaximDL to take images through their automation interfaces.

-Ray