Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Building 10Micron models and the things that affect models

  • Please log in to reply
118 replies to this topic

#51 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 9,407
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 10 February 2021 - 02:01 PM

Still a little unclear, if you can't go longer than 20 minutes, that means there is some drift after some time, and the target won't be centered anymore...am I not getting something?


You just "goto" the target again at intervals. The pointing doesn't drift but the tracking might wander in this time. My goto interval btw is twice a night - the second after  a meridian flip!

 

 

 

if you can't go longer than 20 minutes


You can with hybrid guiding - leave dual tracking on but do a auto-guider correction at a low rate. I use 4 corrections a minute.

I've tested this up to 2 hours single exposures and it works very well. However you do need a very good temperature tracking focuser to do this and have calibrated the temperature/focus position profile well. I use an Optec Gemini to keep the focus sharp over these huge time span exposures.


Edited by Tonk, 10 February 2021 - 02:02 PM.

  • EFT and psandelle like this

#52 dan_1984

dan_1984

    Mariner 2

  • -----
  • Posts: 245
  • Joined: 05 Jun 2016
  • Loc: Europe

Posted 10 February 2021 - 02:13 PM

Thanks for the explanations. I get it now. Seems amazing

#53 SimonIRE

SimonIRE

    Apollo

  • -----
  • Posts: 1,161
  • Joined: 16 Dec 2018
  • Loc: South East England

Posted 10 February 2021 - 03:30 PM

Is the there any reason to think model accuracy is impacted by sensor size. I just ran a 30 point model using a 6inch refractor and an ASI1600mm. PAE was 23” with an RMS of 4.8”. I swapped over the camera to a QHY600 and its 1’ 23” with an RMS of 2.2”. This is on a GM4000 with about 140kg of gear and counterweights on it so there is almost no way the mount moved.

Any thoughts?

Edited by SimonIRE, 10 February 2021 - 03:30 PM.


#54 SimonIRE

SimonIRE

    Apollo

  • -----
  • Posts: 1,161
  • Joined: 16 Dec 2018
  • Loc: South East England

Posted 11 February 2021 - 03:16 AM

Incidentally,

 

I am looking for some reasonable literature on the rationale for NOT choosing stars near Polaris for polar align routine in the mount and why the best stars to use are those near the intersection of the meridian and the celestial equator.

 

Tonk mentioned in the older 10 Micron thread that the explanation was in the old Losmandy Gemini manual but I couldn't find this manual online. 

 

Anyone point me in the right direction please?



#55 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 9,407
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 11 February 2021 - 05:25 AM

the rationale for NOT choosing stars near Polaris for polar align routine


The non-linearility of longitude. Resolution is very low at pole and max at equator. I.e. subtend the same absolute angular distance East or West near pole and measure the longitude difference and do same at equator - its coarse versus fine. Think about it - whats the longitude when pointing direct at a pole? (is this a Cartesian to polar coordinate problem?)

 

 

 

but I couldn't find this manual online.


I just Googled for "losmandy gemini manual" - first hit is http://www.losmandy....i_manual_l4.pdf !!!!

The section to read is 4.3.2.3 - however other than telling you how to polar align using the model (exact same as 10Micron) and telling you to use an alignment star at the equator/meridian intersection, the deeper explanation I remember is absent - so that must have come from elsewhere.


Edited by Tonk, 11 February 2021 - 05:53 AM.

  • psandelle likes this

#56 SimonIRE

SimonIRE

    Apollo

  • -----
  • Posts: 1,161
  • Joined: 16 Dec 2018
  • Loc: South East England

Posted 11 February 2021 - 05:48 AM

The non-linearility of longitude. Resolution is very low at pole and max at equator. I.e. subtend the same absolute angular distance East or West near pole and measure the longitude difference and do same at equator - its coarse versus fine. Think about it - whats the longitude when pointing direct at a pole? (is this a Cartesian to polar coordinate problem?)





I just Googled for "losmandy gemini manual" - first hit is http://www.losmandy....i_manual_l4.pdf !!!!

The section to read is 4.3.2.3 - however other than telling you how to polar align using the model (exact same and 10Micron) and telling you to use an alignment star at the equator/meridian intersection, the deeper explanation I remember is absent - so that must have come from elsewhere.

Cheers. I found that manual alright but couldn’t find the section in the 137 pages or so. As you note, the detailed explanation you alluded to in a previous post wasn’t in this manual. Thanks for the heads up

Edited by SimonIRE, 11 February 2021 - 05:50 AM.


#57 tocster

tocster

    Vostok 1

  • -----
  • Posts: 198
  • Joined: 25 Sep 2016
  • Loc: Melbourne, Australia

Posted 11 February 2021 - 08:49 AM

 I'm interested to know what astro imaging tools you guys use, and what the workflow is.  I need to setup every time so I'm looking to make things a bit less complicated.     

 

I would imagine that you:

- connect to the mount

- do a polar alignment (I normally would just home my mount and use sharpcap for this)

- initiate a modelling with ModelMaker/MountWizzard in Voyager/SGP etc...

- Tell the mount to go to a target

- Image.

 

How long does the modelling step take for a reasonable model?  I would imagine that a longer FL scope will require more modelling points?

If I have a very good polar alignment, will that reduce the number of modelling points needed?


Edited by tocster, 11 February 2021 - 08:52 AM.


#58 Lead_Weight

Lead_Weight

    Surveyor 1

  • *****
  • Posts: 1,957
  • Joined: 21 Jul 2016
  • Loc: Houston

Posted 11 February 2021 - 09:16 AM

 I'm interested to know what astro imaging tools you guys use, and what the workflow is.  I need to setup every time so I'm looking to make things a bit less complicated.     

 

I would imagine that you:

- connect to the mount

- do a polar alignment (I normally would just home my mount and use sharpcap for this)

- initiate a modelling with ModelMaker/MountWizzard in Voyager/SGP etc...

- Tell the mount to go to a target

- Image.

 

How long does the modelling step take for a reasonable model?  I would imagine that a longer FL scope will require more modelling points?

If I have a very good polar alignment, will that reduce the number of modelling points needed?

I use a Raspberry Pi with StellarMate at the telescope, and run an ethernet cable out to it using a Powerline adapter in my back yard and have a switch at the mount so I can connect it to ethernet along with the StellarMate. 

  • Turn the mount on
  • Turn on the Raspberry Pi
  • Launch EKOS on the Stellarmate and connect to all the equipment, turn on the camera cooler
  • Polar align with the hand control (I've also used MountWizard4 for this, but sometimes get mixed results, likely a user error with not clearing a model)
  • Launch MountWizard
  • Point to a star, focus
  • Load points (pre-saved) in MountWizard, hit run model. I run a 60 point model since my home blocks a portion of the sky and I can't physically get more points into the model. I have the mount set to 8°/sec speed. 15° is too fast for my blood. The model takes around 10 minutes to complete, I have a 3 second settle time, and the EKOS plate solver has been rewritten, and can now solve in a fraction of a second, so it's pretty quick.

At this point, I'm ready to go for the night. I usually plug in target and sequence into the EKOS scheduler, tell it when to start (astronomical twilight, ° above horizon, or specific time), tell it when to end (same parameters), and loop the sequence until auto shut down.



#59 RazvanUnderStars

RazvanUnderStars

    Apollo

  • *****
  • Posts: 1,135
  • Joined: 15 Jul 2014
  • Loc: Toronto, Canada

Posted 11 February 2021 - 11:07 AM

The model is needed in order to find out how the entire system deviates from a perfect system (perfectly rigid, with mount axes perfectly perpendicular, no slack, no atmospheric refraction and so on). The number of points needed is fundamentally driven by two factors:

  • how imperfect the setup is in various azimuths and at various altitudes. The less rigid or less consistent the setup is, the more points are needed to characterize its behaviour.
  • where you image in the sky. If you only plan to image one object in a given night, you don't need model points far from the trajectory of that object in the sky, so you can use fewer. 
  • the focal length. A higher FL will make deviations from pinpoint stars more visible.
  • the exposure time. Longer times will allow deviations from pinpoint stars to become more visible.

 

The above factors do not depend on the accuracy of the polar alignment. Below is a quote from Filippo Riccio, one of the 10Micron software developers (he posted it on the 10Micron user forum) that clarifies things further:

 

 

"Effects of polar misalignment

 

Even if your mount is wildly away from a correct polar alignment, when you make a model the correction of the polar misalignment will be performed exactly.

 

This means that the pointing error and the tracking error will be the same - if you make a model with the same number of stars and the same accuracy - whether the mount is aligned with the celestial pole or not (note that the mount will prevent making a model if the error is too large, but this is only for sanity of the results).

 

This, of course, happens only if the dual axis tracking is selected. When dual axis tracking is deactivated, the mount will track like a normal equatorial mount, by moving the right ascension axis at a constant sidereal speed. The only important thing that will change is the field rotation"

 

 

 

How long does the modelling step take for a reasonable model?  I would imagine that a longer FL scope will require more modelling points?

If I have a very good polar alignment, will that reduce the number of modelling points needed?


Edited by RazvanUnderStars, 11 February 2021 - 11:11 AM.

  • psandelle likes this

#60 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 6,234
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 11 February 2021 - 11:48 AM

Is the there any reason to think model accuracy is impacted by sensor size. I just ran a 30 point model using a 6inch refractor and an ASI1600mm. PAE was 23” with an RMS of 4.8”. I swapped over the camera to a QHY600 and its 1’ 23” with an RMS of 2.2”. This is on a GM4000 with about 140kg of gear and counterweights on it so there is almost no way the mount moved.

Any thoughts?

What might effect model quality is image resolution, not sensor size, unless you can't get enough stars on a plate to plate solve.  Even slightly unfocused stars will be OK.  



#61 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 9,407
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 11 February 2021 - 12:27 PM

likely a user error with not clearing a model


The 10 Micron controller automatically wipes the model after using the built-in polar align step - so no

#62 Lead_Weight

Lead_Weight

    Surveyor 1

  • *****
  • Posts: 1,957
  • Joined: 21 Jul 2016
  • Loc: Houston

Posted 11 February 2021 - 12:58 PM


The 10 Micron controller automatically wipes the model after using the built-in polar align step - so no

I was referring to using MW4 to polar align. So somewhere in there it has caused a problem for me. Using the hand control, I've never had an issue.



#63 mworion

mworion

    Explorer 1

  • -----
  • Posts: 53
  • Joined: 08 Mar 2016
  • Loc: Hohenbrunn / Munich

Posted 11 February 2021 - 02:58 PM

Hi Andrew,

MW4 deletes always the model before starting a new one. But you're right, if you do mechanical polar aligning without building a new model, you kept the old, which is invalid.  

Michel


  • Lead_Weight likes this

#64 Lead_Weight

Lead_Weight

    Surveyor 1

  • *****
  • Posts: 1,957
  • Joined: 21 Jul 2016
  • Loc: Houston

Posted 11 February 2021 - 03:15 PM

Hi Andrew,

MW4 deletes always the model before starting a new one. But you're right, if you do mechanical polar aligning without building a new model, you kept the old, which is invalid.  

Michel

To be clear, I think I am the cause of the problem, wherever it is, when I attempted to polar align with MW4. I have had success at times, but there must have been something I did incorrectly somewhere in the process that it does not always work for me.



#65 mworion

mworion

    Explorer 1

  • -----
  • Posts: 53
  • Joined: 08 Mar 2016
  • Loc: Hohenbrunn / Munich

Posted 11 February 2021 - 03:20 PM

Andrew,

If you actually use the support of MW4: Selecting an align star from hemisphere, click for center it and then using crosshair to adjust -> There is now way how to find out if you finished adjusting mechanically. It would be nice detecting that and deleting the model automatically....

Michel 



#66 Lead_Weight

Lead_Weight

    Surveyor 1

  • *****
  • Posts: 1,957
  • Joined: 21 Jul 2016
  • Loc: Houston

Posted 11 February 2021 - 03:44 PM

Andrew,

If you actually use the support of MW4: Selecting an align star from hemisphere, click for center it and then using crosshair to adjust -> There is now way how to find out if you finished adjusting mechanically. It would be nice detecting that and deleting the model automatically....

Michel 

If I remember correctly, I believe my process was to clear the model, plate solve 3-7 random areas of the sky by running a small model with MW4 after boot up. Then select Polar/Ortho align from the Hemisphere window, right click on a star in the window to slew to it. Use live preview, or 1 sec loop of images to make manual alt/az adjustments to center the star. I have also tried the method by using the recommended turns suggested (this doesn't work for me). Maybe I'm missing a step? 



#67 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 9,407
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 11 February 2021 - 03:45 PM

If you actually use the support of MW4: Selecting an align star from hemisphere, click for center it and then using crosshair to adjust -> There is now way how to find out if you finished adjusting mechanically. It would be nice detecting that and deleting the model automatically....

Michel



So if you follow how the keypad does it - if I select Polar Align to select a star then the mount knows I have finished building the model and wipes the model when I press the Enter key at the end of the Polar Align step (after I have mechanically adjusted to center the cross hairs - It doesn't know if I've actually done that - but pressing Enter will do the wipe because I've been in Polar Align mode) ....

So in the software if you allow the user to declare they are polar aligning (press a button "Polar Align") then how about a modal dialog where the user must press "Enter" to say they are done (and nothing other then the "Enter" button is active) - then the software wipes the model and tells the user that the model must now be rebuilt. You could also do a plate solve after the Enter button is pressed to check the mechanical adjustment was actually done (something the mount keypad route doesn't support)


Edited by Tonk, 11 February 2021 - 03:48 PM.


#68 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 9,407
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 11 February 2021 - 03:47 PM

I have also tried the method by using the recommended turns suggested (this doesn't work for me)


It doesn't work for *anybody* - unless you have graduated the knob with 1/100th turn scale! The only sure way is cross hairs - either software or an astrometric eyepiece
  • psandelle and Lead_Weight like this

#69 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 6,234
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 11 February 2021 - 04:01 PM

While it's not the best method for polar aligning and I never use it, it does "work" if you consider the fact that polar alignment doesn't have to be perfect with a dual-axis tracking mount.  Close is what you need and the mount model will adjust for the polar misalignment.  Field rotation is the only potential issue.



#70 Arie

Arie

    Messenger

  • *****
  • Posts: 494
  • Joined: 01 Sep 2015
  • Loc: Netherlands

Posted 11 February 2021 - 04:24 PM

FWIW

I also am having problems with PA through MW4

When I select a alignment star from the hemisphere window and slew to it, next adjust the alt/az screws to center it and then run a 9 point model, the PAE is multi-minutes.

When I use the HC PA routine and use a star from the HC, the PAE is @30arcsec straightaway.

I can't figure out what I am doing wrong in MW4.

I still have a polemaster laying around which connects with a little magnet to the top cover of the mount.

With the mount powered off and the RA clutch slightly loose I do a PA within 2 minutes.

A 25 point model consequential build wit MW4 gives me a PAE <1arcmin. Which is good enough for me.

 

Besides, my camera atik16200 has looooooooong download times, which makes PA a tedious exercise.



#71 mworion

mworion

    Explorer 1

  • -----
  • Posts: 53
  • Joined: 08 Mar 2016
  • Loc: Hohenbrunn / Munich

Posted 11 February 2021 - 04:30 PM

If I remember correctly, I believe my process was to clear the model, plate solve 3-7 random areas of the sky by running a small model with MW4 after boot up. Then select Polar/Ortho align from the Hemisphere window, right click on a star in the window to slew to it. Use live preview, or 1 sec loop of images to make manual alt/az adjustments to center the star. I have also tried the method by using the recommended turns suggested (this doesn't work for me). Maybe I'm missing a step? 

No, everything OK. The turns are only good if you are far away. Best is if you could center the star. 


  • psandelle likes this

#72 mworion

mworion

    Explorer 1

  • -----
  • Posts: 53
  • Joined: 08 Mar 2016
  • Loc: Hohenbrunn / Munich

Posted 11 February 2021 - 04:33 PM

FWIW

I also am having problems with PA through MW4

When I select a alignment star from the hemisphere window and slew to it, next adjust the alt/az screws to center it and then run a 9 point model, the PAE is multi-minutes.

When I use the HC PA routine and use a star from the HC, the PAE is @30arcsec straightaway.

I can't figure out what I am doing wrong in MW4.

I still have a polemaster laying around which connects with a little magnet to the top cover of the mount.

With the mount powered off and the RA clutch slightly loose I do a PA within 2 minutes.

A 25 point model consequential build wit MW4 gives me a PAE <1arcmin. Which is good enough for me.

 

Besides, my camera atik16200 has looooooooong download times, which makes PA a tedious exercise.

Arie,

thanks for the hint. I will have a look to it. Tomorrow weather should be good after months without clear skies.

Michel



#73 Arie

Arie

    Messenger

  • *****
  • Posts: 494
  • Joined: 01 Sep 2015
  • Loc: Netherlands

Posted 12 February 2021 - 07:33 AM

Michel,

Last time I used Betelgueze for a PA, which was due south at the equator for me.

Slewed to it by MW4. Centered it with the alt/az screws.

Reran a 9 pt model. 6 arcmin PAE.

Slewed again to Betelgeuze by MW4.

Missed it badly on the left side. Centered it. Took quite some adjustment.

Reran 9 pt model. Again 6 arcmin PAE.

Slewed to Betelgeuze. Missed it badly on the right side.

Repeated this a few times, but I kept on swinging from left to right and the PAE did not get any better.

Finally slewed to Betelgeuze with the HC PA. Centered it and pressed Enter.

Reran a 9 pt model. PAE was 20arcsec!

 

This is my experience.

For model building MW4 is next to perfect, but the PA does not work for me.

Could it be that the align star coordinates in MW4 are off?



#74 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 9,407
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 12 February 2021 - 08:59 AM

I still use the keypad for the polar align step. No problems whatsoever and gets me 15" to 20"


  • psandelle likes this

#75 Lead_Weight

Lead_Weight

    Surveyor 1

  • *****
  • Posts: 1,957
  • Joined: 21 Jul 2016
  • Loc: Houston

Posted 12 February 2021 - 10:45 AM

Michel,

Last time I used Betelgueze for a PA, which was due south at the equator for me.

Slewed to it by MW4. Centered it with the alt/az screws.

Reran a 9 pt model. 6 arcmin PAE.

Slewed again to Betelgeuze by MW4.

Missed it badly on the left side. Centered it. Took quite some adjustment.

Reran 9 pt model. Again 6 arcmin PAE.

Slewed to Betelgeuze. Missed it badly on the right side.

Repeated this a few times, but I kept on swinging from left to right and the PAE did not get any better.

Finally slewed to Betelgeuze with the HC PA. Centered it and pressed Enter.

Reran a 9 pt model. PAE was 20arcsec!

 

This is my experience.

For model building MW4 is next to perfect, but the PA does not work for me.

Could it be that the align star coordinates in MW4 are off?

This is essentially what I experience when trying to do PA with MW4. 




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics