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

Question on camera rotation doing mosaic with NINA

  • Please log in to reply
24 replies to this topic

#1 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 09 October 2019 - 11:15 AM

Hi all,

 

Lately I am doing some mosaics planned with NINA and have come to

some strange behaviour ( in my opinion ) such that the panes composing

the mosaic are displaced in between them.

My question now is: do I have to rotate the camera like in the angle as

solved with platesolver2 in order to get these panes aligned between them??

I'm not sure if I explain myself correctly but include 3 images to show 1) the

way I planned the mosaic in NINA 2) the displacement and 3) Platesolve2 result with the angle

 

Should I rotate the camera with this angle so all will be aligned??

 

Any feedback is highly apreciated since I'd like to do the mosaicing well and

not the way it goes nowadays, have to learn!!

 

 

 

Attached Thumbnails

  • Pickering 3P.jpg
  • NGC 6979-100519-3pane-3x30x120-RC8-reduc-Triad-Villareal-V2-CN-question.jpg
  • Platesolve result.jpg


#2 OldManSky

OldManSky

    Surveyor 1

  • *****
  • Posts: 1927
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 09 October 2019 - 12:01 PM

Hey Mert, 

That doesn't look like a rotation issue -- it looks like a plate solving tolerance issue.

In the plate solving setup tab, it has an "error <" field...the default is 2 arcminutes, which is way too much for a successful mosaic.  Try setting that to 0.2 or 0.1 instead, that will get the image within a few dozen pixels of exactly where it needs to be.  Also make sure you have "Sync," "Reslew to Target," and "Repeat Until" checked so that it will repeat the move/platesolve/sync/recenter sequence until it's within the error < value.



#3 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 09 October 2019 - 12:36 PM

Hi Paul,

 

You are right about the tolerance, which at the moment is at 2

arcminutes.

Because I had problems with Platesolver2 I switched over to

ASTAP and after installing the g17_star_database_mag17

library, ASTAP is really fast in solving, so I will try and adjust

the tolerance as you suggest.

My thinking behind the question was if I rotate these 3 panes

with the same angle maybe they can be made to align.

 

So if I understand correctly, you think the planning and the result

are correct apart from the tolerance the platesolving had, is

that correct?

 

Thanks for your feedback!

-Mert


Edited by Mert, 09 October 2019 - 12:36 PM.

  • Swordfishy likes this

#4 OldManSky

OldManSky

    Surveyor 1

  • *****
  • Posts: 1927
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 09 October 2019 - 01:34 PM

Yes.

Look at the planned (in the framing screenshot) second pane down -- in the plan, there should be about 1/8 of a frame space between the wispy part of the nebula at the right, but in your actual shot, it's right up against the right edge.  That tells me that the frame wasn't centered where you wanted it to, and it was off about 2 arcminutes to the left.  Similarly, for the 3rd frame down, the planner shows about 1/8 of a frame to the left of the nebula at the bottom, but in your actual shot it's right up against the left edge -- again, about 2 arcminutes off.  So your mount's default GOTO is off a bit, and the error tolerance wasn't small enough to correct it during platesolve.

 

Note that the rotation in the solved plate shows your *actual* rotation, so you can't adjust anything to match that, it's what you are already at!

 

I think if you make the error tolerance smaller, it will line right up.  I just finished last night the shots for a 4-pane mosaic of this same area, and with a small error tolerance they all line up with no drift to one side or the other.



#5 bmhjr

bmhjr

    Apollo

  • ****-
  • Posts: 1254
  • Joined: 02 Oct 2015
  • Loc: Colorado

Posted 09 October 2019 - 02:08 PM

Yes, your rotation angle is off.  In the NINA plan the nebula aligns vertically, but your image panes when aligned show the nebula across the frames at an angle.  Tolerance for platesolve aside, the angle does not match the plan.



#6 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 09 October 2019 - 02:17 PM

Yes, that sounds like what I was suspecting, the plan looks like the

nebula is vertically aligned with the panes but in reality there is an

angle.

According to platesolve2 that angle is aprox. 165º, or 15º off from

where I have the camera, looks about right doesn't it?

 

So, :thinking: maybe I should take 1 image of a pane, calculate the

angle, rotate the camera accordingly and then start the sequence doing

platesolving etc.

Does that sound ok?



#7 bmhjr

bmhjr

    Apollo

  • ****-
  • Posts: 1254
  • Joined: 02 Oct 2015
  • Loc: Colorado

Posted 09 October 2019 - 02:20 PM

I'm not familiar with NINA and if there are any options to prompt for manual rotation.  I assume the plan is showing 0 deg rotation?  Anyway, I think what you have proposed should work. Nice images BTW.



#8 tkottary

tkottary

    Ranger 4

  • *****
  • Posts: 300
  • Joined: 06 Dec 2015
  • Loc: San Jose,CA

Posted 09 October 2019 - 02:23 PM

Nina follows different rotation angle 

 

Nina 0-360 CW
ASTAP O-360 CCW



#9 deepanshu29

deepanshu29

    Vostok 1

  • -----
  • Posts: 127
  • Joined: 16 Jul 2017

Posted 09 October 2019 - 03:06 PM

This IS a rotation issue. Enable "Manual Rotator" in equipment tab. When you start your sequence, after plate solving it will tell you how much rotation you need to fix this. Keep in mind, NINA mosaic planner calculates the rotation 0-360 CW, ASTAP calculates 0-360 CCW. Since you want 180, you will be fine. Also, when NINA told me to rotate Xº CW, I realized that I had to rotate Xº CCW because image is inverted from camera. 

 

And why this is a rotation issue, look at the object in your top frame vs in your plan, it is titled. 

 

I just ran an exercise couple days back for 8 panel mosaic and learned few things hard way. 



#10 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 09 October 2019 - 03:22 PM

Thanks for confirming this, I was really wondering what was going

wrong.

For example platesolving get's the center of the image to a couple

of pixels centered, even meridian flip after retrying a couple of times

get's within some 10 to 20 pixels, not bad. ( at my image scale that

would equate to roughly 8 to 16 " )

Is there any economical device ( rotator ) that can be controlled

by NINA??

Have to check on that, for the moment I will make some markings

so that I can manually rotate the camera with certain precision.



#11 deepanshu29

deepanshu29

    Vostok 1

  • -----
  • Posts: 127
  • Joined: 16 Jul 2017

Posted 09 October 2019 - 03:32 PM

 

Is there any economical device ( rotator ) that can be controlled

by NINA??

Have been looking for one myself for a while. Only options I have found is either build one myself or just save up and get NiteCrawler one day. 

For now, I am rotating based on best guess and it is not too bad. For mosaic, you only have to get the correct angle only once, same rotation applies to all panels. With in couple of iterations I am able to get with in 1º rotation tolerance that I have setup in NINA. And 1-arcmin for plate solving. I am using ASTAP.



#12 OldManSky

OldManSky

    Surveyor 1

  • *****
  • Posts: 1927
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 09 October 2019 - 03:36 PM

I will (sort of) stand corrected.  The rotation *is* off.  My bad.

But the centers don't match up, either...so it's probably both.

 

Mert, if you scroll down a bit on the left of the framing tab, you'll see a slider for rotation.  Set that to match what your platesolver tells you your rotation is.  Then the plan will match the actual camera rotation.

But also make your plate-solve error distance smaller, so that the centers of the frames match up.



#13 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 09 October 2019 - 04:01 PM

Hi Paul, yes you are thinking in the same line as I do.

I will try to solve 1 frame, get the angle it tells me and

try and set this in NINA.

The bad part is that I am planning sequences at home

and lateron load them into my scope control pc.

So, once I have the sequence planned I don't know

how to redo that in the field without internet conection.

In NINA I can not reload the xml files in the framing tab

and adjust, might be a nice feature thinking1.gif

 

Oh well, just used the Skyatlas offline framing, searched for

NGC6979, once solved created a 3 pane mosaic, loaded

1 frame, solved and then had pane 1 on the screen.

Rotating to 165 degrees almost perfectly got it lined up.

Then moving pane1 of 3 to the image and saved as xml.

Then in the field I'll take 1 image, calculate the angle

and adjust the camera manually.

Maybe that works??

 

Apart from thinking in building a rotator myself, maybe the

Pixis LE is the most economical comercial rotator.


Edited by Mert, 09 October 2019 - 04:11 PM.

  • deepanshu29 likes this

#14 deepanshu29

deepanshu29

    Vostok 1

  • -----
  • Posts: 127
  • Joined: 16 Jul 2017

Posted 09 October 2019 - 04:14 PM

Mert,

 

That's what I do, plan at home, save the sequence as XMLs and then re-use them in field. This was the first time I ever tried to execute a mosaic so thought of running through all panels and captured only 10 minutes per panel. I am glad I did that as I found the discrepancy between how NINA reports the rotation angle vs what ASTAP reports. 

 

Here is my planned mosaic, with Telescopious (BTW, Telescopious uses same rotation standard as ASTAP) - 

ASTAP said rotation angle as 105º, NINA said 75º, I used 75º as I planned my sequence in NINA.

Planned mosaic
 
This is how the plan was executed - 
Executed mosaic

 

This is off by 30º because NINA framing assistant uses 0-360 CW and plate solver ASTAP uses 0-360 CCW. Now I know this, I will still plan my sessions using NINA framing assistant but will manually change the angle to ASTAP standard after generating the sequences. 



#15 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 09 October 2019 - 04:35 PM

One question on this is how you change the angle, do

you edit the xml files?



#16 deepanshu29

deepanshu29

    Vostok 1

  • -----
  • Posts: 127
  • Joined: 16 Jul 2017

Posted 09 October 2019 - 05:11 PM

One question on this is how you change the angle, do

you edit the xml files?

You can change it by either editing the XML files or from the Sequence pane in NINA, it is under the Target section. If you edit in NINA seq pane, it will not be reflected in XML unless you save again. 



#17 OldManSky

OldManSky

    Surveyor 1

  • *****
  • Posts: 1927
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 09 October 2019 - 08:12 PM

Mert, is there an affordable Camera Angle Rotator (manual, not necessarily motorized) available for your scope?  I have the William Optics one that goes in front of the flattener/reducer, and it has markings for being nice and repeatable...that might help.

 

Or, at the very least, you can plate solve one night, put a mark with painter's tape on the focuser and your camera attachment point, use that rotation value to plan, then when you reassemble line up the marks.  That should get you close enough, I think.

 

Below is the Ha channel for my 4-pane mosaic of the same area you're working on that I just got together.  Planned with NINA, with the rotation angle being about 1 degree off, and the plate-solving error value being 0.1 arcsec.  Just to show you it can work smile.gif

 

 

Attached Thumbnails

  • HaLayerAll4_3.jpg

  • Swordfishy and richorn like this

#18 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 10 October 2019 - 04:46 AM

Great information, thanks a lot!!!

#19 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 10 October 2019 - 09:12 AM

Just to be sure, if lucky I can test a bit tonight from the balcony, if I adjust the rotation, when solving in platesolve2 for example the angle it indicates will go to 0 when I aproach the correct angle???


Edited by Mert, 10 October 2019 - 12:43 PM.


#20 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 12 October 2019 - 05:46 AM

In order to have some idea on how much I have to rotate

the camera I made a ruler and taped it to the reducer's

tube.

So when rotating the reducer plus camera, I can estimate

the angle.

Here it roughly devides to 20º between the big division lines.

I hope it will help!

Would NINA when platesolve and conected the manual rotator

give me an angle indicator to rotate??

 

Attached Thumbnails

  • Drawtube devider.jpg

  • OldManSky likes this

#21 OldManSky

OldManSky

    Surveyor 1

  • *****
  • Posts: 1927
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 12 October 2019 - 10:00 AM

"Would NINA when platesolve and conected the manual rotator

give me an angle indicator to rotate??"

 

The platesolve will give you the angle you are *at* in the image that's been solved.  It won't tell you how much to rotate between where you're at and where you've set the rotation angle in the framing module.  You'll have to do that as mental math smile.gif

 

But your scale should help!  Nice idea.

 

I have my camera "upside down" from the normal orientation...platesolve initially reported an angle of -179 degrees when I lined it up by eye.  That was plenty close enough to 180 that I didn't tweak it, and used that setting in the framing wizard.  All the panels line up just fine.


Edited by OldManSky, 12 October 2019 - 10:05 AM.


#22 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 12 October 2019 - 10:07 AM

So let's asume I set the angle in NINA to something like 160º, when platesolving

NINA won't tell me how far I'm off but only tell me the angle I am at?

Did I understand you correctly Paul?

( Hard to understand how an automatic rotator could aid then )



#23 deepanshu29

deepanshu29

    Vostok 1

  • -----
  • Posts: 127
  • Joined: 16 Jul 2017

Posted 12 October 2019 - 10:18 AM

If you have manual rotator connected, and if you have Slew to Target & Center Target box checked and if you have angle values specified under Target Coordinates, NINA will give you a pop up specifying angle to rotate. When you are done rotating your setup manually just hit the 'X' to close that pop up, NINA will take a capture and plate solve again to validate angle. Accepted rotation tolerance is default 1 degree. You can change in Options > Plate Solving.

BTW, how did you make that scale so cleanly? Neat idea!

Edited by deepanshu29, 12 October 2019 - 12:05 PM.


#24 deepanshu29

deepanshu29

    Vostok 1

  • -----
  • Posts: 127
  • Joined: 16 Jul 2017

Posted 12 October 2019 - 10:21 AM

Keep in mind-
1. If you got planned angle from NINA framing assistant and using ASTAP as plate solver, they are both different standards. One is 0-360 CW, other is 0-360 CCW. I change the target angle by 180-target suggested by NINA framing assistant since I use ASTAP.
2. You will likely have to rotate opposite direction NINA is telling you to.
  • Mert, tkottary and OldManSky like this

#25 Mert

Mert

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 6316
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 12 October 2019 - 11:45 AM

Thanks a lot for your help on this!

I have to learn using it correctly and already have seen that

the angle NINA says is different from what ASTAP or Platesolve2

are saying.

For the ruler I used this website: https://robbbb.github.io/VectorRuler/

then saved as a .svg, loaded into Inkscape and printed like a PDF.

( to take to the printer on another PC )

Then I used transparent tape to stick it onto the reducer.


Edited by Mert, 12 October 2019 - 11:48 AM.

  • deepanshu29 likes this


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