Jocular: a tool for astronomical observing with a camera

[mintakax]

Bumping this up just in case

[Martin Meredith]

On 08/11/2022 at 18:10, mintakax said:

Martin, I have confirmed with Software Bisque that this BTP FITS Header entry specifies the East or West placement of the OTA during the image capture, with a value of  0 indicating OTA East (Pointing West) and 1 indicating OTA West (pointing East). I'm not sure if this is sufficient to act as a reference to  the sub orientation.  As far as my subs are concerned, if it is a "1", then it is always the orientation of pre-meridian flip, if it is "0" the subs orientation is rotated 180 degrees.

What do you think?

 

dan

 

Thanks for this info. I'll see if I can add an option to specify a certain flip on load if this header is present.

Martin

[mintakax]

On 13/11/2022 at 00:43, Martin Meredith said:

Thanks for this info. I'll see if I can add an option to specify a certain flip on load if this header is present.

Martin

Thanks Martin!

-Dan

[mintakax]

Is there any documentation that describes how to use the ROI function in Jocular? I can't find any mention of it in the documentation that I'm used to referencing.

[Martin Meredith]

11 hours ago, mintakax said:

Is there any documentation that describes how to use the ROI function in Jocular? I can't find any mention of it in the documentation that I'm used to referencing.

I may not have transferred all the doc across to readthedocs, but it is described in the older version of the docs here: https://transpy.eu.pythonanywhere.com/jocular/#

Essentially, ROI applies to the view you see in the 'eyepiece' at the point at which you click the ROI icon, and continues to apply until you click ROI again. The only thing to be careful with is that there are enough stars for stacking purposes, since ROI throws away anything not in the view. 

This is the relevant passage:

Jocular implements region-of-interest (ROI) as follows. First, use a framing capture and the pan/zoom/rotate options to get the object just how you want it. Then stop the framing capture and click on the ROI icon. Any subsequent captures (for the current observation) will be restricted to that region. Click ROI again to see the full sensor. Note that currently you need to set up any binning options *before* choosing the ROI.

 

 

[Steve in Boulder]

1 hour ago, Martin Meredith said:

I may not have transferred all the doc across to readthedocs, but it is described in the older version of the docs here: https://transpy.eu.pythonanywhere.com/jocular/#

Essentially, ROI applies to the view you see in the 'eyepiece' at the point at which you click the ROI icon, and continues to apply until you click ROI again. The only thing to be careful with is that there are enough stars for stacking purposes, since ROI throws away anything not in the view. 

This is the relevant passage:

Jocular implements region-of-interest (ROI) as follows. First, use a framing capture and the pan/zoom/rotate options to get the object just how you want it. Then stop the framing capture and click on the ROI icon. Any subsequent captures (for the current observation) will be restricted to that region. Click ROI again to see the full sensor. Note that currently you need to set up any binning options *before* choosing the ROI.

 

 

Martin, does this feature work with a "watched camera"?  I think Dan captures all his subs with TheSkyX and uses Jocular in watched folder mode.  His camera has an enormous sensor, so I think he's trying to keep memory usage down.  

[Martin Meredith]

9 hours ago, Steve in Boulder said:

Martin, does this feature work with a "watched camera"?  I think Dan captures all his subs with TheSkyX and uses Jocular in watched folder mode.  His camera has an enormous sensor, so I think he's trying to keep memory usage down.  

Ah, you're right, no it doesn't. ROI is implemented at the point of capture at present (in fact, I use the zwoasi function to set the ROI). It is possible in principle to do it with incoming subs but it will be important for consistency to do it prior to the main capture ie using framing subs, which doesn't necessarily fit with Dan's workflow. 

[mintakax]

Thank you Martin and Steve! 

--Dan

[Odiug]

I am trying to use Jocular with Kstars/EKOS. Images are FITS 16bit RAW from an ASI294mc. Fits header shows:

SIMPLE  =                    T / file does conform to FITS standard             
BITPIX  =                   16 / number of bits per data pixel                  
NAXIS   =                    2 / number of data axes                            
NAXIS1  =                 4144 / length of data axis 1                          
NAXIS2  =                 2822 / length of data axis 2                          
...
BAYERPAT= 'RGGB    '           / Bayer color pattern                            
...

In Jocular under "set up devices" "color space" I selected RGGB. In "channel chooser" I selected RGB. I still see a grey scale image. "filter(s) chooser shows "B", but I cannot change it.

What am I doing wrong?

Regards & CS

Guido

[Steve in Boulder]

6 hours ago, Odiug said:

I am trying to use Jocular with Kstars/EKOS. Images are FITS 16bit RAW from an ASI294mc. Fits header shows:

SIMPLE  =                    T / file does conform to FITS standard             
BITPIX  =                   16 / number of bits per data pixel                  
NAXIS   =                    2 / number of data axes                            
NAXIS1  =                 4144 / length of data axis 1                          
NAXIS2  =                 2822 / length of data axis 2                          
...
BAYERPAT= 'RGGB    '           / Bayer color pattern                            
...

In Jocular under "set up devices" "color space" I selected RGGB. In "channel chooser" I selected RGB. I still see a grey scale image. "filter(s) chooser shows "B", but I cannot change it.

What am I doing wrong?

Regards & CS

Guido

Hi Guido -- The first thing is to stack the images!  Your screen capture is a view of a single sub.  Click on the icon on the top that looks like a stack.  It's under the number of the sub.

[Odiug]

Ah, yes. Thanks Seve. I guess, I need to practice more. But I definitely need smaller subs.

[Steve in Boulder]

2 hours ago, Odiug said:

Ah, yes. Thanks Seve. I guess, I need to practice more. But I definitely need smaller subs.

With a 294 camera you may want to bin 2x2 or even 3x3, especially while you're learning the controls.  Also, you may get better performance by turning off continuous updating for the sliders.

I think performance enhancements for larger sensors are under consideration by Jocular management.  

[Martin Meredith]

I'll repeat my regular caveat that Jocular isn't really designed for OSC cameras. As you've probably gathered, the temporary solution currently implemented is to split each RGB into 3 subs so that it can go on to treat them as coming from separate filters and flip into LAB-space. I guess there are better ways of doing this (less memory/compute intensive) but I haven't had the time to investigate as yet. 

cheers

Martin

[bluebird181]

Hi Martin,

First post on this forum. I've tried unsuccessfully to get Jocular to work on my Windows 10 Pro PC. The latest version won't run at all. The earlier version you list on the website, along with the earlier GUI, starts but does not function properly.  The test files in the Prev folder don't open, for example. During the first start up, I inadvertently chose the wrong directory for the log file to go and now I cannot undo it. I tried uninstalling the program, rebooting, and reinstalling and the log file still goes to the unwanted directory. I would like to use "joculardata" as you suggested.  FYI, I am not a Python programmer or user.  I want to live stack LRGB and this looks like it'll work if I can get it running. I'd appreciate any help.

Thanks,

Larry

[mdrileynyc]

First post on this forum. I've tried installing and using jocular on Ubuntu 20.04. I was able to install after some fixes but it crashes when loading your suggested examples.

Installation issues (for others):

• I installed in a virtual environment
• Needed to apt install xclip xsel
• Needed to pip3 install olefile
• kivymd:
  • With default version, jocular won't start
  • With suggested version 0.104.2, starts but prints huge, continuous warnings
  • With pip minimum indicated version 1.0.2, starts without warnings

Usage issues:

• Downloaded examples, put in jocular capture directory and rebuilt observations
• Clicking on any DSO results in jocular exiting  - logs state:
  • 11Feb24 15:42:25.846 | ERROR    | jocular.jocular           | on_stop                    |  173 | root exception: rescale() got an unexpected keyword argument 'multichannel'
  • .../jocular/lib/python3.8/site-packages/skimage/_shared/utils.py\
  • ", line 316, in fixed_func
        return func(*args, **kwargs)
               │     │       └ {'anti_aliasing': True, 'mode': 'constant', 'multichannel': False}

Any suggestions how to fix the usage error?

[Martin Meredith]

There's been a recent-ish change to the skimage package, one of Jocular's dependencies. They've done away with the 'multichannel' argument. If you're happy to edit the code you can just search for 'multichannel' in 'multispectral.py' and remove that argument. Or send me a PM and I'll send you the updated code.

[mdrileynyc]

Thanks. There was another use in monochrome.py but removing both instances of the arg did the trick.

Lovely UI btw!

[Martin Meredith]

Yeah, I forgot about monochrome but glad you found it and have it working!