Subsequent Jocularity

[AKB]

Somewhat hampered tonight (24-Mar-2019) by a combination of bad weather and ineptitude...

The weather issue was both increasing wind and cloud between 8:00 and 9:00 PM, until it became totally clouded out.  The ineptitude was due to my inability to get Jocular running in real time on my Mac, with the front-end acquisition on the observatory PC.  Managed to crash Jocularity by trying to select the shared drive. 

However, I subsequently perused, aligned, stacked, and enjoyed the results using this magnificent EAA application (thanks @Martin Meredith).  Having played recently with a cool (in both senses of the word) colour CMOS camera, it was a joy to rediscover the simplicity of a mono Ultrastar in combination with a C9.25 Hyperstar.  No calibration frames.  Also, no erudite objects observed, but a few old firm favourites: Horsehead, Flame, M78.    I will explore more exotic targets given the opportunity, and better luck next time.

Anyway... this was what I managed, which after some weeks of nothing was fine by me.

 

     

 

[Martin Meredith]

Nice work and glad to see the tool in use! 

I hope some of the issues you've been having will be sorted soon. OS-neutrality is very hard going but we'll get there I hope. Just to take the opportunity to remind anyone that the tool is available for testing for anyone wishing to do so.

I was out myself with the tool last night collecting RGB, L and H-alpha as I really want to make a start on supporting multispectral soon.

Martin

 

[AKB]

For the record, here are the post-processed images from the above session, redone in PI, with bias and flats, Mure denoise, careful stretching, and a bit of contrast enhancement. 

The full-sized frame shows some nasty reflections for the Horsehead and Flame... I'll blame Alnitak.

I look forward to using calibration frames in Jocular, when I understand quite how to do that (naming conventions??) as these make a BIG difference.

     

 

[Martin Meredith]

Calibration is not yet documented but since you mention it I'll say a few words.

Jocular maintains a calibration library and supports automatic deployment of calibration masters during the observing session. Darks for instance are tagged with exposure and temperature information, and get applied based on user-settable tolerances. Flat and biases are also applied. Flats will soon be tagged by filter too so they can be applied correctly in a multispectral setting.

Jocular also supports semi-automatic collection of calibration subs and the production of calibration masters. Until such time as Jocular has its own capture engine, it relies on a simple naming convention for the incoming subs. Any subs whose first 4 chars are 'flat', 'dark' or 'bias' are treated as the appropriate kind of calibration sub and accumulated just like any DSO, at the end of which a master is produced. All of this is automatic, but the user can intervene to select/reject subs, change the stack combination method etc. As soon as Jocular detects it is dealing with calibration subs it sets up appropriate values for the B/W points, stretch function and stack combination to make any user intervention easier (so for flats the stretch is linear, stack combination is 80% i.e. 20% outlier rejection, and the B and W are either side of 0.5). Minimally, all a user has to do is set up the capture engine with the appropriate name and Jocular will do the rest. Any masters created are available, alongside any created in the past, for immediate deployment. 

Since calibration subs (as well as the masters) are stored in separate folders just like the DSOs, they can be reloaded at any point in the future and reprocessed, including remastering, if required. Nothing is thrown away. Here's one I collected last night (I always use twilight flats and find them very effective so long as they are combined in the right way to get rid of star trails). I just reloaded the subs and did a snapshot of the stack. Now, if anyone can tell me what I need to do to my optical path to deal with the nonuniform illumination I'd be grateful! (Tilt somewhere?)

Getting masters created with other programs into Jocular is possible but currently requires editing a .json file, so is not ideal, although it simple and a one-time process in the sense that once done, that master is in the library and available for ever. Creating masters from subs collected with other programs is easy though. Just treat the subs like any other capture and reload them, whereupon a master will be produced.

The goal is to minimise user involvement in the process. It is very easy to collect a dark library almost without thinking about it over the course of a few sessions, and then one can just forget about calibration. During the session I keep an eye on the temperature and set it on the Jocular interface, and the system just selects the right masters based on that and exposure time given the aforementioned tolerances. If temperature changes during a session, the optimal darks are selected automatically. Having collected calibration subs over the past few months of usage I don't do anything at all during the session except flick a switch to say whether I want to use darks and flats, and like everything else that decision is reversible during the session, so if they are not behaving properly they can be de-applied quickly/automatically.

The more complicated issue is handling calibration of previous captures, but that's another story...

Martin

 

[AKB]

On 25 March 2019 at 17:52, Martin Meredith said:

Now, if anyone can tell me what I need to do to my optical path to deal with the nonuniform illumination I'd be grateful! (Tilt somewhere?)

I'd just assumed that this was just a gradient in the twighlight sky?

On 25 March 2019 at 17:52, Martin Meredith said:

Getting masters created with other programs into Jocular is possible but currently requires editing a .json file, so is not ideal

Not a problem, in my case, since, bizarrely, it is something I do all the time in another pastime (home automation software development.)

[Martin Meredith]

Thanks for the suggestion of it being twilight gradient. I wonder if there is any easy way I can check that? (optics are not my thing!). When I Iook back thru my older flats I see a similar pattern but that doesn't exclude twilight gradient I suppose since I tend to do the same thing each time. If it is twilight, that would imply that using these flats ought to produce poor results when calibrating images taken post-twilight, and I don't seem to see this, but maybe I haven't noticed or am cancelling it with gradient removal!

I was assuming something like tilt because I've noted (for years!) a (slight) issue with star shapes on one side of the chip with my optical path configuration but haven't got round to fixing it as the centre of the image plane generally looks ok and the autocollimator hasn't seemed too bad. It will soon be time to disassemble and reassemble the mirrors for cleaning so I will make a careful attempt to line things up and see if that makes a difference. I'm very happy with the Quattro for EAA and it would be great if the setup wasn't yet optimal because that would mean things will get even better...

Martin

[AKB]

10 minutes ago, Martin Meredith said:

I wonder if there is any easy way I can check that?

Rotate the whole OTA.  If it’s the sky, then the gradient will rotate, if not then it will remain stationary with respect to the frame.

Flats taken like this should work fine for removing dust bunnies and vignetting, but they may leave a residual gradient in the calibrated image.  This should be easily removed by appropriate processing.

If you really want to analyse the optical train, then CCD Inspector may be the way to go.

I love my Quattro, although I’ve had to do a lot of fettling, blacking edges, etc.  Stopping light getting in the back and around the focuser has been important.  I recently bought an auto-collimator to tweak even further.

Tony

[Martin Meredith]

Thanks Tony. I will do that before looking for more complex solutions.

I'd love to use CCDInspector but last time I looked it was Windows only and there are limits... but I like its functionality and I am tempted to try to reproduce parts of it myself at some point, given infinite time.

I've fettled my Quattro a lot too with flocking and preventing light ingress, though I need to do more around the focuser for sure now I'm hanging a reasonably heavy EFW off it rather than just the lightweight Lodestar.

Martin