New Feature : Planetary Live Stacking

[admin]

Hi,

just to keep things interesting, I have added a few new features to planetary live stacking with today's update (4.1.12925):

* A 'De-ring' slider. Using this may help get rid of (or reduce) the bright ring that can sometimes appear around planets as a result of sharpening. It will not help with any rings inside the edge of a planet or solar/lunar disk however

* The option to automatically create an animated GIF version of a timelapse when the timelapse is complete. This will create a GIF that plays over about 5 seconds if it is just playing 'one way' and a total of about 8 seconds if you select the forward/reverse option (4 seconds each way). These animated GIF versions are great for sharing on Social media without needing any further processing/editing.

* An option to additionally save an unsharpened copy of a timelapse, which could later be sharpened/processed consistently throughout.

Screenshot 2024-12-23 153907.jpg

Note that the animated GIF option is only available if you choose the main timelapse save format to be either SER or AVI.

Additionally, from this update, SharpCap can also read and play back AVI files via the 'Test Camera 2 (High Speed)' camera - this is limited to simple AVI files in RGB format (ie not compressed AVI files with MJPG, H264, etc compression). AVI files recorded by SharpCap should certainly play back OK.

cheers,

Robin

[Borodog]

Thanks for continuing development on this tool, Robin.

I am curious, what is the "de-rind" method you are employing? During development of waveSharp we ended up with a tool that allows the user to create a mask that reduces sharpening at the limb of the planet. The controls are fairly complex; a threshold value that defines the location of the limb, two separate controls that affect the feathering of the mask (one that feathers only inward toward the planet center and the other operating both inward and outward), a control for the shape of the feathering, and two controls for opening a gap in the mask and setting it's orientation (mostly for Mars and Venus, which often have terminators where the de-rind tool can damage the image by preventing sharpening where it shouldn't).

I am very interested to see how your single control de-rind approach works. I just installed the update and will test it with some old SER files.

[Borodog]

Hmm. Perhaps I am just missing it, but, where exactly is the de-rinr/de-rind slider?

Edit: Found it. It was hidden by the height of the bottom control panel.

[Borodog]

I like it, at least for Jupiter. Very similar results to what we achieved in waveSharp with just a single slider.

dering.png

Thanks for this.

By the way, I don't think I've mentioned this lately, but if the seeing is good enough, I feel like the results I can achieve live in SharpCap are approaching what I can produce with Autostakkert and post-processing. Your efforts on this tool have not been wasted. Profound thanks.

[admin]

Hi,

glad you got the new de-ring sorted and are getting good results. Actually it's dead simple - it sets a brightness threshold. If the original image brightness at a pixel is above the threshold then the sharpened pixel value is used. If it is below the threshold then the minimum of the original pixel value and the sharpened pixel value is used instead. That stops the sharpening from *increasing* the brightness of the image in the dark area around the planet, and therefore helps prevent the ring, but it allows the sharpening to *decrease* the brightness in that area to give a clean cut-off at the edge of the planet. So simple that maybe I shouldn't have explained it

Actually I've really been learning from using the timelapse how critical the seeing is - you can see the timelapse going from basically rubbish to rather good with nothing changing but the seeing conditions.

Another thing that I have snuck in that might be handy for planetary is a new mode for the planetary focus score controls. If you had selected to maybe average over 2s sample times, the previous options were to select either the average score of all of those frames or the best score - now you also have the option of measuring the 'average of the best 25% of frames'. I think this is beneficial, especially in variable seeing conditions, as the 'best' score could be higher than it should when out of focus simply due to one or two frames coming in very sharp due to seeing. Conversely the average score can be dragged down by the worst frames that you would be excluding from stacking anyway...

cheers, and Merry Christmas!

Robin

[Borodog]

FYI, I have recently switched from using the Contrast/Edge Detection method to the Histogram/Brightness method and the improvement in the parabolicity of the focus score curve is marked; it is much less sensitive to the seeing.

[Borodog]

PS, regarding de-ringing or de-rinding, the method I favor is a bit too complex for Cor to want to implement it in waveSharp. But essentially what I would really like to see is a method that performs a feathered Gaussian blur, or at least prevents sharpening, but only in the radial direction, centered between the last dark fringe and the outer bright fringe. That preserves the sharpening in the direction parallel to the sharp edge but reduces it perpendicular to the edge, which is the ringing direction. But your method is very simple and easy and produces a good enough result on Jupiter, and it's probably a good case for the KISS principle. I suspect it will not perform so well on Mars and Venus because of the terminator issue (although it should perform well on Mars near opposition), which is what we saw in waveSharp; hence the need for the controls to open and place a gap in the de-sharpening mask.

[admin]

Hi,

Mike - sorry I missed replying to your last post. While I expect that radial smoothing would help reduce the ringing effect, it would be a tricky thing to implement with any degree of performance. I also expect that you would need to manually pick the positions of the two inner/outer fringes, as automatic detection of that sort of subtlety is not going to be an easy thing to make reliable. I can see it being useful in some sort of image post-processing tool like Wavesharp or Siril etc, but it's not really suitable for SharpCap.

I have a new update uploaded today that has some fine tuning and tweaks to Planetary Live Stacking based on user feedback and some good Jupiter imaging sessions earlier in the week. To summarize the changes:

* Tweaks to how much the up/down buttons adjust the various sharpening and colour/brightness adjustments. Those buttons are supposed to make a change that is small enough that it is hard to see the effect, but using them 2 or 3 or 4 times will let you get the result you want. In some cases I have made the increment smaller to try to achieve this, in others I've made it larger where previously you'd have to press them many times to get any visible effect

* Fix to the ROI tracking not working if the camera flip feature was used

* Fix to the flip (after dark/flat) option not working when planetary live stacking was activated (actually it was being applied twice, so the image was flipped then flipped back to the original orientation)

* A history graph of the quality of the stack (prior to sharpening/enhancement) has been added in a new tab. I think this should record seeing quality over a session, but haven't had a chance to try on real data yet

* When creating an animated GIF version of a timelapse there is now just one tickbox - when enabled, both the forward only and forward/reverse GIF animations will be created. I found myself creating one then wondering what the other would look like with no easy way to create it.

* Under/over exposed warnings will show on the left in planetary live stacking if the input frames from the camera become under exposed (all channels have peak brightness at less than 50%) over overexposed (any channel has peak brightness above 95%). I have been wondering about adding an auto gain adjust to try allow the brightness to be auto-corrected in these cases...

* Fix a display stretch that was configured but disabled being applied wrongly to the frames written to the timelapse if the 'use display stretch' option was checked

* Add a reset button to clear the filter history graph - handy if you have made some changes to the camera settings or focus that affects the frame scores and want the graph data cleared rather than a step change causing all new frames to be used (step increase) or none (step decrease)

* Add an option to enable/disable the use of the selection area to specify the region of the full camera frame to stack. This is now off by default, but can be enabled.

I should emphasize that these sort of small but important tweaks often help make a big difference to the usability of the software - three really quite decent clear nights let me use pretty much every part of live planetary stacking and spot the things that were annoying me by not quite working right. I do value feedback from SharpCap users on this sort of thing, since I can't give this sort of detailed workout to every feature in the software regularly!

cheers,

Robin

[Borodog]

Robin,

I agree that the method I am describing is much too complex for SharpCap, or to be done automatically at all. It's more of a Photoshop type process.

Thanks for your continued work on this tool. If only I had the weather to make us of it any time soon . . .

[AstroMathmo]

Thanks for these tweaks - the default save region has helped me out.

I wondered if there were any cunning tricks to manage a MF while doing one of these. Clearly I'm looking through the least atmosphere around then. The ROI tracking is great at making sure I can capture right up to the flip as the sudden acceleration of the ROI movement is my clue to stop capture. Is there a cool way to minimise the join jump?