New Feature : Planetary Live Stacking

[Borodog]

The stacking is always continuous, unless you pause stacking with the new feature to do so.

[admin]

Hi,

changing the number of frames while stacking has the following effects

* Increasing the number of frames : SharpCap slows down the erosion of old data in the stack, so that as new frames continue to be added the effective length of the stack increases and the 'Current Stack Length' will gradually rise towards the new target

* Decreasing the number of frames : SharpCap increases the erosion rate of old data, so that even with new frames being added, old data is being removed more rapidly, and the current stack length will decrease towards the new target.

I'm not sure about the need for a separate histogram in the planetary live stacking - the adjustments are far less critical than in deep sky stacking (you are not applying a strong stretch, just tweaks to the black & white levels mostly).

If the focus drifts then the options are either to try to correct while still stacking (watch the frame filter graph and input frame contrast) for indications of improvements/worsening as you move the focuser, or you could stop stacking and return to focus score. If you want a long, uninterrupted timelapse then it's probably worth setting up temperature compensation on your focuser ASCOM driver (if it has that feature).

cheers,

Robin

PS. Will fix the stabilization option not being saved.

[DrBobAZ]

Hi Robin

Just checked out the Dec 4th version. Thanks for all the great feature upgrades. Especially liked the running graph of 'quality' and the ability to include the display histogram settings in the saved time lapse.

I had a bit of trouble with replacing feature tracking with the track planet with camera ROI option

When I do that, everything starts out fine, but very soon after initiating the ROI jumps (in one step) to a location somewhere else on the chip, and the planet of course stayed pretty much where it was so I lost the planet from the ROI.

I was able to reproduce this several times, so it just wasn't a one off case. I ended up just using the feature tracking (no mount guiding) as I did before.

Bob

[admin]

Hi Bob,

that's a shame, but not entirely unexpected - I tucked in that feature on the basis of the selection area tracking working and surely updating the camera ROI would be just the same conceptually. The cloud gods did not give me an opportunity to test for real, and it's a tricky one to test in simulation.

The code looks correct (no silly swapping of X and Y or anything like that), but I wonder if the problem is that the camera doesn't respond instantly to the change - that would result in the next few frames coming in with no movement, and SharpCap repeating the adjustment for each frame until it has gone far too far...

I will put in some logging and a speculative bit of code to prevent a second ROI movement for 1 second after any previous one - even if the clouds remaing until next week, we should learn more next time around.

cheers,

Robin

[DrBobAZ]

Thanks Robin!

Bob

[ChrisR Oz]

Hi Robin,

I've been processing old SER files of Saturn and Jupiter through the new Planetary Live Stacker using Test Camera2 and browsing for my SER files, getting used to controls, and it has been working very well. Looks like a great tool for live planetary viewing. Can't wait!

But suddenly, returning to the Saturn SER files that I used previously, I only get perhaps 1 frame stacked after several minutes. I can reset the stack, set to accept 100% and even disable the Wavelet and Display controls (expecting to see something like stacked original images) and get nothing for quite a while, then only 1 frame stacked out of about 8,000 frames. Checking "show unprocessed frames" shows the SER data frames OK. No preprocessor controls were added, just the source SER file.

Only 1 frame stacked out of thousands

Planet stack1.jpg (90.34 KiB) Viewed 116438 times

I tried restarting SharpCap (v 4.1.11433) many times, opened Test Camera 2 again, opened the Live Planet stack, and it continues to misbehave on Saturn, like there is a remembered bad parameter or it continues to test new frames against some exiting data (in GPU?), instead of starting afresh.

I've uploaded one Saturn SER here (not so pretty): https://drive.google.com/file/d/1Mk6bzE ... sp=sharing
Don't know what that would show, as it was working fine first time through before playing with Jupiter.

Cheers, Chris.

[admin]

Hi Chris,

the thing that comes to mind is to ask if you had adjusted the gain/exposure in the test camera when playing back the video? Apart from the quality filtering, the other main reason that it will not add frames to the stack is if the alignment detection code fails to find the center of the planet, which can happen if the image is very dark.

If you have applied a display stretch then that can make the image look reasonable, but in fact be dark (perhaps because you turned down the exposure in the test camera as well).

Certainly I have no problem getting the saturn SER file to stack correctly, so worth trying again with the default camera settings I guess. If you still have no luck, save the stack (even if only one frame) and include the capturesettings.txt for the saved stack, which will reveal all your stacking and camera settings.

cheers,

Robin

[PeterC65]

I finally managed to use the new Live Planetary Stacking feature last night. It’s amazing, a complete game changer for planetary EAA making it possible to do for the first time really.

Usually I just give the Planets a cursory look when I do EAA but last night I spent an hour observing Jupiter, coming back to it twice during the session and seeing the Great Red Spot for the first time.

I used the RAW8 colour space and a small ROI about four times the size of the planet, then set the planetary stacking tool going with a stack length of 1000 frames, just using the fine sharpening setting set to 1.000 and auto adjusting the brightness and colour. This is a fairly simple setup but gave great results on Jupiter.

Shorter exposure times seem to give a crisper view, I was using 6ms, and an IR pass filter improved things further, but at the expense of colour.
Here is the live stack from last night with a UV/IR cut filter …
[attachment=1]Jupiter Visible 20ms x D09_12_2023 T23_51_30.png[/attachment]

and another with an IR pass filter …
[attachment=0]Jupiter IR 5.3ms x D09_12_2023 T19_49_40.png[/attachment]

I was using an Explorer 200PDS at its native F5 as I was testing this new scope. Next time I will add a Barlow and try at a more appropriate (for my camera) F12.

[admin]

Hi,

just to let people know that there are some more tweaks to planetary live stacking in this week's SharpCap update (4.1.11491).

* Added options to rotate and flip the planetary live stack image. Note that rotating will happen around the planet center, so may not rotate the whole frame if you have a rectangular image

* Fixed a couple of settings not saving (note - tidied up the internal saving of planetary stacking settings, so some will reset to default with this update, but future changes to them will save and restore correctly)

* Hopefully fix issue with tracking using camera ROI

* Bin the image 4x4 before calculating image quality to reduce the impact of noise on the quality measure

cheers,

Robin

[Borodog]

Thanks for another great update, Robin!