New Feature : Planetary Live Stacking

[admin]

Hi,

I was wondering about this last night too, since I could get 5 or 6 hours on Jupiter if I could run the timelapse through a meridian flip. For me the flip involves:

* Stop Planetary stacking
* Flip the mount
* Go to full resolution on the camera, 16 bit mode, 5-10s exposure
* Plate solve+ resync
* Re-goto jupiter
* back to 8 bit, ROI , short exposure
* Re-activate planetary stacking

Of course, the steps for re-finding Jupiter will vary depending on your setup, field of view, etc. If your mount is perfectly polar aligned with no cone error then you might even get it back in view after the flip, but somehow I doubt it.

I guess part of the solution would be to have a button to 'Freeze' the stack (for want of a better word) while you do the flip - a frozen stack would be retained even if planetary stacking is closed or the camera mode is changed, and would come back when you restarted planetary stacking, possibly with a message 'resume frozen stack?'. The code would also need to know to flip the input frames after the mount flip (not hard to do).

I also found myself thinking that an undo/redo for camera setting changes might help - you could change the settings to do plate solving then CTRL+Z a few times to undo the changes back to the original planetary settings.

I'll keep thinking about that one - I have some other changes all ready to go for Monday and a few more that I would like to do, so I'm not sure if this will find a slot in the near future.

cheers,

Robin

[ChrisR Oz]

Hi Robin,

It might be simpler than that. In many systems at long focal length, a plate solve is out anyway (unless you mean with a guide camera, ...). So the "freeze" feature might be all you need, plus the "image flip" after the meridian flip. Do the freeze. Then do the meridian flip. Re-find the planet, however you might approach that, and re-enable ROI centring (e.g. I use feature tracking ROI centre). Then with the planet centred, "resume from the freeze".

I don't think you'd ever get the meridian flip to succeed, find the planet and resume, all automatically, if long focal lengths are used. But your "Freeze" and "Resume from freeze" would permit the time-lapse to continue with little fuss, after finding the planet again.

Cheers, Chris.

[admin]

Hi Chris,

yes, that's pretty much how I was thinking of starting out - the freeze/resume would allow for the timelapse/stack to continue and give the user the chance to do the flip however they find best. I just about get plate solves to work at 3600mm FL, but I'm cheating and using an IMX571 camera for planetary, so I get a big FOV and then cut down to a small ROI.

cheers,

Robin

[admin]

Hi folks,

further improvements to Planetary Live Stacking in today's update (4.1.13010) - some really good stuff this week (at least I think so!)

Firstly the one I like the most - 'Auto Sharpen'...

autosharpen.jpg (41.02 KiB) Viewed 4706 times

There are two new buttons 'Auto Sharpen' and 'Auto Sharpen More' at the top of the sharpening controls section. Using one of these buttons will analyze the current stacked image to determine the level of detail at different pixel scales and then try to set the sharpening controls automatically to give a decent, sharpened image. The results can be seen below (unsharpened, auto sharpen, auto sharpen more).

The base 'Auto Sharpen' usually gives a result that is biased a little towards being slightly less sharp with less noise, where as the 'Auto Sharpen More' tends to give a result that is sharper but at the expense of more visible noise. Obviously you can adjust to taste from either to fine tune the image to your preferences. So far I have seen good results with a wide range of planetary images and lunar images too. Solar H-alpha - particularly with a fine pixel scale - seems not to sharpen enough (at least to my taste). If anyone finds any other cases that don't work so well, please let me know (and share enough data so I can reproduce the issue and test).

nosharpen.jpg (6.44 KiB) Viewed 4706 times

asresult.jpg (10.43 KiB) Viewed 4706 times

asmoreresult.jpg (14.42 KiB) Viewed 4706 times

Next, a slight re-organization of the main controls

The layout of the left-hand panel has changed and now looks like this:

maincontrols.jpg (33.42 KiB) Viewed 4706 times

The main goal here was to fit in an extra value - the 'Stack Contrast', which is a measure of how much contrast is in the stacked image before any sharpening/processing is applied to it. This value should be a good measure to indicate the recent quality of frames going into the stack, so will help indicate focus if you are still focusing and also show how good the seeing has been recently.

While some things have moved around to help accomodate the new readout, I've also added

• An info icon next to the stack length control - hover over this and it will show some information on the length of time that is being represented by the current stack state - should be helpful to ensure that the time is short enough not to cause problems due to the rotation of Jupiter showing
• A 'Half-and-Half checkbox next to the option to show unprocessed frames - this shows the processed image on one side and the unprocessed frame on the other side. If you uncheck and check it again, the two will swap sides.

halfandhalf.jpg (12.2 KiB) Viewed 4706 times

The Stack Quality History graph

This new tab appeared last week without much explanation, but I am planning to make a bit more of this in the future, so it's worth explaining...

stackqualityhist.jpg (29.33 KiB) Viewed 4706 times

This graph shows how the contrast level of the stacked image (interpreted as the quality) varies over time. If you leave all the alignment/stabilization/stack length settings alone then this graph is basically going to be responding to changes in seeing, transparency and any focus drift. A peak in the graph meas that the stack was a particularly high quality at that point, which in turn means that the frames leading up to that point in time were of higher quality than normal.

The value recorded in the graph is the same as the 'Stack Contrast' value shown on the left hand side - that value on the left is the instantanoue latest value of the graph.


Measuring Frame Quality

This week I have changed the details of the algorithm that is used to measure frame quality - this means that the numbers seen in the 3 contrast measurements on the left, in the Frame Filtering tab and in the Stack Quality History tab will no longer match the values shown for the same conditions in previous versions. Higher values still mean more contrast (usually better quality), but where you might have seen values from perhaps 5 to 7 before, now maybe they will be 12 to 17.

The goal of this change was to make the measurements less sensitive to changes in the frame brightness without reducing the sensitivity to focus/seeing quality. We already have a good tool to measure brightness (the histogram), so having the quality measurement jump up when the exposure is increased 10% wasn't really helpful. I ended up capturing Jupiter videos at a range of different peak brightness levels (50,60,70,80,90% on the histogram) - one set by adjusting gain and the other by adjusting exposure - then analyzing those files to be sure that new approach really is less sensitive to brightness changes... You can now expect to see a slight drop in the measurement values when the brightness is increased rather than a large increase.


Other Things...

• A bug which meant that values of sharpening sliders above 50 were not saved/restored properly is fixed
• Improvements to how exposure of the camera frames is measured - reduces false warnings of the camera frames being overexposed
• The maximum value of the sharpening layer sliders has been increased from 256 to 320 to allow more sharpening (particularly useful at fine scales)
• The sub-area of the image that is updated with new sharpening/processing settings while the sliders are being adjusted was incorrect if the flip options were used - this is now fixed
• There is now a checkbox on the timelapse tab that allows timestamps to be added to the timelapse frames

Hopefully that will keep everyone going for a week or so - I have more ideas to come

cheers,

Robin

[PeterC65]

Hi Robin,

Thanks for the latest updates. I am particularly looking forward to trying Auto Sharpen as I'm never sure where to set the Level 1 to 4 controls.

One change I would really like to see is to the file naming used for planetary live stacking. At the moment it doesn't take account of the user defined file naming templates used for DSO live stacking, indeed it seems completely different from even the DSO live stacking default (as far as I can remember). As a result I have to move files and rename them when I do planetary live stacking to keep the naming convention aligned with what I use for DSO's.

Is this something you could look at?

Peter

[admin]

Hi Peter,

yes, I got the feeling that a lot of people were struggling with the sharpening - too many adjustments and no obvious way to set them without experience. This approach seems to get a good degree of fine sharpness without the dreaded 'heavy crayon contrast' issues that you get when you turn up the higher layers too much.

The filenaming kind of bends the rules for normal file naming a bit, since the normal rules didn't seem to quite fit. The primary goal was to make sure that all the captures from a single stack end up in one directory - regardless of whether they are video timelapses or still snapshots - normal filenaming rules don't guarantee that. So, it begins by setting up a folder for the captures based on normal filename rules and then uses other bits of the rules to set the file names within that folder.

If there are bits of the filename of the initial folder name that aren't being set up right then I can look at that, similarly if there's something not right about the file names within the folder I can investigate, but it is critical to keep the 'one folder' design.

cheers,

Robin

[Borodog]

Thanks for continued updates on this tool Robin.

One thing I am looking for is a way to use Automatic Refocus that doesn't break the time lapse being generated by the tool. So the idea of a "freeze" and "unfreeze" of the time lapse would be applicable there as well. You might drop a time lapse frame or two during the refocus, but that is inevitable anyway and it would save you the hassle of concatenating multiple time lapses in post, e.g. with PIPP.

Cheers,

Mike

[PeterC65]

Hi Robin,

When live stacking DSOs I only ever save snapshots with adjustments using the template:
Live Stack (Processed): {TargetName}\{TargetName} {Filter} {SubExposure} x{Gain} {StackedFrames}frames D{Date} T{Time}.

The other templates I have set are:
Single File (Video): {TargetName}\Captures\{TargetName} {Filter} {Exposure} x{Gain} D{Date} T{Time}
Sequence (Stills): {TargetName}\Snapshots\{TargetName} {Filter} {Exposure} x{Gain} D{Date} T{Time} I{Index:2}
Live Stack: {TargetName}\Stack@{DateTime}\{TargetName} {Filter} {SubExposure} x{Gain} {StackedFrames}frames
Live Stack (Raw Frames): {TargetName}\Stack@{DateTime}\{TargetName} {Filter} {SubExposure} x{Gain} Frame{Index:4}

So using the M42 test camera I get a file named “M42 8.0s x12.5 2frames D13_01_2025 T17_13_18” in a folder named “M42”. The filter name is missing as my EFW isn’t installed.

As an aside, when I do DSO live stacking without the EFW I have to manually rename the filename to include the filter name (Visual or Clear) fitted, or not, directly to the camera. I assume I could avoid this by using the Manual Filter Wheel in SharpCap?

Using the Jupiter test camera for live planetary stacking with these templates I get a file named “Snapshot at 17_12_01 of Jupiter_00001__WithDisplayStretch” in a folder named “Snapshots” which is a subfolder of a folder named “Stack of Jupiter”.

This doesn’t seem to bear any relation to the templates I have set.

So far I seem to be able to rename the snapshots and time lapse files that I get from planetary live stacking using the same format as I use for DSOs.

Peter

[kaymann]

So I am in the middle of the Mars occultation and I cannot believe how easy it has been... Click auto sharpen -perfect. Click auto adjust brightness/ color - perfect for both Mars and the Moon. Stack count of fifty keeps Mars round and the not overexposed. Both save as buttons right where I want them... Wife and mother in law are in complete awe...

Thanks Robin for this ultimate tool. I have attempted and failed on other occultations and failed miserably so it is no doubt in my mind SharpCap is the key to tonight's success... Getting ready for time lapse of the reappearance can hardly contain myself!!

C9.25 with HyperStar and ZWO ASI533MC Pro

[admin]

So I am in the middle of the Mars occultation and I cannot believe how easy it has been... Click auto sharpen -perfect. Click auto adjust brightness/ color - perfect for both Mars and the Moon. Stack count of fifty keeps Mars round and the not overexposed. Both save as buttons right where I want them... Wife and mother in law are in complete awe...

Thanks Robin for this ultimate tool. I have attempted and failed on other occultations and failed miserably so it is no doubt in my mind SharpCap is the key to tonight's success... Getting ready for time lapse of the reappearance can hardly contain myself!!

C9.25 with HyperStar and ZWO ASI533MC Pro

That's the sort of report that I really like to receive! Looking forward to seeing some of the images too when you have them ready

@Peter,

I will have a dig into the file naming issue - one thing to note is that the 'Live Stack' template only applies to deep sky live stacking (when I wrote the settings page there only *was* one type of live stacking, so the distinction didn't need to be made). Planetary live stacking is following the 'Sequence of stills' template with some modifications for the snapshot images and the 'single video file' for the timelapses, or at least it should.

cheers,

Robin