Auto recenter in live-stack units (pixels currently)

[lowenthalm]

This feature is pretty cool, and seems to work well, at least when my plate solve is working reliably. One request though:

When I switch binning on my camera, I have to go in and change this up and down. Since it relies on plate solve, the image pixel scale is known, so could Sharpcap internally store the drift limit in arc seconds, so that it could automatically change the drive limit pixel count when you change binning? Its seems like a minor point, but in practice when I am using different binning on different targets during the evening it gets tiresome remembering to to change this each time.

[admin]

Hi,

I guess I could make the unit 'unbinned pixels' so that if you set 100 and are using bin2 then it would recenter when the offset came to 50 image pixels. Using arc seconds is actually awkward with the way it works at the moment, since I put off doing the plate solving until the first time recenter is needed, so the arc second measurement isn't known until then.

PS. Do make sure you are using an up-to-date version as there were issues with previous versions that would almost always run the offset twice in a row.

cheers,

Robin

[lowenthalm]

The unbinned pixel option would work. I had the feeling using arc seconds might be a a difficult approach, even if you did have the plate solve data.

I did notice that it would thrash if the plate solve failed, repeatedly trying after every frame captured and then solving failing again. If there is already a plate solve solution before you start live-stacking, why not just use that as a best guess at pixel scale for recentering?

I am using the March 6 4.1 beta release.

[admin]

Hi,

good point on the retry issue - I have seen that once myself but did not get around to doing anything about it.

I will have to investigate whether it can use pre-existing plate solving information. The issue is (I think) that any drift between the plate solve and starting stacking is undetectable if doing it that way and I think it matters...

cheers,

Robin

[lowenthalm]

If I recall the behavior when I was using it last week, you are using a goto, rather than an offset command. The original plate solve does give you a good solution for pixel scale and sky orientation, so could you do an offset only using this information if you its an EQ mount. Still better than no action at all!

It also strikes me that you could save the last good plate solved image in a buffer and go back and use that as your star alignment reference frame. You might even be able avoid having to repeatedly do plate solving to recenter the image, which slows things down and can fail.

[MikeHuerto]

Hi Robin
Just came across the auto - Recenter image using Platesolving option in SC 4.1 - fantastic tool!

But I am just curious about the speed of the operation to check for drift in pixels. Is Sharpcap doing a full platesolve, or is there some kind of image to image comparison that it is doing to check for drift? The reason I ask is that many GOTO Dobson users have very sort 4-15 sec exposures - so some kind of quick check routine would probably be called for.

Best
Mike

[admin]

Hi Mike,

the calculation for 'how far has the image moved' comes for free out of the live stacking - SharpCap is already calculating that info to align each new frame with the stack.

The first time a recenter is required, SharpCap has to do a plate solve to work out the scaling between screen pixels and sky degrees and also the orientation of the image onscreen. That might lead to a pause while the plate solve happens. Fortunately there is no need to re-do the plate solve on subsequent recenter operations in the same stack - the same data can be re-used.

cheers,

Robin

[MikeHuerto]

Wow! That's really clever - well done. I have yet to try it, but I think it will be a real plus to Dobson imagers - of which there are a growing number!
I have promoted it on the following Dobson Astrophotography site: Facebook site:https://www.facebook.com/groups/1940588322919476

I'm wondering whether it can be adapted for planetary re-centering. If I rig up a digital finder camera, aligned to my planetary camera - as I describe in this link - https://www.cloudynights.com/topic/8815 ... o-dobsons/

As I've posted in this forum previously, SW Dobson mounts do not respond well to pulse guides by SC Feature Tracking and Firecapture autoguide tools - this recentering tool, might be just the trick - I will give it a go!

[admin]

Hi Mike,

the problem with doing something similar for planetary imaging is that the camera settings that show the planet correctly will not show any stars for plate solving I suppose you could plate solve before starting planetary imaging, but even that is an issue with a Dob (or any AltAz mount) because field rotation means that the relative orientation of the camera axes and the mount axes change

That will require some sort of new idea to make it workable I think - not sure what yet...

cheers,

Robin

[MikeHuerto]

Would that still be an issue if a seperate autocenter and mount control camera with its own lense was piggybacked on the Dob? There would be 2 instances of Sharpcap running at the same time: One for the planet imaging camera looking down the focus tube of the Dob which would only be shooting videos. The other for the mount control camera which would be livestackingand autocentering the mount (but not saving images). Its sole function would be to keep the planet centered in its field of view. This second guide camera would be connected to the mount drive. It would require precise alignment of the guide camera and imaging camera- so that the planet would be centered in both cameras. Similar to the set up I described in the cloudynights link above.
https://www.cloudynights.com/topic/8815 ... o-dobsons/

Meanwhile, Here are two images from test run of the autocenter tool - worked quite well, until it could not longer align close to the zenith. I restacked the saved files in Pixinsight - it did a pretty good job of getting ride of field rotation artifacts.
Both are the entire FOV no cropping, with 988 x 4 second exposures. SC dropped about 30% of frames due to failure to align.

PS-Stack_988frames_3952s_WithDisplayStretch.jpg (173.09 KiB) Viewed 13166 times

STARSTARLESS-nx.jpg (446.61 KiB) Viewed 13166 times

I might retry with my fieldrotator attached.

But thanks again for developing this. Its a great tool with a lot of possibilities!

Best
Mike