On-chip Interweave Guiding [Deep Sky] Acquisition techniques · Tony Gondola · ... · 55 · 1178 · 0

Jared Willson:

I'm not saying that people who are guiding are wrong because that's the vast majority of astrophotographers out there. In my case, I have been forced to shoot with a mount that's not capable of guiding with any accuracy and for a long time, couldn't replace it. I shoot unguided out of necessity and have learned how to make it work. I see nothing wrong with exploring alternatives.

What mount are you using that can handle 30s sub exposures unguided at 400mm focal length without any issues, but can not be effectively guided? Not trying to cast aspersions, but I am surprised by this particular combination. I would think that most mounts capable of solid tracking for thirty seconds at a time, even at a relatively short 400mm focal length, would also be capable of decent guiding performance. 

I'm not really a proponent of the, "Guiding Needs to Die" approach since I think guiding is actually a pretty elegant solution to the real problem of wanting to make tracking corrections in the middle of an exposure. But I certainly agree that for lots of imagers it's not a requirement, and it may soon go away entirely if read noise drops much from its current levels. I would love to get to the point that we are all just live stacking a few million 0.1s images!  For me, it wouldn't be about getting rid of guiding--it would be about incorporating "lucky imaging" into even faint deep sky subjects and improving resolution. If we could spin up an AWS server with the appropriate level of GPU power to do real-time image calibration, complete with error rejection, such that my "computer" could keep up with 10 frames per second and just give me a final stack, I think that would be really great. Then I'd need a really good internet connection to my telescope, but that's about it. No home NAS system, no terabyte size directories of raw files, and all the resolution planetary imagers are used to but applied to galaxies, clusters, and nebulae. We're not quite there yet. I know my observatory couldn't handle transferring 10 full resolution raw files per second to an S3 storage bucket, so right now I would have to pay for the latest and greatest NVidia has to offer stuffed into a desktop at the observatory, but there are already Python repositories out there that can keep up with live stacking 61 megapixel files at the rate of one every second or two as long as you can bring a high end graphics card to bear. Let read noise drop a bit more, and guiding will lose relevance.

It's an old Meade LXD-75, remember those? It is one step above the 55 because it has bearings! I won't say that its 30 sec tracking is solid, just something I can live with. On average, I have to throw out about 20% of the subs. The thing that saves me is my polar alignment is excellent. I am pretty much just dealing with RA.

Aaron Dalton:
If we're recentering on a scale small enough that individual subs won't be affected by tracking effects, there will presumably be not dithering motion which is a recipe for nasty artifacts when you go to stack. Hopefully that doesn't open a whole new can of worms.

I actually do dither, N.I.N.A. has on "on mount" option that will dither every so many frames without guiding.

Francesco Meschia:

Charles Hagen:
This simply isnt guiding, this is unguided with extra steps. That is the fundamental issue here. You are not doing anything to actually correct for potential error that will arise within an exposure, which is the entire point of guiding. Star alignment takes care of the registration of the frames in processing, as long as its close it does not matter in the slightest whether or not one subframe is aligned with the next, in fact we dither to avoid that happening at all. The solution is just to guide. The hardware required is not problematic, if this is an issue you are having I would think this is an edge case or a personal issue.

Additionally, if your mount can go 30+ seconds at a time unguided you shouldn't need to babysit it. If you do need to babysit it, I highly doubt it is actually able to perform well unguided for even 30 seconds. 

Just guide. Its not hard.

I say we go back to film ;)

Attacking people like @Charles Hagen who are offering comments with the argument that they’re against innovation is telling. Are you familiar with the Dunning-Kruger effect?

What Charles and I have been trying to convey is that if your equipment is good enough to produce good unguided subs for the sub duration that you want, you don’t actually need to guide at all. The interweave guiding system you proposed in your first post would re-center the frame between subs, before the next sub begins. That is actually not a problem of any relevance in astroimaging, because image registration (which is anyway mandatory before stacking) is already taking care of any drift. No one worries about that. It might actually be detrimental to the quality of the stacked master frame, because the same astronomical features would always fall on, more or less, the same pixels, making hot pixel rejection and fixed pattern noise removal very problematic. That is why we use dithering, which by itself does not mandate guiding.

I hope it doesn't seem that way, I'm not here to be negative and if it was taken that way, my apologies to Charles.

On the rest, what happens with my rig is that over time the RA error would build up to a point where I manually re-center because I don't want to loose too much of the frame during stacking. If the system re-centered for me I would be able to leave the system un-attended. I have since found out that all of the commands I really need are available in N.I.N.A.'s advanced sequencer so I'll be having a go at that.