On-chip Interweave Guiding

Dont have anything to contribute to the discussion on how this could be implemented, but will say that your 30s unguided estimate is off roughly by a factor of somehwere between 3-10.

The new harmonic mounts require guiding at short intervals, like a second or so so even further from the estimate. The 2 Skywatcher mounts i have used both needed rather short guide exposures + Predictive PEC in PHD2 to control the aggressive RA period. I dont see how any of the common mass produced mounts can go anywhere near 30s without major compromises to data quality.

Oskari Nikkinen:
Dont have anything to contribute to the discussion on how this could be implemented, but will say that your 30s unguided estimate is off roughly by a factor of somehwere between 3-10.

The new harmonic mounts require guiding at short intervals, like a second or so so even further from the estimate. The 2 Skywatcher mounts i have used both needed rather short guide exposures + Predictive PEC in PHD2 to control the aggressive RA period. I dont see how any of the common mass produced mounts can go anywhere near 30s without major compromises to data quality.

I have been running un-guided 30 sec. exposures on an old LXD-75 so surely, more modern mounts can do better? I would take 30 sec. as a minimum, at least with traditional mounts.

What’s the advantage of this and rational behind?

if your mount can do 30s or longer unguided and you want to do short exposure anyway there is no need for guiding. Just for re-centering?

how long a mount can go unguided depends primarily on the mounts price tag (unfortunately) but big time also on the resolution of course.

Die Launische Diva:
How many successful corrections an autoguiding system typically performs within those thirty seconds of exposure? If the answer is more than one then your proposal will only recenter the scope as you have already said. It is equivalent to registering the frames without having the ability to account for small field rotations. (My question is a honest one as I have never used an autoguider )

And the answer is none. Guiding or re-centering would only take place between light frame exposures.

Maybe I wasn't clear. Assume that you are using an existing autoguiding system and doing the astrophotography business as usual. How many guide corrections the autoguider completes during the exposure time of a single frame? If the answer is more than one, then the autoguider system will perform better than just taking an exposure unguided, recentering, and then taking the next exposure.

Oskari Nikkinen:
Dont have anything to contribute to the discussion on how this could be implemented, but will say that your 30s unguided estimate is off roughly by a factor of somehwere between 3-10.

The new harmonic mounts require guiding at short intervals, like a second or so so even further from the estimate. The 2 Skywatcher mounts i have used both needed rather short guide exposures + Predictive PEC in PHD2 to control the aggressive RA period. I dont see how any of the common mass produced mounts can go anywhere near 30s without major compromises to data quality.

I have been running un-guided 30 sec. exposures on an old LXD-75 so surely, more modern mounts can do better? I would take 30 sec. as a minimum, at least with traditional mounts.

*How do you determine how good it does? I would be very surprised to hear you get good HFR/fwhm stars at 30s unguided. Have you measured your unguided tracking error with PHD2 guide assistant?

If you were going to say because of round stars, then i will say that getting round stars is trivial and not the point of guiding. The point of guiding is to get a small round star. Big difference. You get a round star simply by having equal error in RA and DEC, so you could have 2'' of guide error in both axis and get a round star but of course the data is actually junk.

Alex:
What’s the advantage of this and rational behind?

if your mount can do 30s or longer unguided and you want to do short exposure anyway there is no need for guiding. Just for re-centering?

how long a mount can go unguided depends primarily on the mounts price tag (unfortunately) but big time also on the resolution of course.

If you are running un-guided, automatic re-centering would be very helpful because you wouldn't have to babysit the session. Many mounts do have trouble with traditional autoguiding so I see this as an alternative. The advantage is you have less hardware, less weight, less cabling and no differential flexure. The downside is it limits the length of your lights and reduces the time efficiency of the session. Modern CMOS cameras work great with shorter integration times because of low read noise although you will have more data to crunch. That said, storage capacity and computer speed is increasing all the time. I don't have a solution for session efficiency so it will take a bit longer to get any given imaging time.

Arun H:

I think we are discussing options and yes, from what I see on forums, some people do struggle with it for a number of reasons. I think I've already laid out the possible advantages, even when compared to traditional OAG.

I'll leave someone who has actually tried it to comment.

Perhaps there are advantages versus OAG. Certainly, someone who invents something new can always think of advantages. But there are disadvantages as well. Such as much less frequent corrections and hence more onus on the mount. Ideally, for highest performance, the mount would use an accurate model to account for flex and different movement rates in different parts of the sky. Such mounts can go unguided entirely, and there are quite a few imagers who do this. There would be greater overhead due to the need to solve and center. Etc. And to answer the question of if this will be better requires an understanding of why people struggle with conventional guiding. This is likely from a variety of reasons, including not understanding the basic principles of guiding, poor hardware, poor connections, bad calibration, etc.  People who have these issues are just as likely to struggle with some new guiding method.

in a way, with fewer corrections, it can be said that there's less onus on the mount and more on polar alignment accuracy. Yes, you still have whatever P.E. the mount has but you're  not asking it to bang around in both axis every few seconds, something that a lot of lower tier mounts with lots of backlash struggle with.

Quinn Groessl:
NINA can already auto recenter with the imaging camera. It's not as streamlined as you would like, but it can do it. Unless you can invent a way to have one sensor take simultaneous exposures of different lengths, this isn't really all that feasible.

How quickly can N.I.N.A. do that? If it has to plate solve to do it then it wouldn't really be useful. It seems that N.I.N.A. has all the bits internally to make this happen in a very streamlined way.

Quinn Groessl:
NINA can already auto recenter with the imaging camera. It's not as streamlined as you would like, but it can do it. Unless you can invent a way to have one sensor take simultaneous exposures of different lengths, this isn't really all that feasible.

How quickly can N.I.N.A. do that? If it has to plate solve to do it then it wouldn't really be useful. It seems that N.I.N.A. has all the bits internally to make this happen in a very streamlined way.

Correct, it platesolves, and then it's bound by how you set the tolerance and how accurately your mount slews.

I have been running un-guided 30 sec. exposures on an old LXD-75 so surely, more modern mounts can do better? I would take 30 sec. as a minimum, at least with traditional mounts.

Regarding what modern mounts can do; it is very well known that the harmonic mounts (at least the "cheaper ones") struggle a lot with unguided tracking and you will get elongated stars really fast.. You could probably do a few seconds with a telescope, but not much more than that. EQ mounts does this a lot better. I have no experience with mounts that have encoders though, they would probably tell a different tale.

My AM5 can't even do 10s unguided if memory serves me right, I haven't tried my Juwei 17 in the field yet but I don't expect it to do much better, neither do I see any reason for it as guiding is a fairly simple and cheap addition to a setup. In comparison I used to run my old HEQ5 for 60s subs with a 80mm refractor without guiding, but I also had to throw away quite a bit of data due to it. Shorter sub exposures also increases read noise, so I don't really see the need myself.

But that's just my take on it though. I get what you're trying to discuss, and though I'm not a programmer of any sort, I don't really see why it shouldn't be doable somehow.

Oskari Nikkinen:

Oskari Nikkinen:
Dont have anything to contribute to the discussion on how this could be implemented, but will say that your 30s unguided estimate is off roughly by a factor of somehwere between 3-10.

The new harmonic mounts require guiding at short intervals, like a second or so so even further from the estimate. The 2 Skywatcher mounts i have used both needed rather short guide exposures + Predictive PEC in PHD2 to control the aggressive RA period. I dont see how any of the common mass produced mounts can go anywhere near 30s without major compromises to data quality.

I have been running un-guided 30 sec. exposures on an old LXD-75 so surely, more modern mounts can do better? I would take 30 sec. as a minimum, at least with traditional mounts.

*How do you determine how good it does? I would be very surprised to hear you get good HFR/fwhm stars at 30s unguided. Have you measured your unguided tracking error with PHD2 guide assistant?

If you were going to say because of round stars, then i will say that getting round stars is trivial and not the point of guiding. The point of guiding is to get a small round star. Big difference. You get a round star simply by having equal error in RA and DEC, so you could have 2'' of guide error in both axis and get a round star but of course the data is actually junk.

I think it would be very difficult to get a consistent equal error and correction in both axis to get round stars as you've proposed.

Possible future cameras might have a per pixel shutter. So could be used for guiding at say 0.1 to 5 seconds and rest for much longer.