initial test of an SX AO at f4 - calm conditions - windy conditions added

[Shiraz]

Peter (Poita) very generously let me use his SX AO unit to see if it would work OK for hi res imaging with an f4 Newtonian imager.

Initial testing, designed to study the effects of AO with minimal mount errors, yielded the following:
1. There is slight chromatic aberration from the 13mm corrector plate at f4, but it is not enough to worry about in normal imaging conditions.
2. When active, the AO generally reduced the peak brightness of stars (by up to about 30%) and increased their sizes (by up to about 25% in FWHM) when compared to results with the AO off. This was on a good mount (EQ8) in windless conditions and with average seeing of about 2.5 arc sec FWHM (see attached image for typical un-stretched results)

Discussion:
- An AO would be expected to increase star sizes by a factor of up to 1.4 outside of the isokinetic patch (eg with off axis guide camera). The AO did better than this, so it was presumably also correcting some mount error. However, the AO still degraded the imagery by a noticeable amount.
- The difference between AO and non-AO imagery is detectable at the sampling of 0.93 arcsec/pixel, but with the data resampled to 1.86 arcsec, there is no discernable difference with/without AO.

Opinion:
1. On a good mount under reasonably good seeing, with no wind and with near-Nyquist sampling, there was a perceptible overall drop in resolution, so AO does not appear to be beneficial under these circumstances.
2. However, if seeing is really high quality, it is possible that the beneficial effect of the AO in correcting mount errors may outweigh the loss in resolution. The 2.5 arcsec seeing was nowhere near good enough for this possibility to be tested, so this is conjecture.
3. If the mount is of lower quality and/or the wind is significant, the mount correction by an AO may provide a nett benefit in many other seeing conditions. This is also conjecture, but others have reported that this is the case.
4. If the imaging is undersampled (eg ~2 arcsec is widely used), the effect of an AO may only be noticeable if it corrects significant mount errors.

The next stage will be to test the system on an EQ6 and under windy conditions. For now though, the answers to two of the basic questions would seem to be:
"can an SX AO be used for hi res imaging at f4 ?" - yes
“does an AO increase resolution in average seeing ?” - probably not, likely depends on wind and mount quality, but you may lose resolution in calm conditions

Thanks for reading. Would be very interested to know if these results tally with the experience of others. And once again, thank you Peter for letting me use the AO. Regards ray

Test configuration: SX AO with an off axis Lodestar guide camera at 2x2 binning and 0.1 sec/frame under PHD2. 200f4 Newtonian scope, on an EQ8, with no coma corrector and with a QHY5 L2 as the imaging camera (at 10 second frames to minimise the mount error while still allowing the effects of AO to be assessed in isolation). Imaging sampling was 0.93 arcsec. Test sequences were 10 frames with AO active, followed immediately by 10 frames with it off.

[RickS]

Ray,

I did some experiments and got a reasonable FWHM improvement on a GSO RC10 at f/8 and an image scale of around 0.56 arcsec/pixel. IIRC, it was windy the night I was testing.

Martin Pugh is also getting some fantastic results from AO but that's on a big scope (17", I think?) at SRO with very good seeing.

Not surprisingly, it seems that AO is most beneficial when you're pushing the limits of your mount.

Cheers,
Rick.

[Shiraz]

Thanks Rick.

Seems to be a consistent pattern emerging - if the mount error is dominant (eg either the mount is lower quality, the seeing is exceptional or it is windy) then AO can be beneficial. If seeing is dominant, AO can make matters worse.

My test was designed so that seeing dominated and the limitations of off axis guiding, with uncorrelated noise in the guide channel, were evident. It showed quite clearly that AO as implemented can actually exacerbate seeing errors. Next stage will be to show what it can do to fix mount problems.

I wonder if we ever get good enough seeing in Aus for AO to be worthwhile on a good quality mount in still conditions?

regards Ray

[RickS]

Have you read this article, Ray? https://www.sbig.com/about-us/blog/differential/

[Poita]

I haven't used it much, but when I had a CGEM it was a massive improvement in image quality on that mount.

I think AO works like a champion with a big scope and an imperfect mount. I bought it more to fix tracking errors than for seeing issues.

[Shiraz]

Yes thanks Rick - Peter Ward pointed it out. It's worth a (very careful) read. Nice to see tech stuff peeking through the clouds of marketing hype. The main thing I took from it was that SBIG are aware that the currrent method of guiding on a single off axis star is not too good - hence their apparent intention to move to multiple stars, which will overcome the problem of noise injection into the guide channel.

Hi Peter. I agree that mount correction is where these things shine, not seeing correction, which is how they have been sold. Maybe there is a marketing problem - ie how do you sell something that fixes mount problems to people who believe that their mounts are perfect?:lol:

regards Ray

[multiweb]

Stars are noticeably rounder with AO on. What's the guide rate you managed you say? 10Hz? Must have been a very bright guide star...

[Shiraz]

yep, it was bright. The lodestar exposure was 0.1s, but the actual update rate was less than 10Hz. I was doing testing, so chose a bright guide star by just scanning around to find one with a high SNR and then imaged whatever happened to be in the field at the time. I imagine that finding a bright enough guidestar would be a major practical issue.

The unguided EQ8 was producing nice round stars, so the AO could not improve on them. Have you ever found that the stars can be rounder, but a bit bigger, with AO on?

[multiweb]

In practice I think the best guiding rates I've achieved were around the 4Hz mark. Good enough to correct for wind buffeting and mount errors. But I doubt it did any improvement with seeing. Although at long FL 3m and over at an image scale of say 0.7 second of arc per pixel your stars are definitely not as fat. So good for any typical SCT with a large aperture.

[Peter Ward]

Interesting result.

I'd question whether all AO's are created equal.

The groundwork to SBIG's devices was done by Brad Wallis (Awarded a NASA prize his programming skills during the Cassini mission) and Benoit Shillings, ( BeOS creator and Software engineer with few equals). SBIG went through a very thorough evaluation process to determine what benefits an AO might offer, some time prior to offering it to the amateur market. ( I recall being asked to provide data for this early assessment some 15 years ago)

If the system hysteresis, acceleration and dampening is not well understood, modelled and controlled, it would not surprise me in the slightest to see tip/tilt systems making images worse.
(BTW AO acceleration and displacement is user/software adjustable with SBIG's systems)

But this has not been my experience using AO7, AO8, AOL and AOX units. I have always found stellar footprints were as small and intense with AO enabled as without. When the seeing has been good, I have constantly seen 30-50% improvements in both. As has been pointed out earlier, one need only look at Martin Pugh's recent AO data and compare that to his earlier, non-AO, images to see the benefits.

[Octane]

Would an AO unit be of any practical significance imaging at 530-850mm focal length?

H

[Peter Ward]

I'd say no, but frankly have never tested such short focal lengths with an AO.

[Shiraz]

I rather think that you have completely missed the point Peter. The testing was done specifically to show that the concept of guiding on a star outside of the isokinetic patch can in some circumstances lead to negative results - AO's cannot be expected to "fix" sub-standard seeing because they suffer from this problem, although it is often claimed that they can. If I had used an SBIG AO I would expect the same result - unless they have somehow managed to bypass the physics of the atmosphere. As an SBIG distributor, you will naturally favour that product, but it seems a bit offside that you view this evaluation in terms of a shootout between competing AOs, when only one is being tested.

As was pointed out, if an AO is used under good seeing conditions, it should be able to fix residual mount/wind errors - I presume that is what Martin Pugh is finding, although I doubt that I will ever get seeing that is anything like good enough to confirm that idea.


Should be useful for correcting major mount errors (eg wind) in good seeing H. My testing was done at 800mm fl and the AO did what any AO would have done under the circumstances. The fl is immaterial (provided there is enough back focus) - what matters is the angular pixel scale, since the job of an AO is to modify the angles of the light rays. If you use small pixels at these focal lengths, AO should be advantageous under some circumstances - if you use big pixels and undersample, AO will be of less use.

[Shiraz]

Perhaps it would help to set out a mudmap of when (I think) single star guided AO probably makes sense and when it is counterproductive. The graphics will not win any awards, but show the general idea - it is a best guess distillation of the little that I know about how AO works and what others have noted. The shaded area has been tested as reported on in this thread. Other areas will be tested if possible when conditions allow, in an attempt to validate that which is currently based on anecdote and surmise. Maybe this is a bit of a Quixotic approach, but hey, this is a fun hobby, so why not.

The results of the current test should be read in the context of the attached diagram - this is not intended to be a blanket condemnation of AO, nor of the SX unit, just a demonstration of what happens under one of the sets of conditions in which AO is not beneficial.

[multiweb]

I like that Graph. Simple and concise. Bookmarked. :thumbsup:

[PRejto]

Disregarding a post I'm about to make re my problems with Protrack, I've been wondering if it makes any sense to consider a combination of Protrack and AO to avoid sending any guide corrections to the mount whatsoever?

Expectations are that Protrack on a Paramount can yield 5-10 min unguided images with perfect stars. Some users here report that they are able to do this but I have never succeeded. But I would guess that the small amount of egg shaped stars could easily (and possibly more effectively) be dealt with using AO.

How much drift can AO account for before it sends a mount guiding correction?

Thanks,

Peter

[multiweb]

I suspect a mount such as a PME with protrack would not benefit from any SX AO. But SBIG AO tilt mirror is far superior and might make a difference because of the update rate. It's chalk and cheese.

[PRejto]

Hi Marc,

I don't doubt what you said re SBIG vs SX AO. But, my idea wasn't necessarily to improve on any mount errors per se, but to avoid using the mount motors to provide guiding corrections. Wouldn't AO offer superior and much faster and smoother corrections regardless of brand? If the goal was to say avoid drifting a few pix over 10-15 min wouldn't AO have a better chance of doing this more smoothly than the mount?

Peter

[multiweb]

AO is another component in the imaging train interfering with the light path. In the case of the SX AO a 10mm thick piece of glass. Better straight light than some other surface to go through if you have a top mount and can help it. That's where I'm coming from. Less is more.

The AO will do very fast corrections but if you have a consistent drift you should sort that out first. When the star goes out of range of the AO the mount is bumped to re-center the system.

[Poita]

Happy to loan it to you Peter if you want to see if it will help with your setup.

[clive milne]

Ray there may be some benefit orienting the off axis guider such that it is pointing to the sky in the upstream direction (of the air flow) relative to the imaging chip. (typically west)


I could elaborate, but I am sure you can understand why this would be preferable.

[gregbradley]

I am surprised by these results. I thought the whole point of an AO unit was to help get tight stars in weak seeing. If it doesn't do that and adds some CA then that is pretty pathetic.

Martin Pugh is getting amazing results from an SBIG AOX unit at Sierra Remote. Peter Ward's recent Eagle shows AO benefits even at moderate focal length.

So is this more the SX unit is not very good?

Greg.

[Shiraz]

Hi Clive. Now that is lateral thinking:thumbsup: - will try it if I have time.


Hi Greg.

all single element refractive AO devices add CA. However, it appears that it is not a major problem even at f4 and with the relatively thick plate of the SX, so it is not an issue for your system.

As far as I can tell, the main benefit of an AO is to give even better results in good seeing by tidying up residual mount errors - surely that is not too "pathetic". Rick hit the nail on the head - AO helps most when you are pushing against the mount limits (eg the seeing is <1.5arcsec and you are imaging at much less than an arcsec/pixel).

The penalty of sampling the guide signal from a star that is a long way from the image means that the total seeing error increases. All amateur AO systems do this - even the SBIG paper referenced earlier shows that to be the case. So current AO does not really fix poor seeing - because the geometry is wrong and the update rates are too slow. "Fixes the seeing" still seems to be used for some marketing purposes though. At least SX only claim "Overcomes rapid gear errors to stabilise even difficult mounts".

AO can tidy up mount errors due to lower build quality or wind induced motion, even in poor seeing - lots of evidence that it can provide a nett gain, even with the inherent reduction in resolution.

I would certainly not say that the SX AO is "not very good" - it works as expected. I have no idea how it compares to the SBIG alternative.

As an aside, the errors introduced by using a single off axis guide star are apparently well understood in SBIG. According to the paper referenced by Rick, they are working on a way to use multiple guide stars and an on-axis guide scope, which will get around the problem of decreased resolution - but I don't think they are there yet. If they can pull it off, they probably won't be able to improve resolution by much in poor seeing, but at least they will not make it worse.

[Peter Ward]

I remain puzzled as to why you are seeing better performance with the SX AO turned off, as opposed to it being active. My intention was not to bag SX, but your data does not reflect the results Alan Holmes has published, nor my own qualitative results or that seen by many other (SBIG) AO users.

Alan's data shows a clear improvement & correlation, in both Dec and RA that slowly decreases with angular distance from the guide star ( if only mount errors were being corrected, that would not expect a Dec correlation).

I think it incorrect to say, the physics predicts a first order AO will make things worse.

Quite the contrary.

If the atmosphere's isoplanic patches on the night are at least as large as the telescope aperture, tip-tilt or first-order corrections, provided they are fast enough, will be beneficial....this is exactly how professional systems work prior to applying complex wavefront corrections for higher order components via deformable optics.

On nights when higher order turbulence dominates the image, sure, I'd expect zero improvement...but imaging during tragic seeing, regardless of how good your optics/mount/etc. are, is a futile exercise IMHO.

Suggesting residual mount errors/wind buffeting is all that is being corrected by the AO doesn't offer much insight ( with a system similar to mine )as my mount's PE is around 1-2 arc sec and, being in a dome, there is zero buffeting. Begging the question: if not isoplanic waves, what is being corrected ? That said, I am now curious.

It will be interesting to gather some data, with and without high-speed first-order corrections ( in my case being applied by an AO-8 and AO-X) and similarly quantify the results. I'll happily upload the results when done.

[gregbradley]

I would be interested in that Peter. There are really only 2 AO systems on the market SBIG and SX. I was always somewhat suspicious of the SX unit seeing as the company is more of a middle ground type product maker, but willing to have it be otherwise.

I have often noticed Leo from Italy's images are very sharp with tiny stars and he uses an SX AO unit but on a GSO RC with relatively long focal length.

The SBIG units have many great examples of tight images.

By the way the term is Isoplanetic (also spelt Isoplanatic) Patch: which means the angular section of atmosphere where electromagnetic waves are parallel.

I think Ray referred to it as Isokinetic Patch which is another term for the same thing.
Greg.

[Peter Ward]

Isoplanatic

:lol: quite correct! I typed that while offline and my ipad autocorrect didn't like either

[Shiraz]

Maybe Alan's own words can convince you: "Figure Six shows the star image that would result if the errors were only due to stellar motion. If I didn’t guide at all, but had a perfect drive, I would obtain the result given by the "Slow Guide curve. The higher the peak value, the better the star image. With Two Star guiding, the result actually got a little worse. This is what I mentioned might happen before – the guide system is chasing the seeing, but one frame behind. " ie Alan's published results match very well with what I observed - the star images are "a little worse" with an AO than without. And they were taken in similar conditions of: "These measurements were made on a night with only a slight wind." Alan then goes on show how reducing the framerate can actually improve performance and then to explain how much better multi-star guiding would be, but that discussion has nothing to do with current systems.

As for mount errors, Alan's analysis method specifically excludes them (he uses differential measures). I excluded them by measuring on a windless night and with short image exposures (10sec) on a pretty good mount. And we came up with the same answer - AO (of any brand) is quite capable of reducing resolution. If you accept this, then the next question is "well why do they help at all?" - what you get probably depends on the spectrum of the seeing error, but I suspect that an AO can tidy up a myriad of mount wobbles, producing a nett benefit only when the improvement from the tidy up is more significant than the increase in seeing noise (ie the seeing must be good to start with). This is pure conjecture though and the complete answer is probably nothing like that simple.

Agree that pro systems must use tip-tilt in conjunction with deformable optics, but I am pretty sure that they would never even dream of trying it with a guide star outside of the isokinetic patch and at bandwidths of 10Hz.


In any case, would really look forward to seeing some hi res FWHM measurements :thumbsup:. I also hope to get some more measurements soon in conditions that are more suited to showing that improvements are possible with AO - not just degradation. In the end, a theory is only of value if it is backed up by measurements - however, if we can get a clearer understanding of what exactly is going on, it may be possible to push the limits a bit harder. In particular, it may be possible to choose an optimum update rate, since faster is not always better.

On isokinetic and isoplanatic patch, I think that they are generally regarded as being slightly different: the isoplanatic patch is that region over which all turbulent effects are correlated. The isokinetic patch is that over which only the low order tip-tilt stuff is correlated - it is bigger than the isoplanatic patch.

Ref: https://www.sbig.com/about-us/blog/differential/

[Shiraz]

Last night was windy with over 20knots of gusting sea breeze. Set up the AO system (SX AO on f4/200 Newtonian) in the open on an EQ6, a situation that would normally rule out any imaging due to the uncontrolled wobbling of the mount with wind loading.

In the worst conditions of the night, neither AO nor unguided imaging was possible – there was just too much movement. However, after a few hours, the wind abated slightly and the AO could manage the jitter - results at:
http://www.astrobin.com/full/113137/0/?real=&mod=
The top row of 10 heavily cropped sub images shows what happened unguided in these conditions – I would only class one of the subs as usable. The lower row of subs shows how well the AO corrected mount movement – over half of the subs were usable. The AO did a good job, since the better quality subs had FWHM of around 2.5 arc sec and reasonable star roundness – I would regard these as high enough quality to be useful for either colour or luminance. Very bright stars showed some smearing caused by uncontrolled motion (before the AO reacted), but these images could still be usable with median stacking.

As an additional test, I tried everything with the RCC1 coma corrector in place – there is just enough back focus to fit the AO, a thin OAG and a thin filter wheel. There did not appear to be any hidden "gotcha"s.

Although the scope is an f4 with 800mm fl, it is being used for high resolution imaging – the small pixels yield a system with sampling of 0.93 arcsec, which is about equivalent to a 2m fl scope with 9micron pixels. From this perspective, the AO did a good job in windy conditions and the results validate the suggestion in the earlier diagram that an AO can be useful in wind. The other suggestions that AO works well with a lower capability mount and also in very good seeing, and that it is not worthwhile in poor seeing, are all borne out by the observations of other users.

And finally, thanks again to Poita(Peter) for letting me use the AO.
Thanks for reading. Regards Ray

[alistairsam]

hi Ray,

Interesting results, I have the older Orion Steadystar AO where the OAG prism can actually be tilted to point to a different area in the FOV, would this help at all if tilted to point toward the centre?

Secondly, do you think using the lodestar in 2x2 binning had any effect on your results? the lodestar has fairly large pixels at 8.2micron. when you bin, don't you lower the resolution of the guiding component and would this affect centroid calculations especially if it is a bright star?

I had a brief chance to test the AO on my 10inchF4 with the stf8300 and it showed very promising results, this was with an NEQ6pro, and not very windy conditions. I normally use an OAG.
did you do any tests with 1x1 bin of the lodestar?

I got a lodestar X2 and had no issues with getting a guidestar at 100ms from a very LP backyard.
cloudy as hell, but I'll post some results as well when it clears up.

Looks like the benefits are more evident in a system with a higher resolution, typically less than 0.7 arcsec/pixel.

Cheers
Alistair

[Shiraz]

Hi Alistair.

Don't know if angling the prism will move the aimpoint much - have you tried it?

I tried the Lodestar at 1x1 and 2x2. The only perceptible difference was that the update was slower at 1x1, so I used 2x2 (as recommended by SX) for all testing. The centroid finder in phd is apparently able to resolve to much better than a pixel (below 1/10 pixel with good SNR). I used stars with >30 SNR, so expect that the centroids were accurate to a small fraction of an arcsec.

My impression is that AO has most benefit when using a mid quality mount (eg EQ6) with a fairly heavy load under most seeing conditions - and when using a high quality system under very good seeing conditions. Agree that pixel scale is important - there is probably not much point in improving the guiding by 1/2 an arcsec if you are imaging with a pixel scale of say 2 arcsec - you won't be able to see the difference.

What sort of improvement did you see when using AO on your system - better resolution or rounder stars?

Regards Ray

[Poita]

I don't think any of the Active Optics systems (SBIG or SX) out there truly correct for poor seeing, true Adaptive Optics systems that deform the mirror to cope with poor seeing do a great job, but we won't be seeing those in amatuer astronomy any time soon.

What they can do is correct way faster, with better accuracy and with less momentum shift than mount-motor corrections. This makes them very useful for slightly windy conditions when you couldn't otherwise image, and for where your mount is at the limits of its capability. Both situations are common in our hobby, and AO gives a real benefit.

With crappy seeing however, I just go back inside and catch up on IIS :)

AO works best when seeing is good, and when seeing is good, that is when you really want to be imaging, so it is a win-win to me.

[marc4darkskies]

Interesting thread - thanks Ray. I won't dispute your findings because I've never done any scientific-like testing!! However, I have used an AO-L quite a bit (only at F11.7 on my Tak) and never experienced a degradation unless the AO guide parameters weren't correctly tuned (aggressiveness and, to a lesser extent, slew rate).

I set aggressiveness depending on guide star brightness (guide frequency) and seeing. The higher the guide frequency, the lower the aggressiveness and I moderate this based on seeing. The point being I never try to correct any one guide error with a single correction because that likely means you're chasing seeing (regardless of how good that might be). The objective is to minimize "wander". Wander is a measurement that indicates the amount of position error that has not been corrected (and it can never consistently be zero). Using TheSkyX and an AO-L it's possible to monitor Wander and adjust the AO settings to ensure wander is minimised and also diminished from the values observed without the AO switched on. The intent is to ensure there is at least some benefit to switching the AO on.

Slew rate just limits the maximum correction that the AO is allowed to apply (arcsecs/sec). The maximum correction is the slew rate divided by the guide frequency (Hz). Ie a setting of 10 at 10Hz limits any correction to less than 1 arcsec. This will help avoid over correction when the seeing is bad.

In your case, I'd be interested in seeing your guide error graphs with and without the AO.

Cheers, Marcus

[Shiraz]

Thanks Marcus - great post and a breath of fresh air.

I had been concerned that AO is being sold as the panacea that can "fix the seeing" - something it certainly does on pro systems - but I could not see how it could do that in the way we use it.

Amateur AO seems to be a bit like putting a generic shock absorber on your car - it may help a bit but it can also be worse than nothing. My test in calm seeing was with default settings for the AO and was simply to show that AO certainly can make things worse in some circumstances by introducing uncorrelated seeing noise into the tracking loop. It was never intended as a blanket bucketing of AO, although it seems to have been interpreted that way.

You provide a sound strategy for optimising AO parameters to get a balance between fixing the slow (mount?) errors without introducing excessive seeing noise. My attempts to optimise the AO were not as systematic, being limited to trying different update rates and setting the minimum error at which correction was attempted. I did not keep the error logs - did not realise that they would interest anyone.

Your strategy seems to suggest that you view seeing as the enemy and you attempt to keep it from doing too much damage while still getting some guiding gain. That raises the question of what the AO may actually be correcting for, if it isn't the seeing. Do you have any thoughts on that? Could it just be mount wobbles or could it be dome seeing or tube current effects?

regards ray

[Shiraz]

Good summary Peter, should have just asked you in the first place :)
regards Ray

[marc4darkskies]

Well, I'm an empiricist :). I know (because I've watched it) that seeing contains both high (>10Hz) and low frequency (around 1Hz and longer?) components.

If I have a bright guide star that allows 10Hz corrections then a properly tuned & calibrated AO unit will reduce the amplitude of seeing induced guide errors of less than 10Hz. I know it won't remove the errors (especially the near 10Hz components) because the guiding is reactive, not predictive, but it has to attenuate them. You can monitor this to some extent by watching wander and guide error graphs. The limitation with this monitoring is the sampling frequency of these graphs.

If I can only get 1Hz corrections (a fainter guide star) then at least I know that the low frequency seeing components will be attenuated and I don't have to rely on mount corrections for any lower frequency wander. There's no way an ME can correct faster than about 0.5Hz anyway and even that is pushing it with a heavy scope and probably meaningless from an imaging perspective since most seeing is much faster than this.

The only way to verify an AO's performance (which I don't feel compelled to do because I've seen it working :D) is to compare residual error graphs for multiple runs with and without 10hz guiding. If a guide error graph has increased amplitude after turning on an AO then it's likely the AO was not properly tuned.

Cheers, Marcus

PS: Naturally, seeing is not just about guide star displacement (ie residual error or wander). Bad seeing is usually accompanied by "boiling" in which case, I always switch off, pack up & go to bed! :) AO will only reduce wander.

PPS: I make no claim about the source of the wander I'm trying to minimise - there's no way for me to know that - although at frequencies greater than 1Hz it's unlikely to be from the PME.

[gregbradley]

Now I am wondering even more if this is really the difference between a Starlight Express AO unit and an SBIG one. From what little I gather about the SX unit it seems 10hz corrections are unlikely although they do have the new Lodestar which is 2X more sensitive than the old one which was about the most sensitive guide camera around.

Greg.

[alistairsam]

I have the Lodestar X2 and its a lot more sensitive than the lodestar which I had previously.
with the X2 and my Orion AO, I get atleast 5 stars with good SNR almost anywhere in the sky with 100 to 250ms exposures. with bin 2x2, SNR improves. but its very clean and hassle free.

Here's a quick 3min test of my Orion AO, at 3Hz. Stars are elongated due to bad collimation.
https://www.dropbox.com/s/6pyiuhiq4u1ya32/M42-AO_Ha-stack1-fnl1.png?dl=0
and this was without a single mount correction.

my OTA has a lot of flexure so the AO certainly helped reducing the size of the stars and dropped my fwhm, primary benefit that I can see are improvements of an average mount and bad OTA.
Hoping to do some more tests and exposures this week.

the Orion has a Maxim plugin so everything is done from within Maxim which helps a lot.

Cheers
Alistair

[Shiraz]

Hi Greg.

I only did two tests of the SX AO, so I have no basis at all to compare it with an SBIG one. However, it worked well and behaved exactly as expected under two conditions: it made the stars slightly worse in short exposures, on a good mount in still conditions: it did a good job of cleaning up stars in bad wind. Others have found that AO is generally worth having in more usual conditions, particularly with lesser quality mounts, so I did not test under those conditions.

I used the SX with manufacturer-recommended settings (with minor tweaks). Marcus has pointed out that some software provides mechanisms for optimising AO performance to match the mount and conditions, but I did not have any such software. It may well be that the SX AO could have performed better with alternative software.

I ran the Lodestar at 0.1 sec exposure, but the actual update rate (of any AO) was less than 10Hz since it depends on a wide range of factors:
guide exposure +
USB download+
centroid and offset calculation+
serial comms to AO+
step AO to demanded position (variable depending on offset)+
serial handshake to PC+
begin next guide exposure

Plus there is unpredictable Win8 scheduling before it does anything new.





Thanks Marcus. Your empirical approach of "use it because it works" is eminently sensible. However, I have a different approach - I want to know what it is actually doing, not for academic interest, but because that may give some insight into how to improve the system in other ways, as well as (or even in place of) AO. Amateurs use AO in ways that appear to be in conflict with the way that the pro's use them - I want to know why they work as well as they seem to and where the limits are.


Hi Alistair. Good results and look forward to seeing more. Looks like AO is definitely worth doing with your system and conditions. Looks like Maxim helps with optimising the AO - what procedure do you use?

Regards Ray

[bert]

Just for the record I have used an Sbig ao @32hz on a core 2 duo machine... Which is hardly a powerhouse.

Currently I'm using an aol with an onag xt at f11, guided only in ir. That seems to work pretty well. I have read anecdotal evidence that guiding in ir may have some benefits... But still tuning.

[marc4darkskies]

I actually don't understand what you mean Ray. I'm correcting two dimensional guide star position error by tip/tilting a piece of glass in two dimensions - that's what it's doing. I'm not trying to be flippant here. It's simple feedback control, essentially, to cancel position noise. There is no point trying to understand anything else IMO and certainly no "need" to understand the physics of seeing - that's for the pros who are deforming huge mirrors to compensate for wavefront error. If you accept seeing can cause say low frequency (1Hz) to high frequency (>10Hz) guide star position errors then cancelling out some of that error with simple feedback control will be increasingly effective as the tip/tilt frequency increases.



Pros have large deformable segmented mirrors and while I think they do use tip/tilt on mirror segments (??) for coarse position error corrections, the really cool AO happens because they can analyse the wave front distortions (generally using laser generated reference stars) and correct for them with very clever computation and fast actuators. Pros may also be doing feed forward control.

Amateur AO is limited to correcting position errors only and then only at a relatively low frequency. The effectiveness is completely dependent on the frequency of correction (guide star brightness) and tuning. The main assumption is that the guide star is experiencing approximately the same positional abberations as the rest of the field - a pretty safe assumption given the apertures we use.

Cheers, Marcus

[Shiraz]

Thanks for the heads up Brett.

As I understand it, the old model 7 was pretty fast, but the newer stepper designs can be a lot slower and update rate can vary with how much correction is required. The AOL manual states that the max update for AOL with a 4 pixel offset is about 7 Hz (driven by software overhead, imaging time and time to move the corrector). Does that sound reasonable? The SX (also stepper) with 100ms frame exposures yielded around 3-4 Hz, depending on the required correction, but I have not compared the SX side by side with anything else, so please don't take these figures as comparing apples to apples..

IR ONAG seems to me to be a very good idea for AO, but I don't think it is practical at f4 - there just isn't room.



quite a few issues Marcus, so another long post - sorry.

I did a limited assessment of the performance of an SX AO on an f4 scope, set up as recommended in the SX manual. It was established that CA was not a major issue at f4, that I could fit in the AO + guider + filter wheel, that the AO worked well in wind and that it produced slightly worse results in very calm conditions with average seeing.

I was not surprised that the AO degraded the results in some conditions - at f4, the guide star is well over a degree away from the centre of the field of view, so it seemed quite reasonable that uncorrelated seeing noise in the guide channel would cause problems. At the time my thought was that the situation could possibly be improved with tuning, but the SX documentation provided no info on how/if to do so. So the fact remained that I had set up the AO pretty much as suggested by the maker and it resulted in worse performance in one particular test environment - I posted the result.

You and Peter indicated that you always get some benefit from AO. My understanding is that you tune the filtering of the guide data to strongly suppress the fast fluctuations and guide on the remaining low frequency stuff. You stated that you do not know what causes this low frequency "wander". Some of it could possibly be due to atmospheric turbulence, but probably not much (eg from the SBIG AOL manual "Based on our experience and knowledge, AO-L and AO-7 rates help only slightly with reducing ground layer atmospheric turbulence, but are quite effective in reducing guide errors and wind buffeting"). Apart from guide errors and wind, some wander may be due to dome seeing, minor mount/hardware vibrations and tube/mirror seeing. I am very interested to know which of the possibilities dominates and under what conditions, because some issues can be managed more effectively in other ways. For example, dome seeing can be cured by cooling the dome interior ( FWHM have been ~ halved in this way by some pro observatories). Mirror seeing may be fixed by fitting front fans to break up the boundary layer. I hope to be able to isolate and measure the effects that are influencing resolution and then find ways to deal with them up front. Then, if it still makes sense, I will fit an AO to reduce what is left - at least that's the plan. Should be fun :)

As an aside, I have no interest in professional systems per se, it's just that the pros take the time to test and document what they do - we amateurs tend to rely on impressions. The pro systems that seem to best relate to our systems may be "ground layer AO" - there seems to be a fair bit of activity in that area and they face the same issues we do. The big differences are that they use multiple guide stars, that their guide stars are much closer to the fofv (or in it) and the bandwidths tend to be higher (all of which which sounds like a blueprint for SBIG AO R&D). Many use deformable mirrors, but some are based on simple tip/tilt like we use and these are usually applied for correcting tracking errors (eg due to wind or mechanical components).

Regards Ray

[marc4darkskies]

No worries at all Ray - good luck in your quest mate! :thumbsup: Look forward to seeing your experimental results! :)

[Poita]

If anyone else would like to borrow this to have a play, just send me a PM.
I won't be able to use it for the next 2 months, so am happy for someone else to get some time in with the AO unit. I always learn something new when people borrow gear, so it is a win-win.

[Paul Haese]

Ok I have a question or two.

If my guiding is now at 0.2 of a pixel on a system that is at 0.78" per pixel, will an AO produce sufficient improvement to warrant the cost?

Now if I have good to great seeing, and my mount can only guide at 0.5hz (post above) will an AO be worth getting?

I get what an AO essentially does, but I just want to clarify things in my mind.

[Shiraz]

As I understand it, an AO should help your system in good seeing, because it should reduce local slow turbulence effects (dome/tube) and guide/wind errors. If the seeing is not so good, seeing blur will dominate over wind/guide errors and an AO may be less use (ie it won't fix bad seeing). If you are guiding at 0.16 arc sec (assume RMS), the mount induced FWHM component is about 0.4 arcsec, which is probably just enough to be noticeable in really good seeing.

However, people who regularly use AO indicate that it always helps, so it could be worth it under a wider range of conditions - even 10% improvement in resolution has to be worth having. Cost of these things is relatively small cf the rest of a top end system?