Anyone tried adaptive optics at f4 or thereabouts?

[Shiraz]

Has anyone used a refractive AO at f4? I did some calcs to see how much SA is introduced by a flat plate at f4 and it seems to be about 1/4 wave, which will be well within the seeing blur. Not sure of CA though - a back of envelope shows longitudinal CA of 77microns assuming BK7, which is a bit too much for hi res broadband imaging with my f4, but would be fine for narrow band or even RGB. Grateful to hear from anyone who has tried refractive AO with a fast scope - or if anyone has the ability to ray trace such a setup.

very grateful for any advice - I don't want to buy one and then find out that it produces lovely round stars, but degrades resolution. Cheers Ray

[alistairsam]

Hi Ray

I just got one and am keen to test it out with my 10inch F4, are you able to explain your calculation and finding of CA (in simplistic terms If possible), is this due to the refractive element? How would it degrade resolution with round stars?
They say results are more evident at longer FL's but if you're able to guide at 7Hz, wouldn't that constitute an improvement even at f4 from usual 1Hz guiding?
I'll be using it with an RCC1 and stf8300m, mostly for narrowband.

Thanks
Alistair

[Shiraz]

Hi Alistair. Sounds like a good setup.

Because the refractive element will have dispersion, it's optical thickness will vary with wavelength and red will focus at a different position to blue after passing through. I am not an optics designer, but a back of the envelope calc showed that this longitudinal CA will be enough to produce enlarged spots at blue and red at f4 if imaging over the full 400-700 range. All is not lost though - if the camera has dropping sensitivity at both ends of the spectrum, it may not matter much. My camera has high blue sensitivity, so I was wary of getting an AO - I could not get any manufacturer's advice on how much CA to expect. If you are using filters though (RGB or narrowband), this is not an problem at all and there also should not be an issue at higher f numbers.

The bigger issue is that the AO will not be guiding on the same part of the sky as you are imaging in, so the seeing-induced motion that the guide cam is tracking will not be correlated with that in the image - eg if the seeing is moving your guide star left, the galaxy that you are imaging could well be moving to the right at the same instant, so the correction from the AO could easily make things worse. In the extreme, AO will make the stars nice and round, but 1.4x as big (from memory, I think that Rick found slight star enlargement with an AO). If you use the AO at lower frequency so that you are not influenced much by seeing, it should take out any residual mount errors, particularly those induced by wind. I don't think that AO can help get rid of most seeing induced tracking error with the systems we use, but it will certainly help with mount guiding errors at lower frequencies - be very interested in what you find. Regards Ray

[multiweb]

I did on my 8" newt and it somehow worked but my native guiding is pretty shocking at best. It does wonders at FL longer than 2m though and worked quite well on the C11 at prime.

[bratislav]

There is no issues with aberrations even at f/4, unless you do critical high resolution planetary work. Here's a 6mm BK7 plate (same as in AO-L units) at 85mm from focus at 420-700nm .
40 microns of chromatic focal shift may sound like a lot, until we remember that all uber expensive APOs would easily have five times that (or even more)!

[alistairsam]

Hi

I am using a filter wheel with lrgb and NB filters so will have to focus between filters.

My only concern is the vignetting and light loss cause I have to move the mirror up for the 94mm spacing reqmt of the rcc1 and I had to move it up for the 2.5" moonlite's 74mm height.

Edit: your post in another thread answered my question on how the guide star moves in one direction and the galaxy in another.
this makes it clear. scary though. is ONAG the only way around this if we want AO to negate or improve bad seeing?

http://www.footootjes.nl/Astrophotography_Seeing/Astrophotography_Seeing.html



Cheers
Alistair

[Poita]

Ray, I have an AO unit on my shelf at the moment, too crook to use it. You are welcome to borrow it if you want.
Bounce me an email (I don't check my PMs as often).
I'd be keen for you to try it if you share the results here with others.

[alistairsam]

hi Marc,

which AO did you use and do you have any results or comparisons on the 8inch? was this F4 or F5?

Cheers
Alistair

[alistairsam]

I've been reading about seeing and limitations of AO and can see your point Ray about the guide star being away from the target and high and low frequency components and how AO could make the stars bigger in some cases.
But wouldn't this be less pronounced at F4 as its a wider FOV?

That video of bad seeing with the moon, must've been at a long FL.
my FL is 1m and I haven't noticed anything as turbulent as that, so it may work out well at F4. remains to be seen.

I had a look at Multistar guiding with Jim's plugin for Maxim, but I don't think it works with AO.

My Baader LRGB filters aren't exactly parfocal, it'll be interesting to see how much the refractive element of the AO offsets focus for each filter.

I should get focusmax working, so with autofocus between filters, the longitudinal CA you mention shouldn't be an issue.

you've mentioned SA was less than 1/4 wave. will the quality of the coma corrector exacerbate this at all?
I remember reading that the MPCC had worse SA than the RCC1.

So if SA with the RCC1 is negligible, only component left is the refractive element of the AO. interesting...!!

will post results once I get my Lodestar X2 and RCC1.
Hope to try both high and low frequency guiding should the Lodestar X2 live up to the claim of increased sensitivity.
I don't have a rotator, so will just have to live with adjusting frequency based on guide stars.

Cheers
Alistair

[Shiraz]

thanks for the advice Marc

thanks very much for taking the time to do that Bratislav - it is great to have reliable information. Much appreciated.


vignetting should not be too bad - I tried my old 200f4/RCC1 with my QHY8 and it was usable over the full APSc frame


That's a really generous offer Peter - will email. Sorry that you are crook.

I think that you will be fine with your proposed setup now that Bratislav has provided a reliable assessment of CA. The RCC1 is very good and should not add any problems.
The isokinetic patch is generally quite small (eg arc minutes) and any guiding system that uses a guide star that is outside of that patch will suffer from the additive nature of the errors from the two paths with uncorrelated turbulence. It is not just a problem with AO, but AO works at high frequencies, so there is less integration to reduce the worst effects. I don't think fofv is an issue, the thing that matters is the pixel scale relative to the scale of the seeing jitter.
Multistar guiding would be ideal for correcting mount tracking problems with AO - by using numerous stars, it would be possible to get a "seeing free" estimate of mount pointing.

regards ray

[Peter Ward]

Your insightful analysis has saved me much trouble. :thumbsup:

I regularly use AO with an F3.8 system +16803 sensor.

The simple fact is: stars have more intensity and better FWHM's with AO across the entire field, than without AO.

Further with AO any CCD focal-plane tilt often masked by a standard guiding blur circle, becomes obvious with the AO.

As for CA....trivial at best...I've never detect any difference with or without.

[Paul Haese]

I am rather interested in this myself as I am going backwards and forwards about getting a field flattener, which means I lose the rotator but I could then have an AOX (Peter ;)). The seeing at Clayton has been a little average of late but I am sure an AO would make a huge difference to the subs I have coming out of the system. There is certainly a vast difference between f4 and f8, but still an interesting read.

[Shiraz]

Thanks for the info Peter. I would be surprised if you had much seeing effect at 2 arcsec sampling - seeing blur will be smaller than the pixels in good conditions. But, if you are getting tighter stars with AO, then it has to be worth having.
You would not expect to see any variation across the field, since seeing correction only occurs in the isokinetic patch surrounding the guide star - the imaging sensor will be well outside of the isokinetic patch and the whole field will be equally uncorrected for seeing.


Not convinced that it is going to make a huge difference Paul. There is a comprehensive literature that shows that seeing correction is only possible using AO over a small region surrounding the guide star. There is a bit of marketing hyperbole out there suggesting that AO systems can "correct for seeing", without saying under what conditions. The physics shows how limited that correction really is. AO is likely to be very useful for cleaning up wind and mount jitter/tracking effects, particularly if you take long subs. But as far as I can tell, it cannot correct for seeing over wide fields (eg > 1 arcmin) - it would be great if it were otherwise and I would be very pleased if someone could demonstrate why the limitations identified in the theory apply to professional systems, but not to ours.

As for the seeing of late, there have been a few periods over the past few weeks where the seeing was below 2 arcsec over this way - exceptional. Much of the time I get somewhere around 2.5-3 arcsec, but there are also times (like early on last night) where 4 arcsec is as good as it gets.

Regards ray

[bert]

I have been using SBIG Adaptive optics for some time and I can see value in them.

That said, I was recently at the Hale Palomar 200 inch observatory with their head engineer, and the techs were installing the adaptive optics unit at cassegrain focus. (AO is only at cassegrain not prime).

The adaptive optics they use has 1000 actuators deforming a mirror at 1000 cycles per second. And all the cables between the computers to the adaptive optics unit are hand made, they are called dreadlocks, because they sure look like it. To get the computing power to cause 1000 actuators to correct 1000 times per second is extraordinary. Thinks about this, it reads a sensor makes 1000 calculations to where best 'atmospheric correction' for each actuator, sends that signal for the actuator, after it moves, checks the result recalculates and repeat.... 1000 times per second. There is huge amount of processing power needed to to do this and to get that power, There are racks and racks of computers hanging below prime focus that RUN AO corrections ONLY, no other functions. Because these computers generate a lot of heat, (and heat released into the air is a bad thing for seeing around the telescope) the 200 inch drags glycol coolant pipes around the floor to external coolers to shed the computer generated heat.

Amateurs are using a single corrective prism in 2 axes in an ao unit at around 1 to 10 hertz, that uses a fraction of your desktop computers processing power. So it not at all comparable to professional systems.

So in reality, the amateur AO is not in the same league, So Is it worth it?

I think so. Here's why:

To make an auto guiding correction with a mount requires a movement of mass of all half your mount and all of your optics and imaging train. Then there is cables that may resist movement and you have over effects like oscillations, backlash etc.

Or you could move a small prism weighing a few grams at 10 times the speed....

Do I think amateur AO correct for seeing? Don't know. But it makes one hell of a fast auto guider and better stellar profiles. That will do for me.

I am just setting up a system with AO and the new Large format ONAG. I am very keen to see how this works, because infrared is less affected by seeing than the visual spectrum. So Ray according to your reading, As the onag guides from the centre of the cameras field of view, not off axis, it is more likely to do correct for seeing. What do you think?

Brett

[Peter Ward]

Ray...amateur AO systems clearly have no capacity to correct for anything other than isoplanic errors.....i.e. tip-tilt errors.....but as this is the first order correction of any comprehensive AO system it IS worth doing....as there is no way you can get, in my case about 100kg of mount/telescope/counterweights to move reliably in any direction at 15Hz!

AO really does pay the dividends as I mentioned earlier.

[PRejto]

I read both sides of this argument here and elsewhere and I don't get it. Logically what Ray is saying has to make sense; seeing will be different in different parts of the FOV. If the AO makes a correction based on seeing only around the guide star - and - seeing is different (or moving in the opposite direction elsewhere in the FOV) then part of the image should improve at the expense of areas in the remainder of the FOV.

However, if the mount is tracking improperly, then what the AO sees around the guide star applies to the rest of the FOV providing that seeing is reasonably good.

So, if I've got that right, I don't understand why many are seeing improvements with high end mounts (like ME, MEII, APs etc) where protrack and great polar alignment can provide terrific 10 min unguided round stars.

The only explanation that comes to mind is that perhaps seeing really doesn't always vary so much over the FOV on average. Certainly I can see the benefit to using AO on lesser mounts. I just don't understand from any theory why a really high end mount properly set up should see such improvement. I'm not doubting anyone's statements to the contrary. I just want an explanation.


Peter

[PRejto]

Peter,

I'm really curious about this and not doubting your results. Do you think you get better results because you are just making quicker corrections to "normal" tracking errors (not mount errors)? If so, then I would guess that it takes good or better than good seeing to achieve this. Otherwise wouldn't the mount be bouncing all over the place at high frequency chasing the seeing?

Do you think that a set-up that is capable of going 10 min unguided (as some claim) would also show improvement (outside the isokinetic patch) using AO?

Thanks!

Peter

[Shiraz]

I think that it depends Brett. the ONAG uses NIR, so it will not see the same seeing motion as you get in the image. If AO only really provides a better auto-guider, that's a good thing, since you want to measure only the mount pointing, not seeing motion. However, it will be less useful for correcting seeing, since, although the isokinetic patch will be in the field of the main camera, the guide cam will not be recording all of the seeing motion.


If AO systems work then that's a good thing, as I said. However, I don't see any way that what they do is related to full AO systems - which operate in the NIR and typically have isoplanatic patches (over which they actually work) of a few arc seconds. Even with simple tip/tilt, the region of operation where seeing is corrected is small and outside of the main camera field for off axis guiding - amateur AO systems are generally not set up so that they can correct for conventional seeing. My question then becomes - "well what are they actually doing that gives the improvement that you see?".


I totally agree Peter - a lot of this does not add up.

If AO cannot correct for conventional seeing as amateurs use it (I'm pretty sure that this is what the physics says), then what does it do that gives benefit to users? Of course you could quite reasonably just say "I don't care", bolt one on and go for it, but I think that trying to understand what is going on has more than acaemic interest for the following reasons:

1. If AO just provides a fast autoguider, maybe it can do so quite effectively on all mounts - so could it be that we don't really need red anodised ones - or is there a minimum mount quality that is required before AO pays off? In any case, how come high end mounts that can turn in great results without any guiding at all, still benefit from AO?

2. The physics suggests that correcting for seeing outside of the isokinetic patch should make things worse, so is there a tradeoff between improved tracking and reduced resolution? ie, it might be worth knowing when AO is likely to pay off and when you switch it off - particularly for robotic systems. If AO primarily provides a high resolution guiding capability to tidy up the mount errors, there will be advantages in using multiple stars for determining average mount pointing - maybe we should be asking the guide software people to include multi-star sensing if possible.

3. If AO is actually dealing with some form of local seeing, what is it caused by? - If it produces correlated effects across wide fields of view, the source has to be close to the scope, so maybe it is just due to tube currents/boundary layers or local heat columns. If we could identify something like this, we may be able to improve our systems without needing to use AO - or get them working better in conjunction with AO. Maybe you can get the advantages of AO just by using a couple of high flow fans for example.

4. If it can be demonstrated that guiding within the isokinteic patch with existing AOs does make a significant difference to resolution, could it open up an exciting new area of high resolution DSO imaging for amateurs?

Lots of questions, no obvious answers. regards ray

[PRejto]

Thanks Ray!

I like your analysis, particularly your ideas in #2 and #3.
Thanks,

Peter

Edit: I speculate that improvements - due to AO on high end mounts - are only going to be seen in good to very good seeing to start with (such that the AO system is not so much following seeing variations within a small isokinetic patch but mount tracking errors due to polar alignment and/or poor uncorrected PE that is greater than the blur lost to seeing. I would love to see proof that AO could improve an image on a system capable of long unguided exposures. Probably most imagers using mounts capable of long unguided exposures are not using the mounts in such fashion and thus depend on guiding to make tracking corrections. Obviously AO is going to do a better and smoother job making these corrections up until a certain limit is reached in the AO system at which time a mount correction is issued. Probably just the fact that the mount is making fewer corrections over the exposure duration accounts for the benefit seen.

[Shiraz]

That makes a lot of sense - if you are right, this could explain what AO is doing and when it will work best :thumbsup: Thanks, Ray

[Peter Ward]

Perhaps this will answer some:

http://www.sbig.com/about-us/blog/differential/

Alan's analysis I believe is quite thorough.

[stevous67]

I have no doubt my AOX works @ F8, I wouldnt bother imaging without it:

M8 (https://farm8.staticflickr.com/7204/14088559881_0e16b0cd6e_o.jpg) [Taken last month, majority of the data collected from a suburb of Melbourne, 12.5" / F8 HaRGB]

Steve

[PRejto]

Peter,

Thanks very much for posting the link to the Alan Holmes article. Tha really answers quite a lot of questions I've been asking.

Peter R

[alistairsam]

That's stunning Steve, perhaps someone can take a 10min exposure with and without AO.
or a video of the main camera's FOV to show seeing effects.

Alan's analysis was very informative, but I guess the bubble that he talks about might be smaller in an F4's FOV, end result is to be seen.

I've been trying the Multi star guide plugin in Maxim, just need to get it working with the AO as the guider.

Maybe someone here can write a multi star guide plugin for AO!!

Cheers
Alistair

[Shiraz]

thanks Peter - very useful summary and insight into SBIG philosophy and it has crystallised a number of ideas that had been previously unsupported by data. :thumbsup:

However, I found that it warranted some very careful reading to be sure what it is saying and what it is not saying and to understand what the data actually represents. Overall (and perhaps rather surprisingly), I think that it confirms most of what has already been said on this thread:

1. The measured tracking data shows the expected high speed seeing fluctuations, (uncorrelated over large angles) plus an underlying slow variation that is widely correlated. The author assumes that the slow component is due to some form of seeing, but seems to dismiss the (more obvious?) possibility that it is underlying residual mount error - not sure why.

2. Regardless of what the slow component is, the “fast” simulation in figure 6 shows that the performance drops when two star tracking is used – let’s say that again: the data clearly shows that the stars would be bigger if you turned on an AO under these conditions. This is explained in terms of the system chasing the fast seeing – exactly the problem identified earlier in the thread.

3. The “slow” simulation data in figure 7 shows that if you slow down the AO update rate, you can integrate the seeing and get some gain from using an AO. ie, there is likely to be an optimum update rate that will vary with mount characteristics and seeing – faster is not necessarily better.

4. Both simulations show that multi-star guiding is best – this was indicated as a possibility in earlier posts, but primarily if the AO was being used to correct mount errors. In any case, as far as I know multi-star is not an option at present, so the main performance gains shown in the figures refer to unavailable technology - looks like it would be a good idea for the future though.

5. The simulation graphs 6 and 7 look like they are star profiles and you could easily assume that to be the case – however, as the author states later in the text, that is not what they are. Star profiles will only be obtained by convolving these tracking functions with the seeing function. The measured track functions all have quite small FWHM. So the effect of AO correction for the test system will at best be only slightly noticeable in the star profiles, and then only if the seeing is very good (eg below about 2 arc seconds FWHM). AO correction will not be of any consequence at all for worse seeing, where the tracking will be completely overshadowed by the seeing. Looks like Peter’s theory is on the money.

6. It appears that SBIG is working on a complicated way to incorporate multi-star tracking and to have the tracking stars in the imaging field of view using a separate guidescope – clearly they have identified limitations in the current approach.

In summary, based on this paper:
- AO will tidy up mount tracking/slow seeing, but only if the AO update rate is at low enough, or based on multiple stars.
- There is likely to be an optimum update rate based on seeing and mount characteristics and it will probably be between 1Hz and 10Hz, based on the limited data available.
- AO will not tidy up conventional fast seeing outside of the isokinetic patch. If the update rate is too high it has now been demonstrated that it will actually make matters worse (by a factor of up to 1.4x from the theory).
- There is a slow component to the tracking error that is correlated over large angles and time. It could be either residual mount error or some (previously undocumented?) form of slow seeing – the verdict is still out - but whatever it is, this is the only part of the tracking error that current AO will correct.
- If you have a good quality mount, AO correction will probably only have a noticeable effect in very good seeing – it certainly will not fix up bad seeing, but it could fix up a lower quality mount and help with a good mount if used with care.

[Poita]

My unit is in the post to you, so you will be able to find out for yourself soon :)

In my case it made my not very good Celestron mount outperform my G11 that had been tweaked up and fitted with an ovision worm. THe Celestron without the AO unit was fairly average. YMMV, I'll be keen to see the results of the loan.

[Shiraz]

lovely image Steve - can't argue with that.

[Shiraz]

thanks :)

[Poita]

Which mount/scope combo will you be using it on?

[Shiraz]

a 200f4 on an EQ6 at this stage - without a CC initially, to make sure that it is working at full resolution. Will be interesting to be able to test some of the theory, rather than rely only on logic - time to stop talking and put some rubber on the road. :)

[Peter Ward]

I guess I'd call my PME mount good quality, (raw PE about 2 arc sec) that said, my experience has been using an AO has never made images worse. But I also don't bother imaging on nights where what I call "fuzz-ball" seeing is in effect.
Such nights show so much bloat on stars to make you think the telescope's optics are defective...begging the question, why even attempt deep sky imaging when conditions are so woeful.

However, my reference tome on Adaptive Optics has a discussion on the size of isoplanic patches within the atmosphere, which can, on good nights extend out to well beyond a metre in aperture. This is precisely the slow seeing tip-tilt AO is designed to correct.

If it was only a slow mount tracking error, then increasing the frequency of mount corrections would also show an improvement...yet I have consistently found this not to be the case...the end result often being eggy stars from chasing the seeing, albeit way too late, rather than correcting a real drive error.

[gregbradley]

I'd like to hear about performances of various AO units. SBIG has the most experience here. The latest AOX sounds very good. How does it compare to the earlier models?

How easy are these to use? Does it depend on having an appropriate guide star to get enough hertz to make a difference? Is this easy or hard?

I am thinking my CDK 17 is screaming out for an AO unit and I don't have a lot of confidence in Stalight Express. Their AO unit is about US$2000 so its not cheap either. SBIGs one is about the same $ but they have the runs on the board. I saw a first light from Martin using STXL11 and AOX and its pretty amazing. I was surprised to read he is excited about using this as I thought he was already pretty much at the peak with what he had.

Greg.

[alistairsam]

Maxim DL6 has multi star guiding support built in. Just change the dialog box size to multistar guiding.
Not sure how to use it with AO though. it might work, will try it out.
downside is that it chooses the full frame, you can't select a subframe, so USB download limits may hinder running it at over 5Hz

had a few questions -
All AO units have the OAG pick off prism behind the refractive element, won't the guide star position get affected as the glass tilts?
is there a fixed focal position for the refractive element or can you have the imaging CCD at varying distances based on your configuration?

Cheers
Alistair