A comparison between an AO and without

[Paul Haese]

Whilst working on this project I thought it might be interesting to compare the set I started with last year and the set from this year. Since installing the AOX to my SBIG STXL11002 camera I have noted several completed images have shown distinctly sharper detail from traditionally guided data. Well at least I think it is distinctly sharper. It's not so much the larger detail components, it's the fine detail components that look sharper to me.

I have used identical processing techniques on both sets of data and it should be noted that one set contains 4.5 hours of data and the other contains 2 hours of data. Seeing was also very similar on both sets. The traditional guided images contains images which range from 2.45 -2.63" (4 in all) FWHM. With most being around 2.52". The AOX guided set contains FWHM values from 1.89 to 2.32" (9 in all) with most being around 2.11". I measured the values using CCDinspector. Now that might not seem like much but overall it seems to have an impact to the look to the image. The AOX set was hampered by high cloud in several subs. I have been imaging at 0.2 seconds with the AOX which equates to 3.6hz. It can image far faster rates but guide stars are hard to find which allow that.

I guess the point I am trying to make here is that even at modest guide rates the detail is sharper than traditional guiding. I don't think the AO works well in poor seeing though and that would seem to indicate that it does not cheat the seeing, so you need good seeing to get the best out of it. I don't know if it is purely eliminating mount related defects either. It just allows faster guiding without making many mount guide corrections (or at any rate far fewer).

Sorry about the different orientations, but this one of the downsides to using an AO.

Click here (http://paulhaese.net/NGC1097.html) for image without AOX

Click here (http://paulhaese.net/NGC1097RC.html) for image with AOX

Feel free to discuss. For me I have vindicated the cost outlay I made for the unit, but is it really difference worth that much in the minds of others?

[Slawomir]

Hi Paul,

Thank you for taking time to show us the comparison.

I looked quite carefully at both images and could definitely notice sharper stars in the image acquired with AOX, in particular in the central region with faint stars with the galaxy in the background. Also, I have noticed better star separation - with two bright stars near the galaxy, one has a faint "companion". Not sure why but in the image without the AOX, stars seem to have more distinct halos.

IMO from WIHS the AOX is a whorthwhile investment. Small improvements at each stage of creating an astroimage do add up and inevitably help to elevete final compositions to new heights.

I was looking into getting an AO unit from SX, but I would not be able to reach correct spacing with my current setup, so I dropped that idea. Maybe with my next scope...:lol:

[RickS]

Hi Paul,

The stars look harder in the first image. I presume that's a processing artefact? Are the images down sampled?

Cheers,
Rick.

[Somnium]

great stuff Paul, thanks for posting this. to me it seems it is definitely worth the outlay and the rest ! i wanted to view the images prior to finding out which one was AO to see if i could pick it up, in order to avoid post hock rationalisation, but the difference was clear. particularly with the fine detail in the core and star size. if there was a single demonstration to show the benefit of AO, this is it. results speak much louder that theoretical conjecture!

[Paul Haese]

I processed the first image last year. I had thought I conducted the sharpening the same way but I reckon I might have done a star reduction on the data. Hence the reason for the halos I think.

Tuning a system is an incremental process in my experience and it can have both steps forward and steps backward.

I had to spend a bit of time making sure I could fit everything in and in the end I changed the mirror separation a little to get everything in to the equation. So if you can more optical elements, then you might be able to get one in to your system.




You're right the stars are harder. Most likely I did not selectively exclude them when doing the sharpening. Star shapes are not that great either. One thing that an AO does provide is round stars.

These are 100% crops Rick. No resampling.




I decided to try this upon a comment that Greg made a couple of months ago. I don't really know if this is a definitive demonstration but I thought it showed something for everyone to think about.

[alpal]

Hi Paul,
The AO image does look sharper but
your comparison is not scientific because more than one variable has been changed.
May I suggest that you select your best AO suframe crop of the galaxy & the best non AO subframe,
rotate one so they are the same view - stretch both of them as a pair
& that's all.
Basically it's looking at the RAW unprocessed data.

cheers
Allan

[Paul Haese]

Ok, so here is a 100% crop from the best of each run. The first is taken with the AOX and the second is taken without the AOX. The first was affected by high cloud. FWHM's are 1.89 and 2.45 respectively.

[marco]

Interesting Paul, the difference is not very big but noticeable, these AO devices are very interesting! Pity there is not one for my camera (FLI) :(
Clear skies
Marco

[Paul Haese]

Marco, yes the differences don't appear to be very big with slower guiding. I am expecting that at some point I will be able to guide at 10hz or faster and that should have a significant impact.

FLI not having an AO incorporated was specifically the reason why I went with SBIG in the first place. I have seen enough images from FLI and the very deep cooling of the FLI cameras to have considered it in the mix. The FLI camera was a very high contender. In the end it boiled down to the AOX as the major reason for going down the SBIG path.

[Stevec35]

Hi Paul

The difference is fairly small but it is there. My experience with the AO-X has generally been favorable but it does depend on conditions. Guiding at 10Hz sounds like an impossible dream for me.

Cheers

Steve

[gregbradley]

Thanks for posting that Paul. Its been a long term question for me. I have a SX AO unit (the smaller one) I plan to implement with the Sony 694 and see if its worth it.

Anything that makes your basic data better has to be worth going for. Although I think for many tweaking up their autoguiding and polar alignment, PEC is going to make a bigger impact. But if all that is done as you mentioned then it takes it another step forward.

AOX is a good selling point for SBIG but the size of the 16803 camera, need for 65mm filters (cost of a small camera by itself!) the reliability issues of STX from the last few years (perhaps sorted now by Doug) and overall higher cost are the negatives. The KAI11002 versus KAF16803 is a debate by itself. The KAF16803 gives slightly higher QE in visible and Ha, lower read noise,but mainly gives a large FOV and has a larger full wells. Its a stunningly good sensor and arguably the best out there. But the KAI11002 has aged well and still is a great sensor and I have been surprised by the recent images using it. Its still top notch. I suppose its an oldie but a goldie!
KAI sensors have no RBI but they can suffer from timing issues.

No one sensor is perfect though, much like scopes.

AOX/AO units may be make more sense to those imaging from dark skies with routine good seeing and at longer focal lengths.

Greg.

[Eden]

I'm glad you took the time to do this, Paul. Thanks for sharing it.

I think it's perfectly reasonable to assert that the AO set is the sharper of the two. To me, it's a considerable difference; more than just slight improvement, but I respect that others may see it differently.

I've wanted to do a similar comparison myself but my SXV-AO-LF is yet to come back from the UK and I feel that such a undertaking would be more successful on better optics than what I can offer at this time.



Although not completely airtight with respect to scientific testing methodologies (hah, not like I could do any better...), you've produced compelling evidence that there is an improvement in image quality with AO enabled.

The need for good seeing may at first seem illogical; one could argue that the improvements came from the seeing and not from the AO, a conclusion which I myself considered at one time. This isn't the case. Even on a high-end mount such as your own -- which I imagine is defect-free or close to it -- a properly configured AO unit should be able to achieve corrections which are both more responsive and more accurate than those of the mount, since we're talking actuator-mounted glass lens vs something much heavier and less agile. This is why the guide rates need only be modest.

You've obviously nailed down the other relevant guiding parameters too, because AO doesn't automatically work well just because of good seeing.

Is there any chance that you might consider doing an AO/non-AO set back-to-back on the same night? It'd be interesting to get an idea of exactly how much better AO is when differences in seeing are removed from the equation.

P.S. I was in Adelaide a couple of weeks ago and got a first hand look at how much better the skies are there compared to this miserable city! I can see how it is that you manage to get so much imaging done :P

[Eden]

I can't recall what you're using there Slaw, but it might be helpful for you to know that some of the default parts that come with the SXV-AO can be replaced with slimmer aftermarket alternatives (I got mine from Precise Parts). This applies to the OAG which ships with the unit also, which is very robust in terms of backfocus cost.

[gvanhau]

Good comparison Paul.
The improvement is easyly visible.

I have a SX AO unit from times when I only had a CG5 mount. May be worth thrying if with my current setup...

Geert

[alpal]

Thanks Paul,
I can see that the AOX data is better.
It's certainly worthwhile using it for the maximum sharpness.
As others are saying:
I'd love to see what it could do on the same night
with the same seeing but at 10 Hz.

cheers
Allan

[vlazg]

A comparison with ONAG would also be interesting.

[Paul Haese]

Like you Steve I wonder when 10hz will be a reality. In any event round stars are assured so long as guiding is maintained.



Greg I am sure the AO with the 694 will produce good results.

The STXL series is the one that has had the issues. To the best of my knowledge the STX series has been fine. Both the STX and STXL can use the AOX, so there is some room there. Whilst the 16803 sensor has lots of attributes going for it, my essential reason for deciding on the 11002 sensor was actual the aspect. I prefer the rectangular view. I suppose the price was also an influence too.

You do need good seeing for long focal lengths. Otherwise it wobbles around like an mad man's custard and all you get is large round stars with slightly blurred detail. I have not had a real cracking night of seeing since installing the AOX. I am expecting as the next drought approaches I will get a few nights like that.



You're right I should have taken a more scientific approach to showing the results, but in the end I think people get the message. My last couple of images with this scope and camera combination have been sharper than previous images and that is what I wanted to show.

If I get the time I will try for a night of AO v non AO. Whilst Adelaide itself has lots of cloud free nights, where my scope is located is near the south coast and there can be lots of nights this time of year that its gets clouded out due to the presence of the high pressure systems. It is no doubt better than the eastern states, but still an ever present problem. Given that I like doing mega data images, even one night this time of the year can hurt. So no promises but I will see how I go.

For what is it worth having remote automation is what allows me to get so much imaging done. I can do it from home or anywhere really so long as I can get an internet connection. Even if I get a few hours its better than none. And; I do it every night I can. I have an understanding wife too which helps.



Geert anything is worth an experiment I think.




That will have to wait until I find the necessary star Allan. This time of year there are not a lot of bright guide stars to choose from and my list of projects was delayed by the repairs to the camera/s over the last couple of years.



Whilst I don't have any issue with the ONAG concept, my imaging rig would not support one. I don't have enough back focus. Nor would my budget. I did consider this approach before deciding on the AOX concept. I know there is some belief out there that it cuts through the seeing guiding in IR, and I would expect that guiding would be sharper than standard guiding but I actually wonder if it would produce as sharp results as an AOX. Remember that a small piece of glass has virtually no mass and guiding with the whole mass moving very quickly has consequences.

[Eden]

Paul I was wondering what guide camera you are using with the AO-X?

[Paul Haese]

Brett, the filter wheel has an integrated guide camera in it. It works via a prick up prism. The image data is then relayed to the AOX and software for movement commands. The mount only moves when the AOX reaches the bump value which is about 70% of total movement. It then rapidly moves the mount in a short series of bumps whilst the AOX guides to return both axis to home position and the process begins again. It does this several times over the course of 30 minute sub.

[Eden]

I get that part mate, I was just curious as to what CCD is in the guide camera?

[Paul Haese]

Ah sorry about that.

Specifications here (https://www.sbig.com/products/filter-wheels/fw8g-stxl/)

It is KAI 340 sensor with 7.4um pixels. It is the same one as in the SBIG ST-i camera.

[Eden]

Not a worry, thanks Paul.

[topheart]

Hi,
Is that totally correct?.... I thought you could get away with 50mm square filters with this setup AOX + 16803, at least Don Goldman indicated that to me recently.....

Thanks,
Tim

BTW, Thanks for posting this Paul.....very interesting indeed....I also thought you needed an off-axis guider and remote guide head with the AOX?? Excuse my ignorance....I might change out my FLI16803 for SBIG because of the AOX with my 4.7M focal length!!

Cheers,
Tim

[Paul Haese]

With the STXL series you can get an integrated filter wheel with guider. The STX series is another matter though, it requires a MMOAG, but you can then use an AOX. You would also need the remote guide head too.

[Paul Haese]

It looks like you need the 65mm filter for use of the internal guide chip but can get away with 50mm if you are using an external guider.

https://www.sbig.com/products/filter-wheels/fw7-stx/

[gregbradley]

Thanks for that clarification.

I guess a large square sensor takes up the room that the guide sensor in the filter wheel would normally use.

Greg.

[topheart]

Thanks very much for that info Paul and Greg!

I can see where I got confused now!

Main thing is, Paul you are on a winner with your gear, for sure!



Regards,
Tim