Review: Innovations Foresight On-Axis Guider with Starlight Xpress SXV-AO-LF (1/2)

[Eden]

Introduction

Being less knowledgeable than many of the regulars here at IIS, I have had reservations about doing equipment reviews in the past but this is something that I really wanted to share with the amateur astronomy community and I hope you find it to be informative or at the very least, an opportunity for discussion.

Astrophotography for me has always been much more than just taking photographs of the night sky, it's also a chance to learn about the technology which makes it all come together at the end of the night. I'm also a firm believer in supporting the people at companies like Starlight Xpress and Innovation Foresight, people who are just as passionate about astrophotography as you are and who work hard to make advances in the field which ultimately benefit the community.

With the exception of mounts at the highest end of the market, auto-guiding is an essential component in getting great results and when the opportunity to improve on it is there, one should consider it a priority to do so.

My personal experience with the traditional auto-guiding method of using a guide scope has been satisfactory at best and an utter disappointment on the worst of nights. Unexpected problems (including wildlife!) have all thrown a spanner into the works and it is all the more frustrating when you live in a location when imaging opportunities are at a premium.

Experimenting with alternatives such as the humble OAG proved to be less than satisfying on my particular system. Finding guide stars can sometimes be tedious and is a job which is ideally suited to a rotator, the cost of which is difficult to justify purely for the ability to rotate the imaging train.

Hearing about the ONAG from other forum members last year, I simply could not resist the urge to do away with the OAG coupled to the SXV-AO-LF and to try and marry the AO body with an ONAG.

This results in a system which provides the entire field of view for guide star selection compared to the small off-axis field imposed by the OAG.

This review applies to the standard ONAG, which supports CCD sensors up to 28mm diagonal size. An XT version is available for full-frame sensors, which features a larger aperture and user-adjustable dichroic mirror.

All tests were done using a Skywatcher AZ-EQ6-GT mount, Skywatcher ED120 APO Refractor and the cameras mentioned herein.

How It Works

For those who are unfamiliar with the concept, the ONAG implements a beam-splitting dichroic mirror which reflects the visible portion of the light (< 750nm) received by your telescope to your imaging camera and the near-infrared portion (> 750nm) to your guiding camera.

Aside from giving you the entire field for guide star selection, the distorting effects of atmospheric disturbance are reduced due to the guide star image being composed from light at longer (near-infrared) wavelengths. This results in a steadier guide star image and directly translates to a measurable improvement in guiding accuracy.

Why Near-Infrared?

Around 75% of the stars in the main sequence are of spectral class "M" (Red Dwarf). These stars at the lower right-hand side of the Hertzsprung-Russell diagram emit most of their energy towards the near-infrared end of the spectrum, their output peaking between 750nm and 850nm. The abundance of these older, cooler stars makes the concept of guiding in the near-infrared not only viable, but logical.

Many monochrome CCD cameras have good relative sensitivity at this end of the spectrum, making it possible to image these stars in the near-infrared at similar exposure times to what you would normally use in visible light. By contrast, younger and hotter stars emit less energy at this end of the spectrum and are less suitable for guiding in the near-infrared.

The flexible design of the ONAG also enables the user to reverse the orientation of the imaging and guiding camera, turning the system into a near-infrared imaging platform. Long exposure images at near-infrared wavelengths penetrate gas and dust, revealing stars and galaxies which are obscured by nebulae.

Real-time Auto-focus

Possibly the most innovative feature of the ONAG is the ability to perform real-time auto-focusing.
By taking advantage of how the guide star is presented to the guiding camera in near-infrared light, Innovations Foresight have developed a software solution that is able to monitor your guide-star images for any changes and send fine-grain adjustments to your focuser as you are imaging, ensuring that you remain in focus right throughout the session rather than having to wait until the end of an exposure or until a filter change occurs.

Through the ONAG, guide stars take on the approximate appearance of a cross when correctly focused (See attached image). If and when ideal focus is lost, the guide star will appear either more elongated in the horizontal or in the vertical, depending on which direction focus shift occurs.

The supplied SharpLock software works in real-time to maintain the symmetry of the guide star, thus ensuring that your main imaging camera is always in perfect focus and maximizing the time during which your imaging camera is collecting photons.

For anyone who has experienced focuser shift due to temperature change will certainly appreciate the benefits of this technology, as will those who take longer exposures and postpone focus adjustments until a filter change occurs.

The improved steadiness of the guide star in near-infrared light ensures that SharpLock will not make unnecessary changes to system focus in response guide star distortion.

An in-depth look at SharpLock will follow in the near future.

Physical Operation

Mechanically, the unit is extremely solid and precisely engineered. Weighing in at just shy of 800 grams, it is lighter than all but the smallest of guide scopes (and indeed the accessories needed to support them), thus decreasing the load on your mount. This is an important advantage for people who are already operating their mount at or near the manufacturers recommended weight capacity. Aside from load reduction, moving the guiding system onto the imaging train also alleviates the possibility of incidental flexure at the guide scope rings and the need for dew management on the guide scope itself.

Connection to the telescope is via either female M42 or male 2" and the imaging camera is connected to the top of the unit via male M42 thread with adjustable stopping collar. The guiding camera is connected to the rear of the unit via a specialized grooved focuser with male M42 connector, also with adjustable stopping collar. Treated with temperature resilient lubrication, the focuser glides smoothly in and out of the unit and is secured using a self-centering circular clamp which firmly grips the entire circumference of the tube. Two nylon screws, seated in a groove on the top and bottom of the tube, prevent it from slipping out of the ONAG. This focuser can be removed (as I did) for cameras which do not have an M42 thread, the securing clamp working just as well on a 1.25" nose-piece or barrel camera such as the Lodestar/Lodestar 2x or the QHY5L-II.

The guide camera can be repositioned along the X and Y using a unique dual-axis staging mechanism, for a total travel of 37mm along the horizontal and 28mm along the vertical, in the event that a guide star is not immediately available near your target. Just like the guide camera focuser, the dual-axis staging is lubricated and glides smoothly in both directions for subtle adjustments. Two nylon screws along each axis firmly lock the staging mechanism in place once the desired positioning has been achieved.

The dichroic mirror behaves similarly to a star diagonal in the fashion by which the light is reflected, decreasing the outward focuser travel needed to bring the imaging camera into focus. This further reduces the potential for droop to develop on drawtube systems by maintaining the equipment load closer to the telescopes center of gravity. The ONAG dichroic mirror does not adversely affect image quality at all and is no different from imaging on a straight-through system.

Usage

Once connected, all that remains is for the imaging camera and guide camera to be brought into focus.

The ONAG ships with a selection of high quality M42 extenders to assist with additional spacing which might be required to do this. On my system I was able to bring both cameras into focus with the aid of a single 8mm extender placed in front of the main imaging camera and only needed to rack the focuser out by a few millimeters.

If sufficient inward focuser travel is available, a focal reducer can be placed either in front of the ONAG to increase the field of the entire system or just in front of the guiding camera.

Focal reducers which are positioned in front of the guide camera alone will move with the camera if the staging mechanism is adjusted along the X or Y axis, allowing guide cameras with small sensor size to benefit from stronger focal reducers (ie, x0.5, 0.33) with less chance of introducing field curvature.

Observations

Having never guided in the near-infrared before, I did not know what to expect and had reservations about whether the small aperture of my telescope would be able to supply sufficient light in the near-infrared for the ONAG to function effectively.

I was pleased to discover that there was no shortage of guide stars available in the near-infrared and that the SNR reported by the guide camera was similar to if not slightly higher than what I would expect in visible light at the same exposure time. In some parts of the sky it was necessary to increase the guide camera exposure by a small amount or adjust the staging mechanism, but I never failed to find a suitable star upon which to guide.

At one point when preparing to commence guiding, I noted that the exposures coming from the guide camera were not changing at all, leading me to the suspect that either the guide camera had come unplugged or that PHD Guiding had crashed. In actuality, what I was witnessing was evidence that the ONAG does what it claims to do by removing the distorting effects of the atmosphere. This resulted in the best recorded guiding I have had to date and was confirmed through the statistics generated by PHD Guiding (see attached image).

Whilst most cameras used for auto-guiding today have reasonable sensitivity at 750nm+ wavelengths, a small number of cameras on the market excel in both near-infrared sensitivity and low noise.

Putting a variety of cameras to the test, including the Atik Titan Mono (ICX424), Orion Starshoot (Aptina MT9M001), QHY5L-II (Aptina MT9M034), QHY IMG0H (ICX618) and Starlight Xpress Lodestar (ICX429), the last two of those five cameras were clearly the most sensitive with the Atik Titan, IMG0H and Lodestar all presenting very low noise. This was a surprising find given the larger pixel size of the much heavier Titan (7.4um) compared to the IMG0H (5.6um), both of which have TEC cooling.

Longer focal length systems would definitely benefit from a Lodestar/Lodestar x2 but the considerably less expensive QHY IMG0H (by around $300 at the time of this writing) presents a viable alternative with potentially less noise but slightly lower resolution.

To prevent under-sampling and to ensure that SharpLock is able to perform an optimal analysis of the guide star for auto-focus operations, it is important to select a guide camera with the right pixel size for your system. Put simply, short focal length systems should avoid large pixel sizes.

(Continued in Part 2)

[Eden]

ONAG with AO-- The ultimate combination?

Finally, with the help of OpenPHD, I added the SXV-AO-LF to the ONAG and put the two to the test.

It should be made clear: depending on what you're expecting to get from your AO, this combination may be better suited for larger aperture telescopes given that longer exposure times are sometimes needed in the near-infrared and shorter exposure times are often preferred for AO operation. Larger aperture telescopes have the luxury of being less discriminant when it comes to guide star availability and will obviously be able to extract a higher SNR from fainter M-Class targets at shorter exposure times.

Attempting to run the SXV-AO on my system at a rate of 5-10Hz required a reasonably bright guide star and in the tests I have conducted so far, even with the improved guide star stability in the near-infrared, resulting image quality was degraded because I was "chasing the seeing". Increasing the AO update interval (and ultimately the guide camera exposure time) to 500ms and then to 1s and 2s improved image quality and increased the range of available guide stars with suitably high SNR.

I eventually settled on 2s exposure time and later tested on another target using 3s and 4s. In spite of the AO making less frequent corrections, the adjustments made by the tip/tilt element as opposed to physically moving the mount resulted in slightly better image quality through tighter stars, as would be expected from an AO system.

In the interest of factual integrity, I have opted not to post "ONAG vs ONAG with AO" image comparisons in this review, as I feel that more testing is required, proper FWHM analysis is needed and other factors including transient changes in seeing needed to be ruled out before being able to accurately gauge an improvement over ONAG guiding alone, if any.

Anyone who is interested in seeing preliminary results can send me a PM and I will forward the comparison images to you.

Considerations


-- Because of insufficient inward focuser travel on my system, I am not able to use a focal reducer in conjunction with the ONAG and certainly not with the ONAG and SXV-AO-LF. Although my OTA could be modified to support a focal reducer, on most telescopes this would not be a problem.

-- Without modification, some Newtonian telescopes lack the sufficient backfocus required for the ONAG to operate.

-- Due to the increased load (~800 grams) on the imaging train, a good quality aftermarket focuser may be required and is definitely recommended on low and mid-ranged telescopes. A motorized focuser is obviously required to take advantage of SharpLock auto-focus.

-- Because of the orientation of the imaging camera (vertical, as opposed to the typically horizontal), effective cable management is important to ensure that the telescope can slew without cables getting tangled up or caught on any equipment.

-- ONAG requires 66mm backfocus to the imaging camera and 90mm backfocus to the guiding camera. The ONAG XT requires 68mm and 92mm respectively.

-- The ONAG XT is the better value investment; I purchased the standard ONAG and found that my CCD sensor (28.4mm diagonal effective area) only just fits into the illumination area. This is only a minor inconvenience but worth noting if your APS-C sized sensor is approaching the ceiling of this format size.


Conclusion

Would I recommend the ONAG? Absolutely. The benefits of near-infrared guiding are not simply theoretical, you can see them right there in your guide camera images and in the results that you take home at dawn. Compatible with the majority of systems out there, the reasons in favor of switching to ONAG far outweigh those against it.

I'm glad to be rid of my guide scope!

Acknowledgements

Dr. Gaston Baudat from Innovations Foresight for spending a great deal of time answering my many questions and for kindly assessing my results. His dedication, service and diligence is second to none.

Terry and the team at Starlight Xpress for their work on the SXV-AO, their friendly service and for promptly supplying me with the parts that were missing from my SXV-AO kit at purchase.

Theo from Gama Electronics for his fast turnaround on supplying the QHY IMG0H and for assisting me with the QHY12.

Everyone at IIS for making it such a great resource and a fantastic community to be a part of!

Clear Skies!
Brett

[alistairsam]

Nice Report Eden(Brett)?

are you able to post a video showing the guide star at various intervals starting from 10ms? you might have to use a screen recording software. plenty of freeware around.

at 10ms, the guide star dances all over the place in my system, and just like your results, found the best results to be between 300 and 500ms with an AO as we stop chasing seeing. I'm looking at improving this with an NIR filter http://www.edmundoptics.com.au/optics/optical-filters/longpass-edge-filters/optical-cast-infrared-ir-longpass-filters/1918 as mine's a newt and the CC has a 93mm backfocus.
would be ideal if we could use the ONAG's on newts with an AO, wonder if anyone's done that.
finally, you mentioned backfocus for the camera is 62 and guide cam is 90mm. why is that? physical design of the ONAG?

have you see the focal reducer for the guide camera for the ONAG that allows you to use an FR on the main cam?
http://www.innovationsforesight.com/store/index.php?app=ecom&ns=prodshow&ref=AFR

Cheers
Alistair

[gregbradley]

I would question the assumption that guiding in the near infrared makes an improvement in guiding.

I use a MMOAG guider which is very effective. I read this claim on the ONAG website.

I bought a 1.25 inch infrared filter that is at the same wavelength as the ONAG. I installed on my guide camera (luckily it screwed onto an adapter I had).

I used it a few times. I noticed no difference in guiding errors or improved guiding results at all.

I have 2 infrared filters at slightly different wavelengths. Neither made any difference I could detect except make the guide stars a tad less bright (not massively though).

Anyway that was my experience. Sorry to be negative but its a claim not backed up by my testing at least.

Good polar alignment, good balance, a good T-point model that corrects for slow flexes, a good PEC are far more important in my opinion than any slight gain if indeed there is one from near IR guiding.

Even selection of the right type of guide star can make a huge difference. The first thing I do when I get larger than normal errors is to select another guide star. Often I see a massive drop in errors. Too bright a guide star that is large is not ideal.

Greg.

[Eden]

G'day Alistair,



I can do that, I have nVidia ShadowPlay installed which should allow for desktop recording. You have to be careful how you interpret the results though -- even in NIR, stars which are further towards the horizon from the zenith will still be subject to some atmospheric distortion. It would definitely help with imaging a target like Pleiades from the southern latitudes.

A comparison of stars at various angles would be useful, but people just need to remember that the results are going to vary wildly based on their locations and the local conditions on a particular night. On Tuesday night the seeing in Melbourne was pretty good -- maybe 8 out of 10? Pointing near zenith, the ONAG cleaned up any effects of seeing completely and guide camera image was rock solid.

Guide camera backfocus is 90mm because it's further away from the imaging camera (by an amount equal to the length of the ONAG body) and doesn't benefit from the 45 degree AOI of the dichroic mirror.

I'm familiar with the AFR but I'm not able to connect my existing FR to the ONAG imaging port, it's male M42 on one end and 2" female thread on the other (Orion x0.85). Waiting on a Televue RFL-4087 to replace it.

[PRejto]

Hi Brett,

An excellent and very well written and thorough review. I have used an ONAG now for several years and mostly second your observations.

For imagers wanting to go automated and or remote the ONAG can present a few obstacles. The main issue that I see is the lack of a motorized stage. Yes, it's always claimed that with an ONAG you get the whole FOV of the scope but that is only true if you are willing and able to move the stage around. If you are remote (or even just in the house at the computer) this can be an issue unless you have a way to observe just what moving the stage does in real time. I solved this issue by using a Remote Desktop solution on an Ipad. I can carry the Ipad out to the scope and see what stars I can find by going into focus mode with .5 sec images. But, for a truly remote scope this simply won't work. Now admitedly with my ST-i I often didn't need to hunt for a star but recent attempts at imaging the Horsehead required some hunting and movement to the extreme corners of the stage.....and then if you do a meridian flip the star is lost necessitating another move. This can get frustrating. Since I want to use CCDAutopilot on my system I needed to find a way out of this situation. Gaston has previously indicated that there would be a motorized stage coming, but recently he said that will not be the case and he is looking into a different optical way of increasing the available FOV without moving the stage. Being impatient I have gone to an expensive but good solution with an 11 mm diagonal chip in the new ATIK 414EX. This photo shows my ONAG guider FOV vs the full FOV obtained with the G2-8300 imaging camera on my TEC180. The middle FOV is the Trius FOV on the TEC140:

http://www.pbase.com/prejto/image/159103698

Peter

[Amaranthus]

The other disinclination for me (apart from the weight/backfocus/price) is the necessity. With a TOAG and ASI120MM, I've never had any backfocus issues and never failed to find a guide star (I've never even needed to do any lateral movement to the prism stalk!). Sensitive guide cameras with QE in the mid-70%s are awesome :D

But it's horses for courses, and it's terrific that this has helped you Eden. Happy imaging!

[alistairsam]

Interesting to read Greg's experience . What filters were they Greg? Any links?
I guess the best test would be centering a star and using the same guidecam on the imaging port, looping quick exposures on the same star side by side.

Cheers
Alistair

[gaston]

It may be useful to layout the basic theory of atmospheric turbulence I this discussion.
The seeing conditions are described by two parameters r0 and t0 in the Kolmogorov’s turbulence theory.


t0 is known as the coherence time and it is equals to the time interval over which the rms wavefront error due to the turbulence is 1 radian, or lambda/(2*pi), roughly lambda/6. In short this is the time over which the image of a star can be considered frozen (constant tilt/tip values, other higher order aberrations may still exist though).


t0 is typically in the order of few ms to 10s ms. A true AO system needs to correct at, or near, t0 compensate for atmospheric turbulences. Amateur AO units are much more slower and should be seen as imaging stabilization devices, dealing with only tilt/tip very low rate image variations, such as mount left over mechanical errors (gear box, …),and low frequency components of the local seeing.
For the latter it should be pointed out that the seeing is not constant across the scope field of view (FOV), since the light from two astronomical objects separated by a small angle will travel on slightly different paths though the turbulence in the Earth’s atmosphere. If the objects are separated by a large angle, corrections based on the shape of one object’s wavefront will not be applicable to the other’s wavefront, because their turbulence induced errors will be different (no correlation anymore).

If turbulence corrections are made on the basis of a guide star, the isoplanatic angle gives the angular distance from that star at which the corrections are still valid for other objects. Tilt/tip isoplanatic angles stretch from 10 to 100 arc seconds. Which means that an image with a scale having a wider FOV than this isoplanatic angle will not benefice from AO correction anymore. As a matte fact when a target is much far away from the guide star then the guide star tilt/tip error is added to the seeing effect at this location in the image (no seeing correlation anymore), increasing the overall image blur. Therefore is better not to try using AO for tilt/tip correction at too fast rates with a large FOV (larger than the isoplanatic angle), it could make things worse.

r0 is known as Fried’s parameter, or Fried’s coherence length. It is a measure of the quality of optical transmission through the atmosphere due to random in-homogeneities in its refractive index.
It is worth to mention that both t0 and r0 parameters are a function of the wavelength of the light used for imaging, we should come back to this point later. Usually r0 and t0 are given at 550nm for the visible range.


The Fried parameter has units of length and is typically expressed in centimeters. It is defined as the diameter (not the radius) of a circular area over which the rms wavefront error due to passage through the atmosphere is equal to 1 radian, again roughly lambda/6. Telescopes with apertures larger than r0 will be limited by the seeing, preventing the instruments from reaching its diffraction limit.
The larger r0 the better the seeing, although r0 can reach up to 40cm in some good locations (Paranal Chili ESO), most of us are dealing with r0 values around 5cm to 20cm, with an average around 10cm.
Therefore most amateur telescopes are seeing limited since the scope aperture D is larger than their local r0. For instance for 8” scope D=20cm twice larger than the average seeing coherence length r0=10cm.

Seeing limited scopes FWHM for long exposure time (much longer than t0, which is the case for DSI) is given by:

FWHM ~ =0.25*lambda/r0 (lambda in nm and r0 in cm, FWHM in arc second).

Assuming an average seeing with a r0 = 10cm at 550nm the FWHM seeing is about 1.4”, a bad seeing with r0=5cm gives about 2.8”.
This should be compared to the typical diffraction limit of a 10” scope around 0.4”.


r0 is a function of the wavelength as the 6/5th power of lambda, or lambda^(6/5).
Therefore r0 increases with the wavelength, the improvement from 550nm to 850nm (mid ONAG/guider sensor NIR band pass) is:


(850/550)^(5/6)=1.4

This translates up to 40% improvement in the seeing limit. Since the tilt/tip is the dominant component (~64%) of the turbulence, using longer wavelengths for auto-guiding has the potential to provide smoother tracking, reducing the seeing “chasing” issues.
The best “true” AO results from Earth based telescope are made in IR bands.
It should be remained here that seeing, and therefore r0 for a given wavelength, is also a function of the target altitude since the atmosphere thickness increases when the above horizon altitude angle decreases. Finally tracking and auto-guiding is a complex issue, function of many parameters, beside the seeing.


I hope I helped.



Dr. Gaston Baudat


President

Innovations Foresight

[Eden]

G'day Greg,



You needn't apologize for being negative. One of the primary motivations behind posting the ONAG review was to stimulate some discussion on guiding in the near-infrared, so I appreciate your input.

I also have a near-infrared pass filter here, the ProPlanet 742 which I purchased from Astronomik a few weeks ago. Don't get me wrong -- it is a great filter as I believe all of Gerd's filters are, but like the filter that Alistair mentioned in his post, a quick look at the full transmission profile reveals that it's nothing like the ONAG cold mirror.

Firstly, neither the filter Alistair linked nor the ProPlanet filter I just mentioned block all wavelengths below 750nm (the ProPlanet 742 obviously ending transmission at around 742nm). Alistair's filter transmits from around the 665nm mark.

Secondly, most NIR pass filters transmit small but significant amounts of visible light at various points along the spectrum.

Lastly, although the Earth's atmosphere prohibits UV wavelengths lower than 300nm, these NIR pass filters transmit the remaining UV light between 300nm and 400nm (around 70% at 375nm in the case of the ProPlanet 807).

This is highly significant because of the amount of UV (short wavelength) light coming from stellar sources and the relationship between Strehl Ratio and NIR/SWIR (long wavelength) light.

http://en.wikipedia.org/wiki/Strehl_ratio

Furthermore, the quantum efficiency of CCD sensors at these UV wavelengths is just as high if not slightly higher than in the NIR, which means that the UV light collected by your guide camera is going to cancel out any benefits you would expect to see from NIR alone.

I agree that all of the factors you've listed are very important for good guiding results, particularly polar alignment and properly balancing your mount to minimize backlash, etc.

I don't have a high-end mount and when I tested the ONAG I made a point of not spending 2 hours refining my polar alignment as I have done in the past, it was all very much a 10 minute affair and yet I achieved significantly better results over what I have previously. The PHD2 graph I attached to my article shows an RMS Error of 0.34, which I recorded after leaving PHD to guide for a few minutes. Ordinarily, after spending arguably an excessive amount of time refining my polar alignment, my RMS Error would be around 0.5 on a good night and anything as high as 0.7 to 0.8 on nights with average to poor seeing.

To conclude, your comment about selecting the right guide star hits the nail on the head -- a bright star is more than likely not a spectral class M object and is emitting high amounts of UV. UV light is more susceptible to the distorting effects of the atmosphere and will result in a significantly lower Strehl Ratio and thus significantly higher RMS during guiding. The ONAG simply builds on this by clipping out the UV and the visible.

Cheers,
Brett

[ericwbenson]

The above should read :
FWHM ~ =0.25*lambda/r0 (lambda in nm and r0 in mm, FWHM in arc second).

Regards,
EB

[gaston]

Thanks Eric.

[Eden]

G'day Peter,



Thanks a lot for your feedback. Your comments in another thread were what prompted me to get the ONAG in the first place and I'm very glad that I did. :thumbsup:

I pondered the idea of a motorized stage myself, but I can think of several reasons why it could potentially cause more problems than it would solve. Your ICX825-based camera is an excellent solution, since it offers basically twice the effective resolution than the cameras I mentioned and has a good pixel size and high sensitivity in the NIR. What is the readout speed like for a full frame? For what it offers I think it's quite reasonably priced but wished that ATIK had used the smaller-form factor for it.

I don't do unattended pier flips at this stage but I can see why it could be a problem on some targets, especially at longer focal lengths. I can't however see how it would be any less of a problem on -- for example -- a single-head OAG setup. A dual-head OAG could get the same star after a pier flip, but for the price it goes for plus the cost of two guiding cameras, an ONAG with larger guiding sensor and if necessary, a focal reducer, would be better IMHO.

[multiweb]

Very interesting thread. Learning lots.



That's an interesting point, Gaston. 100 seconds of arc is really narrow. Typical imaging fields for amateurs range from 30min to a couple of degrees which is way bigger than this. Or am I misunderstanding what you are saying?

So a tilt/tip system will only correct for basic mount/mechanical errors only? Nothing that is seeing related. Even a system with a faster correction rate, such as the A07, will not work for a widefield. Correct?

[bert]

I have had a ONAG XT in combination with an stl11k and AOL for some time, in fact I think I got mine from the first bunch.

I only do remote automated imaging, so my comments will be based on that slant.

A motorised stage I think is not great value. I think the better option is to either use a larger format sensor (think ST10 for crazy sensitivity) or a focal reducer for the guide port.

I use a Sbig remote guide head for guiding with a cheap Bintel .5x 1.25 inch focal reducer and a custom adapter to fit the 1.25 filter thread to the t-thread. That not only increases Signal to noise but also increases the field of view. Innovations foresight has an IR optimised adjustable focal reducer but I have not used it.

I love the ONAG and it works well for me, (I have used both an moag and a mmoag in the past) but the HUGE advantage of the ONAG over conventional OAG is the ability to use a larger format camera as a guider. Most OAG's will only illuminate some thing like a 1/3 format sensor. With the onag you can use any camera that uses a t/thread or 1.25 nosepiece and know that the fov will be completely illuminated. This makes guide star selections less of a hassle, especially when AO is used as you need a brighter star for faster guiding. Thats without the benefits of the IR characteristics.

I have never had an issue getting guide star even at f11.

[PRejto]

Isn't that the job of a rotator with OAG and a single guide camera?

Peter

[gaston]

You are correct Marc.

Isoplanatic angles are usually much smaller than the imaging camera field of view for most amateur astronomers. I have attached two images to explain the concept. Credit Claire Max, UC Santa Cruz.
The first one shows the notion of common atmospheric path for two objects at a given angular distance, think about one being the guide star and the other one the target.
The final seeing effecting your scope is the integral of the turbulences across the all atmosphere.
In high altitude turbulences are very different for both objects since the common atmospheric path is minimum. This can be traced to the fact that for a given angle the distance between two turbulent cells is larger (remember the r0), higher you are further apart they become.
Close to the ground turbulences exhibit a large overlap which could cover eventually the all imaging camera FOV. However in the high altitude the overlap is minimum.
The isoplanatic angle takes in account the all atmospheric path, the integral effect.
It is usually defined as the angle, from the guide star, for which the Strehl ratio decreases by 2.72 (or exp(1)) versus the guide star Strehl ratio. This means that if the scope is indeed diffraction limited the Strehl ratio should be at, or above, 80%, in a perfect world 100%. At the isoplanatic angular distance it drops down to ~37%, usually in just few arc seconds. So the isoplanatic patch is quite small.
The second image shows a very good example from the Palomar AO system (R. Dekany, Caltec).
An image, inside of the Lagoon nebula, was taken with (on the right) and without (on the left) AO.
FOV 40”x40”, resolution 0.04” per pixel.
It is false color image made of three IR bands, 1200nm, 1600nm, and 2200nm. This, by the way, brings us back to the point I made before, most true AO systems are used in IR bands where seeing is much less than in the visible. One of the main reason is that in the visible range the isoplanatic angle becomes very small leading to very challenging AO systems.
The two top center cropped images zoom in the same double stars, which is also used as the guide star for the AO. On the left (AO disable) you cannot resolve them, the FWHM is 1.2”, on the right with the AO enabled the FWHM drops down to 0.2” both stars are easily resolved. Obviously this shows the power of a true AO (the correction rate here is few 100Hz though). Now look carefully at the cropped star images A, B and C. They are taken with the AO enable at increasing angular distances from the guide star (GS offset, or guide star offset). You can clearly see that the further you are the worse it becomes, and we have:

A at 5.5” => FWHM = 0.32”
B at 13.3” => FWHM = 0.45” x 0.59” (not round anymore!)
C at 23.2” => FWHM = 0.51” x 0.68” (the furthest and the worst too)

Here the isoplanatic angle is around 8”, not much and yet you are in the IR bans where seeing is reduced in average by about (1700nm/550nm)^6/5 = 3.9 times, almost 4 times better than 550nm!
You can understand from this example why using IR for seeing reduction.
The AO corrections applied to the B and C locations in the FOV increases the image blur quite a lot, and introducing even distortions and aberrations (the stars are not round anymore). This is especially true if the OA system corrects higher order wavefront errors, not only the tilt/tip component.
Seeing is random in nature and should be seen as a multi-variate probabilistic process. Which means that seeing is made of different temporal and spatial components. There is nothing such a unique seeing time value, or a localized source for that matter.
Therefore, with care, amateur AO systems may improve low frequency, close to the ground, turbulences, since their effects are more correlated across most of the imaging camera FOV. Low altitude turbulences are also slower (low frequencies) and therefore using an AO with a half second to a couple of seconds correction rate could improve bad seeing, especially if it is due to local conditions (windy, and related lower atmospheric effects). However for seeing effects from the jet stream, for instance, this is not an option. Of course mount mechanical tracking errors, usually in the second range, are perfectly correlated across the FOV and ideal candidates for image improvement using amateur AO systems at low correction rates.
Form my experience using a tilt/tip compensation AO system (an image stabilizers in fact) at fast rate (below 500ms, or so) is usually making things worst, since our FOVs are much wider that the isoplanatic angle, and especially in the visible range.
AO units are tricky to use since the seeing nature can be different from night to night.

Few words on the ONAG and guider FOV. I worked in several solutions for motorizing the ONAG guider stage, however the design must be rigid, without any play, we do not want to bring back flexure kind of issues here, and therefore eventually it becomes an expensive proposition.
This is a good reason why good rotators are not cheap either. At the end the cost brings the price close to a larger chip camera for guiding, such as the Sony ICX825 (see ATIK). I do not think a motorized stage is valuable economical and technical solution anymore. A larger FOV guider system is the best approach in my mind, providing full use of the unique ONAG large FOV capability.
Most customers having bought the ONAG for remote/automated applications, and they are many, have used this approach.
The price of a good OAG + a good rotator compares to and ONAG with a wider guider ship, yet the former solution still experience a limited overall FOV and the need of moving parts.
Some users having already a rotator have placed the ONAG guider camera slightly off axis, yet with an on axis component (like near one of the guide ship corner), to be able to use their rotator to increase the FOV when needed. In this configuration you are ON and OFF axis. The ONAG is in fact an ON/OFF axis guider. But you do not need to be way off axis, you can remain close to the optical axis where the scope performance are near the diffraction limit.
With larger imaging chips those days, guiding truly off axis to avoid casting any shadow to the main chip, may become challenging, the further of axis you go the more optical aberrations you may have to deal with.
My guess is that larger chip monochrome camera prices will go down, especially in the current (and future) completive market. Cost of a good motorized stage will be difficult to bring down at the same rate than sensors and electronics, and there is still moving parts, which I think we should minimize whenever it is possible.
Also better multi-star guiding (maybe with AO) solutions should become more common over time, back to the discussion on seeing, using stars across the field allows averaging out the seeing effect in the tracking/auto-guiding calculations. For that matter an ONAG with a large guider chip will become an even more interesting proposition.

Clear skies! :)

[gaston]

OK, I am not sure why, but it seems the two attached files did not make it.

Here are links two them:

http://innovationsforesight.com//IsoplanaticAngleConcept.jpg

http://innovationsforesight.com//LagoonNebulaIsoplanaticExample.jpg

[gaston]

In my post on seeing and r0 (Fried's parameter), there was a mistake in the FHWM calculation at 850nm versus 550nm, it should have been:

(850/550)^(6/5)=1.69

r0 scales as 6/5th power of lambda, I used 5/6 instead, sorry for that, I am getting tired :confused2:

As you can see the improvement is almost 70% on seeing using NIR versus visible wavelengths.

With Dr. Mario Motta we did an experiment few years ago to test how much we could improve guiding in poor seeing conditions using NIR.
Here is a link to Dr. Mario Motta's toy, a monster 32" f/6 relay telescope equipped with an ONAG XT and an AO-L (SBIG).

http://bostoniano.info/italiani/mario-motta-world-largest-homemade-telescope-gloucester/

The result is shown at the bottom of this page:

http://innovationsforesight.com/NIRGuiding.htm

[PRejto]

Hi Gaston,

First, thanks so much for your detailed responses about AO.

I'm probably being dense, but, in spite of the stars A, B, C (in the photo you posted) becoming progressively worse isn't the entire AO photo far superior to the photo with AO turned off (even with those distortions)? How does that translate to an argument against using AO even if the isoplanatic patch is so small?

Peter

[gaston]

Hi Peter,

The Palomar AO system makes a huge improvement, the FWHM (near the guide star) goes from 1.2" down to 0.2", 6 times better.
This type of AO systems usually correct for tilt/tip and higher order wavefront errors, at a very fast rate, amateur versions cannot come close of it anyway.
We should remember here that the all image is only 40" by 40", the AO improvement decreases slowly but surly with the angular distance from the guide star already.
Comparing to the central performance the drop is significant at the C position, but of course, relative to our experience, it is still a quite good FWHM, we would dream to achieve this, wouldn't we?
Yet the star shapes are ugly, so what would be the point (beside science) to have tight elongated/distorted stars?
Now with our scope FOVs we are talking about at least a fraction of a degree, like 30', which means 1800", not even close to 40".
The image quality that far from the isoplanatic patch will not be better, likely worse with the AO enabled, than the left image.
Using an AO unit at fast tracking rate will only improve the image in a very tiny area (the isoplanatic path) and than spread the guide star local seeing noise (the isoplanatic path seeing) all over the image, results to a worst situation than without the AO.
In short at fast rates you chase the seeing across the all atmospheric path, making better only the image in few arc seconds around the guide star, blurring more everything else.
This is why you should consider using an AO at low rate to avoid chasing the seeing, at least for us with small scopes and large FOVs.
In my opinion the true value of amateur AO is image stabilization for mount tracking issues (much faster than PE) and in some conditions for low frequency local seeing (near ground) effects.
I would recommand not to go faster the 0.5s, with most mid range mounts the AO will do a good job around one second or 2. Which is also a good choice since you do not need bright guide stars anymore.
I had very great success with an AO-8 and AO-L units and my old Celestron CGE mount, now with a PMX I do not need this help anymore, I can guide with typicality 30" to one minute exposure time, before atmospheric refraction and OTA/mount flexure come to play.

I hope this made my comments a bit clearer.

[gregbradley]

I used a 25mm 720nm IR filter and a 760nm IR filter. I see the ONAG uses 850nm so there's a difference there.

Your mention of the UV is interesting as well. So yes my test is not matching the specs of the ONAG so perhaps there's nothing to be concluded from it.

The imaging camera has these NIR and UV bands filtered out by the dichroic mirror of the ONAG - is that right?

That could also be useful for me in that I find my Trius 694 being so sensitive to IR and probably UV that I get star bloat in some images from brighter stars which per the theory here on ONAG and different types of stars output points the finger at too much NIR and UV affecting my images at times.

If I understand this thread correctly then the ideal setup would be:

1. An ONAG guider.
2. A high sensitivity and larger guide camera ideally a Sony ICX style chip which seem to be the highest QE lowest noise CCDs at the moment.
3. An AO unit before the camera.

So the advantages of the ONAG then are:

1. Better guiding.
2. More guide stars as your guide camera is able to see the same scene as your imaging camera so if you select a guide star closest to the object you are imaging you minimise the seeing differences between patches of sky.
3. Reduces IR/UV that is getting through to your imaging camera also resulting in potentially tighter stars.
4. Ability to select brighter guide stars to enable faster AO corrections thus optimising AO performance.

So an ONAG with a small ICX674 based guide camera, an AO unit like an SBIG or Starlight Express, the continuous focusing system could result in nailing down 2 of the more difficult and troublesome areas of astrophotography - the seeing / guiding as well as obtaining critical focus.

The disadvantages of an ONAG are:

1. Cost.
2. Extra weight.
3. Takes up a lot of backfocus.
4. Having your camera vertical (does this cause balance problems on heavy cameras?) possible cable tangle.
5. Does it reduce the amount of light that arrives at your imaging camera? (I though dichroic mirrors had some loss like the Sony translucent mirror technology they use in their A900 camera - that loses 10%).

By the way this is an excellent thread and we could be talking here about the next step forward in an imaging system which I know I am ready for.

Greg.

[PRejto]

Absolutely! I knew my question was not fully thought out. The blanks are now very well filled in! Thanks!!

Peter

[PRejto]

"That could also be useful for me in that I find my Trius 694 being so sensitive to IR and probably UV that I get star bloat in some images from brighter stars which per the theory here on ONAG and different types of stars output points the finger at too much NIR and UV affecting my images at times.

Greg"

I just looked at the ONAG web pages. It looks like the ONAG transmits from 370nm. The problem I experience with my TEC140/Trius combination also shows up using Baader RGB filters which if memory serves correctly transmit from 380nm. I see clear improvement once I start clipping from 410nm. Thus, I would conclude (with my scope) that the ONAG would not make an improvement in UV bloat (if indeed that is what you are seeing). Now that you have sold your TEC180 are you seeing bloat with other scopes of yours? I've always felt that two issues were involved here. With a refractor there is possible blue haloing from unfocused UV, but this would vary from APO to APO depending on the lens figuration in blue. As best I can tell the Televue NP127/trius combination is immune. But general bloat could very well be due to overflow from the tiny wells. And that I would guess has little or nothing to do with UV. Anyway this is a bit OT for this thread!

Peer

[gaston]

Hi Greg,

I used 850nm for my calculation because it is the average value on the NIR range accessible for most guiding cameras.
This wavelength range starts around 750nm (the ONAG's cut off wavelength) and ends around 950nm, the silicon sensor upper wavelength limit.
For the same reason 550nm is usually used as the average wavelength for the visible range.
This is just a first, but good, order approximation, since for more precise values we would have to integrate (weighted average) the seeing effect with the sensor quantum efficiency versus wavelengths.
By definition the UV range starts below 400nm , down to 10nm. The lower end of most silicon sensors (CCD/CMOS) is around 300nm, which means they can see some UV light.

The ONAG reflects typcially 98% of the light from 350nm to 750nm, above 750nm the relection is less than few percents, below 350nm it is not specified. However color one shot cameras, or LRGB filters with monochrome cameras, usuly cut most of the UV, below 370nm, and NIR, above 650nm. Although the ONAG will remove most of the light energy above 750nm, you may still want o consider removing NIR above 650nm or so, as well as UV below 400nm, for the imaging camera using filters, yet as I pointed out above most cameras or filter sets will do just that.
Using wavelength above 750nm only for guiding using the ONAG will provide smoother tracking, with a marginal lost in SNR.
Having a wide FOV for guiding at, and near, the optical axis associated with a large guider chip provides a unique opportunity for using multi-star (constellation) guiding, for superior performances, which is now available in Maxim DL 6 for instance, just a beginning though.

Although the ONAG option with a large guiding camera seems expensive at first, and I do understand that, it should be compared with other set-ups, such as a good OAG and a rotator alternative (+ a guider). Good quality mechanics and optics are never cheap, especially in relatively small quantity.
Beside the ONAG offers a very large FOV, access to multi-star guiding (with a large chip guider) using one scope, and the only unique real time auto-focus solution in the market (Sharplock), which solves yet another challenge.

Having the main camera vertical is obviously unusual. Yet it may offer a more compact mechanical solution limiting the distance from the scope visual back and the imaging camera, which impacts the gravity center and reduce torque coming for lever arm.
I never experience any balance problem related to this configuration.

[gregbradley]

Hi Gaston,

Yes definite advantages.

Does the real time focusing work with any electronic focuser or only certain types? Is it compatible with the Planewave CDK electronic focuser (which is now compatible with Focus Max)?

Greg.

[alistairsam]

Hi Gaston

Thanks for the detailed explanation, helped answer a number of questions.
but can you explain the effect altitude has on the size of the patch.
You mention that a low altitude, effects of seeing are averaged leading to a larger patch but its smaller at high altitude? Is this correct?
Finally are there any applications of thr onag on a newtonian with AO? The 66mm backfocus requirement does use up a lot of backfocus in most coma correctors including the rcc1.

As for the sharplock maxim plugin, how does it connect to the guide camera when maxim is also connected? And how is it dependent on the onag? Can it not be used with any oag provided the field is flat.

Thanks

AlistaIr

[Eden]

G'day Alistair,



I covered this briefly in my article. SharpLock relies on the ONAG because of how the guide star appears in guide camera through the ONAG -- approximately cross-shaped (+). When out of focus, the cross will appear wider or taller depending on which direction the focuser needs to be adjusted to restore the symmetrical cross shape.

In a ordinary OAG, the star would appear very much the same as it would in the imaging camera (hopefully round), leaving SharpLock without anything to work with.

Maxim allows 3rd party programs and plug-ins to connect to and access it's cameras. (AstroTortilla makes good use of this, as does SharpLock).

[gaston]

The isoplanatic patch is a probabilistic concept based to the all atmosphere effect (including every altitude contributions). The math behind is not trivial, but with some simplifications we can guess the general idea. To explain the concept I made a simple sketch, have a look at:

http://innovationsforesight.com/TurbulanceStructure.jpg

First we should remember that the isoplanatic patch is defined by an angle.
Now let’s, for the sake of this discussion and to make this simple, assume that the turbulence cells have all the same size everywhere in the all atmosphere above the observer. Obviously this is not true, but it does not matter much here to understand the general idea.
Inside a cell the optical properties are the same, which means the cell impacts the light traveling through it, coming from any object, the same way. Of course this assumes that all rays coming from a object are traveling through that cell. Under those assumptions the AO corrections for one object is valid for another one, the only condition is that the light from both objects have traveled through that cell indeed. On the other end, two different cells impacts in a totally different way the light, this means there is no statistical correlation between them, and AO corrections made for one does not apply for the other. This is a bit a binary simplified model, in reality there is a progressive, more or less smooth variation of correlation, from one cell to the next, which is related to the coherence length, but let’s keep it simple.
I divided the atmosphere in just three layers, low (near ground), medium, and high altitude. Since we are talking in separation angle terms here we need to translate it in distances at different altitudes.
Near ground (low) the distance separating two star lines of sight would be d-low for an angular separation theta, in my sketch this distance is about a half of a cell size (just an example of course).
For small angles (which is the case in astronomy with our telescopes) we have:
d=Z*theta

Where Z is the layer altitude, and theta the angle in radian. Therefore d-low = Z-low*theta. From this it should be clear that d increases with Z.
If I use 30’ for theta, or 0.0087 radian, and for Z-low 15m (50 feet), then d-low = 15*0.0087 = 0.13m, or 13cm, or about 5 inches.
From the sketch we should conclude that in this example the cell size is 10 inches, the size of a typical amateur scope aperture. This means that if the seeing was only a function of the low altitude cell contributions my isoplanatic patch would be defined by theta (the coherence angle) indeed, and AO corrections made with the first star (let’s say the guide star), will work just fine for the second star (let’s say my target).
But there are many layers across the all atmospheric path.
For the same angle theta at high altitude, let say 20,000 feet, or 6 km, the distance d-high = 6000*0.0087 = 52m, or about 170 feet, or 2000 inches, there are now many cells involved (200)!
This means that higher altitude layers will create seeing effects which are not correlated inside my angle theta anymore, and now their contributions cannot be corrected using an AO unit across the 10 inches scope FOV. Another way to say this is the isoplanatic angle must be much smaller than theta because the higher altitude cell contributions for the same separation angle.
Now typically, for a given altitude (layer), the all cell structure moves at the speed of the wind (V) which is larger at higher altitude (the atmosphere density drops therefore wind speeds increase with altitude).
This means that the coherence time (the time for a cell to move by its length) is usually much shorter for high altitude seeing contributions than near ground (where friction with the ground decreases the wind speed even more).
In the temporal (or frequency) component of the seeing the low altitude (ground level) contribution is mainly made of low frequencies (a long coherence time). Since the cells are close to the scope, and therefore lead to a relative large coherence angle (with some luck in the order of your scope FOV), you now have some hope for improving the near ground seeing contribution using low AO correction rates. However going too fast with the AO will chase the higher altitude seeing, related to much narrow coherence angles, therefore defeating the all idea.

Let me know if this explanation helped.



ONAG technology with Newtonian is challenging, since those scopes have typically limited back focus. We are working in a solution to eventually solve this problem though.
As Eden mentioned SharpLock (SL) for Maxim DL uses the Maxim DL COM server, therefore it will get the guide star images from Maxim DL directly. Both software applications are synchronized, you could see this as a plug-in.
SL does not work with OAG. SL assumes that the guide star shape changes in a controlled known way with the scope focus. This change must be different before and after the best focus as well, this is a critical point.
The ONAG creates such information for the guider, not an OAG. Also OAGs may face off-axis aberrations (such as coma) and field curvature making the SL concept of real time autofocus more challenging for OAGs, or self-guided cameras.

[alistairsam]

I had a play with the 685nm filter I ordered and I did see a noticeable improvement. http://www.edmundoptics.com.au/optics/optical-filters/longpass-edge-filters/optical-cast-infrared-ir-longpass-filters/1918

My guiding with the AO at 3hz is usually 0.3 RMS but with the filter on, I was between 0.15 and 0.2 RMS.
I wouldn't say it dramatically reduced seeing effects but it did help as far as I could see. conditions last night were similar to those when I've tested previously, maybe I need to test on a more turbulent night.
the lodestar X2 did struggle a bit at 300ms but I realized I needed to use dark subtraction. once I did that, I had a decent number of stars with good SNR.
will keep looking around for the right filter.

Cheers
Alistair

[gaston]

Hi Alistair,

This seems about right.

With most scopes (aperture around 300mm, or 12") the main seeing contribution is tilt/tip. Roughly half of the seeing is due to the near ground layer, the other half comes from the mid to high atmosphere.
Using 300ms will freeze some of this, especially the near ground contribution.

If you would go even faster, let's say around 50Hz the improvement in FWHM, therefore guide star wander, would be around a factor 2 to 3 in the visible, or NIR, versus long exposure time (few to 10 seconds). But this improvement will only applied to the isoplanatic patch now.

Using NIR, above 700nm or so is a good strategy which does not require tracking too fast.
With your filter, you should improve (as mentioned before) your guide star seeing (wander) by up to about 60% to 70%
Which means a 0.3 rms becomes about 0.1 rms, or so, consistent with your observation I would say (considering the model incertitude and local conditions).

The turbulence theory provides a nice model, but it is just a model with its assumptions, approximations, and limitations. We should consider its prediction with some degree variability in practice (like the weather forecast).

For instant near ground the turbulence structure could be much more complex and subtle than the model. Be in a valley, or near structures (building, roads...) may impact the results quite significantly.

I understand that your filter uses plastic substrate, how good is the optical quality?

[multiweb]

Thank you very much, Gaston, for all the illustrations and explanations in layman's terms. Fascinating stuff. :thumbsup:

[g__day]

Great thread guys! An ONAG is on my Christmas list this year.

A simple question - is the oval or X shaped stars you see in the guide star camera any issue for centeroid calaculation in say PHD or MaximDL?

[gaston]

The ONAG guide star astigmatism is quite small at best focus. It is not a concern for auto-guiding and centroid calculation. As a mater of fact using OAG may lead to much more off axis aberrations, such as coma, yet without a significant issues on centroid, unless the aberration becomes extreme.

Fo auto-guiding the right figure of merit is the half flux diameter (HFD), not much the FWHM. It expresees how the starlight energy is spread around the central peak.
With an ONAG the guide star HFD diameters seen by the imager and guider cameras are very close, no worries.