AO ( Active Optics ) for Amateur Imaging
from
Benoit Schillings
Brad D Wallis
The two nuts who brought AO to Deep Sky Imaging.
Wallis and Provin described what seemed to be a resolution limit
in AMOACP (Ch.4) and derived both an expression for that limit
and what seemed to be a reasonable explanation for that limit.
The outcome of that investigation was a formula which said that
the best long exposure photographic resolution obtainable under
optimal seeing conditions was described by :
Rp ( arc-seconds) = 24.3 / APERTURE (in inches)
( as it turns out, the physical explanation we happened upon was a matter of total coincidence as later events have proved )
In 1992 Benoit Schillings and Brad Wallis entered into a series of
discussions, debates, E-mail exchanges over this issue. While
these discussions continued, Robert Provin and Brad Wallis had
managed to BEAT the 24.3/A 'limit' which added fuel to the
ongoing Schillings-Wallis discussions.
The topic of these discussions was basically over the matter of
IF & HOW to best BEAT this 'limit'... and the debates centered on the
real nature of this 'limit' .. Was it primarily driven by OPTICS,
SEEING, GUIDING, FOCUS, or PHYSICS ? The fact that Wallis & Provin
had beaten 24.3/A down to ~16/A voided the explanation put forth
in AMOACP and placed that explanation into the realm of unfortunate
coincidence.
Schillings and Wallis were basically debating fine points of this
problem. Both agreed that SEEING was a major cause of the apparent
resolution wall. Could it be shattered ? What other factors were
major players in the problem ? Schillings proposed an RMS error
solution which had appeared in the literature in various publications.
A look at the matrix created by an RMS study of SEEING, GUIDING, and
FOCUS by Wallis showed that SEEING and GUIDING were quite probably the
major factors in the problem.
Rp = SQRT ( S^2 + G^2 + F^2 )
Yields a 3 dimensional matrix. By looking at this matrix it seemed
probable that FOCUS was a minor player and seemed very likely that
SEEING and GUIDING were players of similar magnitude.
The use of the ST-4 and the first crack in the resolution wall had
hinted at GUIDING as being a larger factor than Wallis & Provin
had previously thought it was. By using the ST-4 and its frequent
guiding corrections ( 2-3 sec vs 10-30 sec ) they had broken the
24.3/A 'limit' and had pushed down to ~16./A as our 'limit'.
By operating at sites that W&P had found to have good/excellent
seeing they could push on this 'limit' more effectively than
would have been possible in less optimal seeing conditions.
The look at the 3-D RMS solution narrowed down the problem. The
ST-4 provided a clue at a solution: GUIDE FASTER ! However, a
typical mount has so much interia that it can not respond to
guiding tweaks at high frequency.
Was there a WAY ?
If there was a way, WOULD IT WORK ??
In October 1995 Brad W. asked Benoit S. if he could get a data file
out of the ST-4 or cookbook camera or ST-7 that would yield a high
frequency sampling of the effects of seeing. Benoit altered the
camera driver code and obtained a data file of about 2 minutes of
30Hz seeing data. With this, Brad set in for some modeling to see if
there was a way to 'fast guide' and to see if such 'fast guiding'
would yield a useful result.
The result is shown in FIGURE 1...
The result was EXCITING.. Both Wallis and Schillings were now
convinced that there might really be a solution.
HOW TO FAST GUIDE was now the question.......
From late October until early December of 1995 they debated the
best way to go about the problem of fast guiding during an
exposure.
In December of 1995 they both were agreed on the best way to attack
the problem and had a device in mind for testing the solution.
At this point, the problem moved almost entirely into the realm of
expertise of Benoit Schillings...His work and expertise with computer
control and device driver programming allowed him to dive into the
electronics and interface problems and to modify ST-7 drivers to
have them take on the control interface for the fast guider unit.
I believe that Benoit characterized this as a "minor problem" when
he began... He might revise that statement these days ! <G>
At this time we made SBIG aware of the potentials of this device
and they were quite enthusiastic. As Benoit began to work up a
prototype for testing, problems arose in the design of a signal
chain amplifier. After a few discussions with Matt Longmire of
SBIG the problem was solved when SBIG volunteered their help in
the design and construction of this component to help speed the
prototype construction. SBIG had already made their ST-4 and
ST-7 drivers available and Benoit set about to enhance the workings
of these cameras/trackers for this project.
Within a month of the intial 'IT's POSSIBLE' concensus, Benoit
had stuck together (literally!) a test unit and was toying with
it in his living room on artificial stars. Within 2 months, a
first light image had been produced that showed us both that this
device had great potential for crashing the resolution limit.
( a side effect is an increase in limiting magnitude )
That 'first light' image added to the enthusiasm level.
The winter of 95-96 produced long stretches of fairly bad seeing
which frustrated testing efforts. An early image of the
Eskimo nebula had pushed down to 1.1 arc-sec FWHM but that
figure remained the benchmark the next 6 months as bad seeing
held the resolution of test images to about 2 arc seconds.
While there were clear differences between exposures with and
without the fast guiding system in poor seeing, it was clear
that good seeing would really be needed to show off this
device's capabilities.
Meanwhile, as the seeing frustrated our ( Benoit's ) testing,
the discussions on how to improve the unit and how to best
use the unit continued. At the same time, we had agreed that
if this device worked we would make a presentation on its
workings and results at the July 12-13 SBIG Imaging Conference.
As the deadline neared, the frustration level mounted over the
winter's bad seeing. We had images that showed that the device
worked, but not enough experience to really know how well it
could work.
Finally, the California weather fronts subsided and in June 1996
a weekend of good seeing in the So.Cal mountains produced some
spectacular results.
Experiences with this device continues to provide information
on how to better use it for best results. A second and third
prototype unit were soon built. During 1996,97 Benoit continues to adjust
and tweak the driver software as Brad and the SBIG folk test the
now prototypes and the developing SBIG AO software .
- At this point, it is clear that a fast guiding system operating
at 5-30+ Hz is capable of producing dramatic improvements of 2X to 3X in
imaging resolution capabilities on small ( 8-24 inches) aperture
instruments. ( The physics of the problem changes as the aperture
increases above 24-inches and a fast guider will probably not
work on scopes of larger apertures ).
- It is clear that such
a unit is not a cure-all. In order for such a unit to provide
the best results, one needs to work under GOOD SEEING conditions.
- And as resolution improves, the matter of big/small pixels rises
to greater import. With a resolution limit of 24.3/A, a C-8 and
an ST-7 ( 9 um pixels ) were a good match but as the 'limit'
passes 10./A then a C-11 and an ST-7 become a good match and
it will require a film like 2415 to take advantage of this
fast guider system on smaller scopes.
Where this device will go will take time to see. There are
various ways to build such a device and optimal configurations
could be developed for smaller and bigger instruments.
AO has come to the amateur community !!
SBIG has confirmed its interest in the potentials of this device
and is evaluating it as a potential commercial product.
Update, Jan 1997: SBIG has built a prototype unit and it is now in testing.
Testing of the new SBIG AO prototype has gone on throughout the summer months of 1997 and has been educational to all of us as we worked to fine tune the capabilites of the new SBIG unit and the new SBIG AO software.
Some NEW IMAGES from the 1997 Season
Update, Sept 1997: SBIG has announced the availability of the AO product.
Update, Jan 1998: The SBIG AO-7 is now in PRODUCTION. A number of units are ow out in the
hands of other amateurs and their images and feedback will begin to guide the future development
of this imaging accessory. Some of the first images with the early commercial units came
from Brad Ehrhorn of Incline Village, CA. Other new AO images soon
followed from Adrian Catterall of the UK.
We look forward to seeing more from AO-7 users as we learn more on the capabilities of this tool
and hopefully learn how to extend its capabilities even move!
All images shown here are Copyright 1996,1997 Benoit Schillings and Brad D Wallis
and may not be used in any commercial applications without
their written consent
Return to the Wallis and Provin Page.
Return to Brad Wallis' Home Page.
Return to the AO Page.