Re: Autoguider correction frequency



The results with the little Orion SSAG and the AP900 are so good that of course all it has done is to show up faults elsewhere in my system. Ten minute exposures have produced tack-sharp stars - but also revealed dust donuts on the images that were never there on the 60 second efforts.

Still - that is what this sport is all about - ratcheting up your system inch by dollar. Good stuff.


Bob Piatek <bobtek@...> wrote: Murray,

On your first question, I think the telescope mount manufacturers are
doing what they do best. I've owned several AP mounts and can't think
of a lot of ways they could improve. Maybe going to a USB interface
instead of the older Serial ports would be my top suggestion.
However, I think there are plenty of opportunities for advancing the
state of the art in auto-guide camera software.

Averaging several short guide star exposures is one of them.

Another technique that I heard of people trying was using multiple
guide stars to calculate position. The idea is that seeing related
changes would be seen differently by the various guide stars in the
field. I haven't seen much data on the results achieved by doing
this, however. My feeling is that small format sensors like that used
in guide cameras would have all stars in the image seeing the same

I did some work myself on something I call predictive guiding. The
idea is that you would collect historical guide star error data over a
period and then look for trends. For star drift due to polar miss-
alignment, for instance, you would typically see a fixed error in a
particular direction. Rather than waiting for one entire guide period
and then issue a single large correction, the software would portion
the same total correction via many smaller corrections. The idea is
to make smaller corrections but more frequent. The results I found
were significant.

What we need is to see more of the software developers out there
devoting more time to the guide problem. It seems to me that most of
their efforts are devoted to the image processing problem.

For your second question... you can certainly use the SSAG as a
primary imaging camera but there will be problems. First is the fact
that the SSAG is limited to only 8 bit data. This will not be enough
for many deep sky objects that need a camera with better dynamic range
to capture the fainter regions of the object. Secondly, the CMOS
image sensor in the SSAG suffers from line noise artifacts that will
limit its performance as an imaging camera. The Starfish design uses
the entire 10 bit pixel data captured by the sensor and has dedicated
image processing hardware to remove the dynamic line noise. So, you
can use it as a primary imaging camera for some of the brighter
objects but you really have to coax the image out of your raw data
during post processing to get good results. You'll make your life as
an astro-imager a lot easier by using a true 16bit camera for your


Bob PIatek

On Mar 31, 2008, at 11:38 AM, Murray Hammick wrote:
> Bob,
> Two questions - and thanks for your comments. I am the guilty one
> for setting this hare running.
> Would not the the best of both worlds be a system that allows you to
> select both the exposure time and the frequency of exposure. That
> will give you the chance to minimise loss of definition of brighter
> stars, but also avoid over-correcting the mount. Alternatively, the
> mount makers could allow you to choose settings to filter
> corrections and eliminate very small inputs attributable to poor
> seeing and other "noisy" variables.
> The other question is whether the Orion SSAG and similar cameras can
> be used as a "CCD" imager. The .05sec exposure setting indicates
> that it might be a very useful CCD subsitute - even if it is
> actually a CMOS device as someone pointed out on this thread.
> Murray

Join to automatically receive all group messages.