Re: Modelling - can of worms


Wiggins, Rick
 

Hi,
I don't want to get into a large debate, but would like to point out
what I believe, for some people, may be a misconception regarding
the definitions of modeling vs. polar alignment.

1. Polar alignment of the mount involves physically aligning the
mount with the polar aixs. This can only be achieved by physically
adjusting the azimuth and altitude of the mount. Any variation from
perfect alignment means that the FOV will rotate relative to the
polar axis by an amount proportional to the polar alignment error
(commonaly called drift by imagers). The image train non-
orthogonality with respect to the polar axis will also drive the
drift error to be higher proportiuonal to the non-orthogonality.
There are various programs such as PEMPro and PoleAlignMax that can
calculate and provide quantitative feedback of the polar alignment
error to the user; however, the user still has to manually move the
mount in azimuth and altitude to better perform polar alignment.

2. Modeling is a common term given to algorithms that calculate
various error terms. Modeling programs "subtract" these error terms
from the non-modeled calculated position of targets to allow the
mount to correctly position the telescope so that the target is
centered in the FOV. These modeling programs can compensate for
refraction, non-orthogonality of the scope relative to the mount,
and polar mis-alignment as well as other terms. They perform this
compensation for the pointing or initial pointing placement of the
telescope on the target. They will do an excellent job of having the
telescope go exactly to the center of the desired FOV following a
slew.

Understandig the brief definitions above allows one to understand
how these terms come into play while imaging. Although the modeling
program will properly position the scope following a slew; once the
tracking begins, the FOV will rotate relative to the tracking star
based on the degree of polar mis-alignment. The modeling will not
help this. In addition, if polar alignment is significantly off, it
will also affect the ability of the guider to correct all the error
terms including periodic error and polar mis-alignment. This is why
I say that modeling is separate from polar alignment. It certainly
can help with pointing, but has no effect on guiding and tracking.

I hope this helps clarify the issue.
Thanks, Rick

--- In ap-gto@..., "Ladislav Nemec" <nemecl@...> wrote:

Good summary of essential conditions for imaging. I just somewhat
disagree
with the statement that 'modeling has nothing to do with polar
alignment'.
It depends on the modeling algorithm. Gemini modeling compensates
for polar
misalignment quite well - the question is, of course, whether
modeling
should substitute for a good polar alignment. Not for permanent
installation
but it may be handy for transient installation - building a model
that
includes compensation for minor polar misalignment may be faster
than very
precise polar alignment in the field.



Also, I still have to achieve significant reduction of tracking
errors (some
of them non-periodic on my Losmandy G11 and thus not possible to
correct by
PEC) that, at this time, are typically less than 8 arcsecs peak-
peak. AP
mounts, perhaps, do not have these non-periodic errors - they are
created
(general belief) by mechanical imprecisions.







-----Original Message-----
From: ap-gto@... [mailto:ap-gto@...] On
Behalf Of
Rick Wiggins
Sent: Tuesday, September 02, 2008 10:37 AM
To: ap-gto@...
Subject: [ap-gto] Re: Modelling - can of worms



Hi,
I would like to chime-in based on extensive use of the AP mounts
in
both observatory and portable setups (ap 400, 600, 900, 1200, &
Mach1.
A. First GPS: There is no real need for GPS. You can get location
coordinates from any map or look them up on the internet ahead of
time. Once programmed in (less than one minute to program during
setup even in the dark), they can be called up in less than 1 sec
during the startup routine. Also, the exact location is not
critical
to imaging.
B. Modeling: Modeling is great for very precise visual or imaging
setup work. It is not necessary for imaging as long as the mount
can
get you within the FOV of the camera. Doing plate solves is far
more
accurate than modeling for centering an image. If you are using
something smaller than an ST-10 and focal lenght over 3000mm,
modeling may be a real benefit. Modeling has nothing to do with
polar alignment. I do not use modeling with my AP mounts when
doing
fully remote imaging. I only use plate solves. With an AP1200
mount,
1200mm FL scope and ST-10 (app 30 x 45 arc mins), my target is
always better than 15 arc mins from center (prior to plate solve)
over an entire nights imaging session of multiple targets ranging
across the entire sky and including multiple meridian flips.
C. Polar alignment: The AP mounts align very well with the polar
scope. I typically get mine better then 10 arc mins with the polar
scope. That is good enough for 20 min subframes at 1200mm FL on a
24mm x 36mm chip or camera. If you need better alignemnt, Pole-
Align_Max, PEMPro, or drift alignment can get you below 30 arc
secs
with a time investment of 30 to 60 minutes.
D. Tracking for imaging: Here is a list of factors using my AP
mounts in descending order of importance (factors affecting
tracing
performance). These are based on monitoring of my tracking logs:

1. Polar alignment with the polar scope. I find this takes less
than
5 minutes for portable use assuming mount has been prealigned with
compass or better during daytime.
2. Exception rigid mechanical setup (based on separate guidescope
and N/A in OAG or internal guiding) with no snagging cables. This
is
a one time setup for each optical configuration if done correctly.
3. Very well adjusted (dialed-in) guider parameters (min/max
corrections, aggressiveness, etc.) Note: McMillain's paper is the
best guide I have seen. This is a one time setup for each optical
configuration.
4. PEMPro or similar PEC corrections. This will reduce AP mounts
from factory performance (I have found between 5 to 2 arc secs PE)
to below 1 arc sec PE. This is basically a one time setup,
although
it should probably be checked yearly or if tracking changes.
5. PEMPro, PoleAlignMax, drift align, or similar polar alignemtn
refinement. This is done during each setup if required.

I hope this data provides some help in understanding imaging
capabilities regarding tracking and alignment with AP mounts.
Thanks, Rick

--- In ap-gto@yahoogroups. <mailto:ap-gto%40yahoogroups.com> com,
Gerald
Sargent <sargentg@> wrote:

In my initial submission starting the modelling topic, I should
perhaps
have noted that I am one of those who has to set up each time,
hence
the need for rapid polar alignment, and I mainly image, using
either an
SBIG with self guide or the STV for autoguiding, both of which
work well,
However both need that the mount be reasonably aligned. My
Gemini on a Losmandy G8 had an excellent routine for getting
aligned
using 3 stars, and it had an inbuilt GPS feature which enabled
me
to
download time/location from an external GPS receiver. Both these
are
great time savers and one could be reasonably well aligned with
the
mount knowing exactly where it was and what time it was in a
matter
of a few minutes.
With my next mount, a Vixen Atlux, a really excellent mount I
had the SS2KPC controller which did not have the GPS facility
which
meant a few minutes "lost", it did however have a very good polar
alignment facility using three stars and one could be very
quickly
sufficiently well aligned for imaging.
One gets used to taking advantages of advanced routines,
image trying to use a car in a city without a self starter - the
crank
handle every time - well that is how I feel using the Mach1 with
no
modelling or GPS facility, sure half an hour or so and I can get
it
aligned but with a lot of "crank handle effort" that I have not
had to
do with my last two mounts. The Mach1 tracks superby once
aligned,
and it is a perfect mount for imaging - but it does lack
the "self
starter"
Gerald.





Join main@ap-gto.groups.io to automatically receive all group messages.