(meteorobs) OT. -Profiling the Sandia Sentinel Camera's vision.

[Chris Peterson]

I'd recommend calibrating your camera astrometrically. One extremely simple 
way to do this is simply to take a bunch of images with the Moon in them 
from your history. Look up the altitude and azimuth of the Moon at each of 
the image times, and fit a scale to it. Azimuth is trivial- you only need to 
find the pixel coordinates of the zenith and rotate around that; altitude 
depends on the lens characteristics. With the Rainbow L163 lens I use, the 
altitude is described by a cubic polynomial. Other fisheye lenses are likely 
to be the same, with slightly different coefficients.

Rigorously converting the Cartesian camera coordinates to accurate altaz 
values (and those into equatorial coordinates) is really the only way to 
determine with accuracy where a meteor is in the sky. Scales are okay for a 
quick check, but not for any serious analysis.

Can't the Sentinel software do most of this for you already? The data that 
Joe Chavez sends out has computed altitude and azimuth values for the meteor 
in each frame.

Chris

*****************************************
Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com


----- Original Message ----- 
From: "Larry" <ycsentinel at att.net>
To: <meteorobs at meteorobs.org>
Sent: Monday, June 15, 2009 4:13 PM
Subject: (meteorobs) OT. -Profiling the Sandia Sentinel Camera's vision.


> The following measurements were found. The methods of arriving at this 
> data
> will be explained below.
>
> The out-of-focus circular fringe area in the Sentinel cameras photographs
> was measured to be 5.8 degrees BELOW true horizon at its first edges.
> (-5.8). The actual fringe in focus could be a -6.3 but cannot be confirmed
> without refocusing a Sentinel camera.
>
> True horizon (-0.9 actual degrees) is located close to the 10 degree 
> circle
> on a linear Altitude scale of my own design fitted to an Azimuth scale and
> made public.
>
> The Azimuth scale I constructed for general public use on Sentinel photo's
> with a later added Altitude scale contains errors in altitude at lower
> elevations due to what I believe is logarithmic compression of the lens.
> This was anticipated at the time of the construction of the linear degree
> circles. At 30 degrees elevation it is quite close, not more than 1.5
> degrees error, measuring no more than 31.5 degrees maximum on a planet or
> bright star at 30 degrees per an older astronomy program called Skyglobe.
> Above 30 degrees the scale is quite precise and is accurate for all 
> intents
> and purposes for a camera aligned to zenith.
>
> At the present time with this public scale fitted to Sentinel images, the
> outside edge portion of the Azimuth scale is measured to be at a minus 0.9
> degrees of true horizon. (-0.9). That bottom portion is scaled 10 degrees 
> as
> a linear scale circle. It was necessary to use this Azimuth-Altitude scale
> to perform the measures as clearly defined fixed points were not available
> in the North-South direction for the Transit.
>
> The transit was leveled at the lower edge portion of the Sentinel camera
> fish eye lens.
>
> Altitude measures were taken with a precision digital level attached to 
> the
> transit as it deviated away from the fish eye lens in new azimuth
> directions. Controlled tilting of the transit was measured for below 
> horizon
> fixed points. The points used were: Camera lens, Roof top edge in S.E., 
> amd
> Roof gutter edge, all located below the true horizon but within the photo
> vision of the camera. The gutter edge is located at the edge of the
> out-of-focus camera perimeter. Local porch lights are slightly below that
> edge and can be seen by the camera occasionally. This suggests -6.3 is 
> very
> close for the maximum below true horizon aperature of the camera.
>
> These measures are very close but not precise due to out of focus targets
> being used to compare to the Sentinel image taken this morning. If this
> information is of some interest or use by others besides myself,  the task
> is made worthwhile. Others may want to make their own measure to confirm 
> or
> argue what I have found to be a workable.estimate.
>
> YCSentinel