Manually
Adding a Camera for use with SkyTools 3
SkyTools
ships with a list of cameras to choose from, but the list is not
exhaustive. We have included all of the cameras that we
could find complete data for and are always adding more.
Unfortunately some manufacturers are less forthcoming with their
data than others. This guide is intended to help you find
and enter the data you need to set up your camera manually for
use with SkyTools. First
determine what type of camera you have. SkyTools supports
traditional CCD cameras, consumer digital cameras, and video/webcams.
It is important to make the correct selection when you enter the
data because each type of camera has its own specialized capabilities
and limitations. A
camera is made up of two basic parts: the CCD detector (or
sensor or "chip") and the electronics in the camera
itself that read, process and store the images. Two cameras
that share the same detector will also share many of the same
specs, so finding out the model number of the CCD detector is a
very good start. The detector itself usually determines
the following:
In
many cases the following parameters are also determined by the
detector, but they are sometimes modified by the camera design:
Basic
Camera Data Start
with the specs for the camera from the manufacturer. You
can often find this information on the manufacturer's web
site. If not, try web searches on the camera model number
plus "specifications", "specs",
"sensor", "pixel size", "read
noise", etc. One really important piece information
to find is the make and model number of the detector (or sensor)
used in the camera. With that you can often find much of
what you need from the detector spec sheet from the maker of the
detector. As a last resort you might contact the camera
manufacturer. Remember, some of the camera specs may
differ from those of the detector alone. Always use the
camera specs if they differ from those of the detector. CCD
Detector/Sensor Data Once
you determine the model number of the detector (sensor,
"chip") used by your camera you can often find a spec
sheet from the manufacturer of the detector by searching on the
model number. An example model number is
"ICX418AKL". Sometimes there is a dash or
space (e.g. it may be "ICX-418AKL") . Make sure
and try different combinations in your search. Google will
often see a dash as a wild card character so it doesn't hurt to
try putting in some dashes. The trailing letters (AKL in
our example) often indicate variations in the model, such as
color vs. mono, so try to match the exact model number used in
your camera. If
you are lucky we will already have the detector/sensor in
SkyTools. If so, you will be able to select the detector
from the pull-down "Detector" menu on the camera data
dialog. This will set the spectral response (quantum
efficiency at different wavelengths). Be aware that the
presence of a glass window in front of the detector will change
the spectral response, so be careful to match the model number
exactly. In some cases the camera may place a glass window
or filter in front of the detector. If that happens this
too will affect the spectral response. In this case you
will need to know the spectral response of the camera/detector
system (together) rather than the detector alone. Camera
Data in Detail Below
we will look at each of the specifications you must enter for
your camera in detail.
-
Pixel
size -- although often supplied in the specs for the
camera the pixel size is determined by the detector, so you
may use the values given in the specs for either the camera
or detector. In some cases the pixels are square so
only one size is given. In this case enter the same
value in both boxes.
-
Number
of Effective Pixels -- this together with the pixel size
is going to be used to compute the physical dimensions of
the detector, which in turn will determine your Field of
View (FOV).
Take care to get the "effective" number of
pixels. A detector is often manufactured with pixels
that are masked and thus don't contribute to the final
image. In some cases the camera design may also reduce
the number of "effective" pixels returned by the
camera. The best way to find the number of effective
pixels is to simply measure the usable width and height (in
pixels) of an image obtained with the camera. Note:
the width of the detector in mm is the number of pixels in
width multiplied by the pixel size (in microns) divided by 1000.
If you cannot find the pixel size, but know the physical
dimensions of the detector, it is possible to use this
formula to compute the pixel size: width (in mm) / (number
of pixels in width) * 1000. But use this as a last
resort because it is not as accurate as knowing the exact
pixel size.
-
Read
Noise -- this is a property of the detector. It is
often cited in the specs for the camera but can also be
found in the specs for the detector. Together with the
dark signal the read noise is used to estimate the noise in
the final image. This in turn affects the SNR
computed.
-
Dark
Signal -- this is a property of the detector. It
is often cited in the specs for the camera but can also be
found in the specs for the detector. Note that the
dark signal depends on the temperature of the detector, so
this value will change if the camera is cooled (it will be
lower). Make sure to note the temperature conditions
for the quoted dark signal, if available. Together with the
read noise the dark signal is used to estimate the noise in
the final image. This in turn effects the SNR.
computed. This
value is important for digital cameras because the optimum
sub-exposure time is usually determined by the dark signal
rather than the sky. Unfortunately it can be difficult
to obtain the dark signal for many digital cameras.
Typical values for modern Prosumer digital cameras fall
between 2 and 20 electrons/sec.
-
Gain
-- this is the amplification of the signal by the
electronics in the camera expressed in e/ADU. It is particularly tricky,
especially for digital and video cameras where the gain may
be varied by the user. For digital cameras the gain is
set via the ISO (or film speed) selection. If
possible, find out the ISO setting used for the quoted
gain. Larger ISO settings will decrease the
gain. If all else fails, use 1.0. For some
astronomical CCD cameras the gain will vary with
binning. Thus there are two gain entries: unbinned and
binned. Most cameras will use the unbinned value so
you can leave the binned value blank (or set it to be the
same as unbinned). The Gain is used for one thing only
by SkyTools: the calculation of the suggested sub-exposure
time when stacking images.
-
Bit
Depth -- this is often a property of the camera rather
than the detector. The camera may return images in
several formats, where each pixel is represented by an
8-bit, 16-bit, or 32-bit integer. For 8-bit images the
value of each pixel can only have a value between 0 and
255. For 16-bit cameras the range is 0-65535.
For 32 bits the range is 0-4294967295 (a very large
number). This is a minor variable used to estimate the
quantization noise and will have only a small effect on the
model. For 12-bit cameras or if you don't know what to
enter use 16 bits.
-
Built-in
Lens -- the focal length of a
non-removable lens. This parameter applies to
video/web cams only. In most cases this field will be
left blank; even for video cameras the lens is usually
removed for astronomical imaging. For a digital camera
without a removable lens add the built-in lens via the Assign
Lenses button on the Select Cameras to use with
Telescope dialog.
Quantum
Efficiency (QE or DQE)
The
quantum efficiency (QE) describes how sensitive the detector used
by the camera is over the light spectrum. Hopefully the
detector your camera uses will be available for preselection from
the list. But watch out for subtle differences between
detectors. Some may have microlenses added and some will
have a window. The QE values will differ in these cases.
The
spectral response determines how much signal is recorded by a
pixel in general and how sensitive it is to different colors of
light. The higher the QE the more sensitive the detector is
and the greater the recorded signal. if the QE is comparably
low in the blue to that in the red, blue sources will require
longer exposures than red ones.
The
QE determines how much signal to expect from astronomical objects
and the sky background, The signal will depend on the color
of the object (or sky) and which filter (if any) is in use.
Ultimately the model sky signal will be used to compute a
suggested sub-exposure time and both model signals together will
be used to compute the SNR. The QE is most important for
traditional CCD cameras meant for long exposures. For video
cameras, which are limited to excellent lunar, solar, and planetary
photography, the QE is much less important.
Usually
the quantum efficiency will appear in graphical form in the
detector spec sheet. It will often be called "spectral
response". The quantum efficiency (in percent) will be
graphed vs. wavelength. You will need to read the values off
the graph to enter into the Camera Data dialog (set the Detector
to Manual Entry). Read the height of the graph at
each of the wavelengths required by the program. SkyTools
uses wavelengths in nm. You may need to convert to other units
depending on those used by the graph. Note that: 1
nm
= 10 (Å) Angstroms = 0.001 microns
A
color camera may have more than one line on the graph. These
lines may overlap (see example below). If so always read the
value from the highest line on the graph.
One
final but important note: the values entered into SkyTools must be
absolute quantum efficiencies (expressed in percent) as opposed to
relative quantum efficiencies. Many camera spec sheets will
graph relative QE instead. To convert your relative QE
values obtained from such a graph you must multiply each one by the
maximum QE of the detector. Unfortunately in these cases it
is often difficult to find the maximum QE of the detector!
This is usually the biggest stumbling block in entering the data
for a given camera. If the detector spec sheet fails to
indicate the maximum QE then you may need to contact the
manufacturer (of the camera or the detector). In some cases
we have had success using online search engines. Search for
the model number of the camera or the model number of the detector
plus "quantum efficiency" or "QE". Often
times the maximum QE will simply be represented as "the"
QE for the detector.
If
you find the data but have trouble interpreting it, please feel
feel to contact me for help.
Example
1: the SBIG ST-402ME
This
is your basic monochromatic astronomical CCD camera. The
graph below is from the SBIG camera spec sheet. It graphs absolute
QE vs. wavelength in nm.
To
read values off the graph we first find the wavelength at the
bottom, in this case 500 nm. Follow the vertical line up to
where it crosses the QE response curve (red line). Now find
the corresponding value on the scale to the left (52%). In
the box marked 500 nm on the SkyTools camera dialog you would
enter "52".
But
sometimes the graphs don't extend far enough in wavelength.
For instance SkyTools wants a QE value for 350 nm, but that's off
the left side of the graph. Sometimes you have to make due
with what you've got. You could enter the same value as we
get for 400 nm (45%). Or extend the graph yourself to the
left by estimating the curve based on the trend in the data.
In this way we can estimate a value of around 35-40%. We'll
call it 38%.
Example
2: MallinCam Color CCD
Our
first example was fairly straight forward. This next one is
much more complicated. Below is the graph found in the
technical specs for the Sony ICX418AKL detector, which is used in
this camera. There are two complications here compared to
the previous example: as a color camera the spectral response is
drawn with separate lines indicating the response for a given
internal "filter" so there is more than one line.
In addition, the relative response (or relative QE) is
plotted on the vertical axis rather than the absolute QE that we
require. You can spot these because they always reach all
the way to the top of the graph. The top of the graph (at
1.0) is the maximum QE for the detector.
We read
the graph much in the same way as in the first example, except
that we always read the value from the highest curve. So at 450 nm
we skip right past the Ye and G curves all the way up to the Cy
curve to read the value. At this intersection we read a value of
0.62 (this is equivalent to 62%).
But now we
have to know the maximum QE for the detector to convert this value
(0.62) to the absolute QE. Sony really doesn't like to share
their QE data so it can be difficult to find the value. In
fact, I was unable to find the source that I used when I created
the spectral response for this camera! Someone may have sent
it to me or I found it in a really obscure location.
Regardless, the maximum QE I have for this camera is 52%.
These video-type cameras often have much lower maximum QE than
standard astronomical cameras such as the one above, which was
approximately 83%.
So we take
our value from the graph at 450 nm of 0.62 and multiply it by 52%
to get 32%. In the box marked 450 nm on the SkyTools camera
dialog you would enter "32".
There are
some assumptions used going in for color cameras which may limit
the reliability of the results, and for a video camera like this
one many variables are, shall we say, "untamed."
The SkyTools exposure model works best for traditional
monochromatic astronomical CCD cameras, but the good news is that
it is these cameras that require the most accuracy from the model
to begin with. Ballpark estimates are usually good enough
for digital camera and video imaging.
|
