7.5 Calibration Frames
85
this, we take a dark frame. A dark frame is an exposure, with the shutter closed, of a
duration equal to the light frame. The dark frame, as with the bias frame, is subtracted
from the light frame to remove any inherent problem with individual pixels. Dark
frames should be fairly linear, i.e., if you double the exposure time, you should double
the count within the pixel. You also should by now have noticed that a dark frame
must also include a bias frame, so in general, the removal of the dark also removes
the bias, but not always. Again it depends on the software you are using. You should
also notice that a dark must be linear only after the nonlinear bias has been removed.
Again, as with bias frames, dark frames should be taken at the same temperature as
the light frame. Typically, they are taken either immediately after or before the light
frame is taken, and in many cases, the software you are using
Not all pixels are created equal. Even on the best science-grade array, some are
more sensitive, some less so. The hole in the secondary mirror of a Cassegrain
telescope, the secondary, dust on the filters, the correcting plate, or the CCD window
and nonuniform illumination of the chip due to the telescope aperture being circular
and the array being rectangular all need to be accounted for. This is done by taking
a flat frame, which, unlike biases and darks, are used to divide the light frames.
Flat frames require that the camera be exposed to a uniform level of light until the
median pixel count reaches a set limit (for most 16-bit cameras this will be around
30,000 counts). This has to be done for every filter mounted on the camera. Normally,
multiple flat fields will be taken, and their median value for each pixel used to create
a science flat. Typically, flats are taken during twilight using a position opposite
the Sun, as that is considered the flattest (i.e., most uniform) part of the sky, with
multiple flats taken and their mean value used as the master flat for that filter. Flat
frames taken in this manner are known as sky flats. Very good flats are extremely
challenging to achieve, and because they account for the contamination of the optics,
they need to be updated much more often than dark or bias frames. Hence, there
is now a tendency to do dome flats. Dome flats use the illuminated inside of the
telescope dome as their light source, rather than the sky. Dome flats may be done at
any time and are more consistent, because the environment in which they are taken
can be controlled. However, they do not cover the entire optical pathway (because
they don’t include the atmosphere), and they may not be as uniform as sky flats.
There is currently considerable debate about sky flats versus dome flats, the methods
used to produce skies and domes, and their impact on many aspects of observational
astronomy, as will be discussed later.
When an image of an object is taken, it is known as a light frame. You will need
to subtract the bias and dark frames from your light frame and divide the result by
your flat frame (minus its bias and dark component) to create a science frame. Many
professional astronomical telescopes will do this as part of the pipeline, a series of
computer programs that together turn an astronomical image into a usable science
image (Figs. 7.4, 7.5, and 7.6).
85
this, we take a dark frame. A dark frame is an exposure, with the shutter closed, of a
duration equal to the light frame. The dark frame, as with the bias frame, is subtracted
from the light frame to remove any inherent problem with individual pixels. Dark
frames should be fairly linear, i.e., if you double the exposure time, you should double
the count within the pixel. You also should by now have noticed that a dark frame
must also include a bias frame, so in general, the removal of the dark also removes
the bias, but not always. Again it depends on the software you are using. You should
also notice that a dark must be linear only after the nonlinear bias has been removed.
Again, as with bias frames, dark frames should be taken at the same temperature as
the light frame. Typically, they are taken either immediately after or before the light
frame is taken, and in many cases, the software you are using
Not all pixels are created equal. Even on the best science-grade array, some are
more sensitive, some less so. The hole in the secondary mirror of a Cassegrain
telescope, the secondary, dust on the filters, the correcting plate, or the CCD window
and nonuniform illumination of the chip due to the telescope aperture being circular
and the array being rectangular all need to be accounted for. This is done by taking
a flat frame, which, unlike biases and darks, are used to divide the light frames.
Flat frames require that the camera be exposed to a uniform level of light until the
median pixel count reaches a set limit (for most 16-bit cameras this will be around
30,000 counts). This has to be done for every filter mounted on the camera. Normally,
multiple flat fields will be taken, and their median value for each pixel used to create
a science flat. Typically, flats are taken during twilight using a position opposite
the Sun, as that is considered the flattest (i.e., most uniform) part of the sky, with
multiple flats taken and their mean value used as the master flat for that filter. Flat
frames taken in this manner are known as sky flats. Very good flats are extremely
challenging to achieve, and because they account for the contamination of the optics,
they need to be updated much more often than dark or bias frames. Hence, there
is now a tendency to do dome flats. Dome flats use the illuminated inside of the
telescope dome as their light source, rather than the sky. Dome flats may be done at
any time and are more consistent, because the environment in which they are taken
can be controlled. However, they do not cover the entire optical pathway (because
they don’t include the atmosphere), and they may not be as uniform as sky flats.
There is currently considerable debate about sky flats versus dome flats, the methods
used to produce skies and domes, and their impact on many aspects of observational
astronomy, as will be discussed later.
When an image of an object is taken, it is known as a light frame. You will need
to subtract the bias and dark frames from your light frame and divide the result by
your flat frame (minus its bias and dark component) to create a science frame. Many
professional astronomical telescopes will do this as part of the pipeline, a series of
computer programs that together turn an astronomical image into a usable science
image (Figs. 7.4, 7.5, and 7.6).