108
8 Imaging
Whether you are taking flats by dome or by twilight, the two processes are broadly
the same. In the case of sky flats, you will be imaging the part of the sky directly
opposite the Sun (which will have set), as this is considered the flattest. If it is cloudy
or foggy in the area of the sky where you intend to take the flats, then you will need
to do dome flats. If you intend to do dome flats, follow the procedures outlined by
the observatory staff, which might require pointing at a particular part of the dome
or using special lighting. You should take dome flats for every filter and binning you
intend to use and, if possible, as many of each as you can in the time you have. A
good flat fills each pixel up to about half its full well capacity, about 30,000 counts
for a 16-bit camera. Your observatory staff should have some idea of what exposure
time you need for each binning and filter to achieve this (note that these times are
also camera and telescope dependent). If you do not have this information, it will
be a matter of trial and error to find the correct exposure time. Do not keep frames
whose exposure times were too short or too long, as that leads to errors. You need
to be only at around half the full well capacity. It does not matter if you are a few
thousand counts out, as long as the exposure time for each set is consistent when you
are taking them. If you are taking sky flats and you can see stars in the frame, the
sky has become too dark for flat fielding. If your flat exposure time is short, perhaps
under 5 s, it means your source is too bright, and you might have errors caused by
the shutter in the flat field. A neutral density filter or neutral density film placed in
the optical pathway can help to increase the exposure time without affecting the flat.
Once you have calibration frames, you will need to create master frames. These
are calibration frames produced by taking the mean pixel value of a number of
calibration frames. This is done in order to reduce noise that will likely be present
in single frames. Depending on the level of error you are willing to tolerate, you can
use single frames as master frames; this is especially true if you do not intend to do
photometry.
In order to obtain your master bias calibration frames, you will need to find the
mean of the frames; this is a pixel-to-pixel mean and not the mean across the frame.
Be very careful when making masters. All the frames used to make the master must
be of the same size, i.e., have the same binning (for dark, bias, and flats), the same
exposure time (for dark frames), and the same filter (for light frames). If you do
not have the same binning level, the process will usually fail, but such is not the
case with filters and exposure time. If you make a mistake at this point and it goes
unnoticed, it will affect all your images. Most image processing software has some
form of image mathematics. If yours does not, then it is reasonably straightforward to
create master frames using Python, although you will need the PyFITS and Numpy
modules installed.
To turn your bias frames into master bias frames, pixel to pixel, co-add all the
bias frames of the binning and then divide every pixel by the number of co-added
frames.
8 Imaging
Whether you are taking flats by dome or by twilight, the two processes are broadly
the same. In the case of sky flats, you will be imaging the part of the sky directly
opposite the Sun (which will have set), as this is considered the flattest. If it is cloudy
or foggy in the area of the sky where you intend to take the flats, then you will need
to do dome flats. If you intend to do dome flats, follow the procedures outlined by
the observatory staff, which might require pointing at a particular part of the dome
or using special lighting. You should take dome flats for every filter and binning you
intend to use and, if possible, as many of each as you can in the time you have. A
good flat fills each pixel up to about half its full well capacity, about 30,000 counts
for a 16-bit camera. Your observatory staff should have some idea of what exposure
time you need for each binning and filter to achieve this (note that these times are
also camera and telescope dependent). If you do not have this information, it will
be a matter of trial and error to find the correct exposure time. Do not keep frames
whose exposure times were too short or too long, as that leads to errors. You need
to be only at around half the full well capacity. It does not matter if you are a few
thousand counts out, as long as the exposure time for each set is consistent when you
are taking them. If you are taking sky flats and you can see stars in the frame, the
sky has become too dark for flat fielding. If your flat exposure time is short, perhaps
under 5 s, it means your source is too bright, and you might have errors caused by
the shutter in the flat field. A neutral density filter or neutral density film placed in
the optical pathway can help to increase the exposure time without affecting the flat.
Once you have calibration frames, you will need to create master frames. These
are calibration frames produced by taking the mean pixel value of a number of
calibration frames. This is done in order to reduce noise that will likely be present
in single frames. Depending on the level of error you are willing to tolerate, you can
use single frames as master frames; this is especially true if you do not intend to do
photometry.
In order to obtain your master bias calibration frames, you will need to find the
mean of the frames; this is a pixel-to-pixel mean and not the mean across the frame.
Be very careful when making masters. All the frames used to make the master must
be of the same size, i.e., have the same binning (for dark, bias, and flats), the same
exposure time (for dark frames), and the same filter (for light frames). If you do
not have the same binning level, the process will usually fail, but such is not the
case with filters and exposure time. If you make a mistake at this point and it goes
unnoticed, it will affect all your images. Most image processing software has some
form of image mathematics. If yours does not, then it is reasonably straightforward to
create master frames using Python, although you will need the PyFITS and Numpy
modules installed.
To turn your bias frames into master bias frames, pixel to pixel, co-add all the
bias frames of the binning and then divide every pixel by the number of co-added
frames.