7.8 Gain
97
7.8 Gain
As stated previously, each pixel on a CCD is, in effect, a small capacitor storing
electrons liberated by the effect of an incident photon. The value that is actually read
is the number of electrons within the well. However, what is reported is the count.
A pixel can often hold in excess of 100,000 electrons, which is more than a 16-bit
analogue-to-digital converter (ADC) can report. Hence, the number of electrons is
scaled. This scaling factor is the gain G. Your observatory staff should be able to tell
you the gain of the camera you are using, and it should also be recorded in your lab
book and your images as part of the FITS header.
As always, no two cameras are created equal, and there may well be a slight
differential between the specified gain and the actual gain. In an ideal world, the gain
should be the maximum well capacity in electrons divided by the maximum number
the ADC can hold. In reality, the gain is seldom that high. In the next practical we
will determine the gain of your camera.
7.8.1 Practical 5: Measuring Camera Gain
In this practical you will find the gain of your camera. You will need the camera gain
when calculating uncertainty in your observations. If new, the camera may very well
be operating near the manufacturer’s published level, but with age it may move away
from that value. You will be using the skills you have developed during the previous
practicals in this book.
Taking and Reducing the Gain Images
1. With the camera down to temperature and the lens cap on the telescope, take a
series of darks with 1, 2, 3, 4, and 5 s exposures. Take two darks for each exposure
time. Ensure that the binning is set to 1 × 1 and subframing is off. If you have the
dark frame option on your control software, turn it on.
2. Take the lens cap off and put the dome lights on. Set the telescope up for taking
flat frames as previously.
3. Take pairs of flat frames for 1, 2, 3, 4, and 5 s exposures. Start with the 5 s exposures. Using FITS Liberator (or another appropriate package), ensure that you
are not entering the nonlinear region of the CCD, hence the need to start with the
long exposures. If you are entering the nonlinear region, swap to a different filter
and try again.
4. Using an appropriate package (for example MaxImDL or a Python script), generate a mean image of each of the pairs of dark frames and subtract these from
the flats of the same exposure time.
97
7.8 Gain
As stated previously, each pixel on a CCD is, in effect, a small capacitor storing
electrons liberated by the effect of an incident photon. The value that is actually read
is the number of electrons within the well. However, what is reported is the count.
A pixel can often hold in excess of 100,000 electrons, which is more than a 16-bit
analogue-to-digital converter (ADC) can report. Hence, the number of electrons is
scaled. This scaling factor is the gain G. Your observatory staff should be able to tell
you the gain of the camera you are using, and it should also be recorded in your lab
book and your images as part of the FITS header.
As always, no two cameras are created equal, and there may well be a slight
differential between the specified gain and the actual gain. In an ideal world, the gain
should be the maximum well capacity in electrons divided by the maximum number
the ADC can hold. In reality, the gain is seldom that high. In the next practical we
will determine the gain of your camera.
7.8.1 Practical 5: Measuring Camera Gain
In this practical you will find the gain of your camera. You will need the camera gain
when calculating uncertainty in your observations. If new, the camera may very well
be operating near the manufacturer’s published level, but with age it may move away
from that value. You will be using the skills you have developed during the previous
practicals in this book.
Taking and Reducing the Gain Images
1. With the camera down to temperature and the lens cap on the telescope, take a
series of darks with 1, 2, 3, 4, and 5 s exposures. Take two darks for each exposure
time. Ensure that the binning is set to 1 × 1 and subframing is off. If you have the
dark frame option on your control software, turn it on.
2. Take the lens cap off and put the dome lights on. Set the telescope up for taking
flat frames as previously.
3. Take pairs of flat frames for 1, 2, 3, 4, and 5 s exposures. Start with the 5 s exposures. Using FITS Liberator (or another appropriate package), ensure that you
are not entering the nonlinear region of the CCD, hence the need to start with the
long exposures. If you are entering the nonlinear region, swap to a different filter
and try again.
4. Using an appropriate package (for example MaxImDL or a Python script), generate a mean image of each of the pairs of dark frames and subtract these from
the flats of the same exposure time.