Gain measured from Poisson noise statistics shows a strong dependency on the exposure level on some CUO cameras. For the W19-(0,0) , the relation found is plotted in figure 2, indicating a deviation from linearity of about 5%.
A more precise measurement of linearity is reached by measuring the ADU level versus exposure time, using a stable light source. Measurements made in this way are plotted in figure 3. By dividing the counts with the exposure time, corrected for shutter delay, no linearity deviation above the detection limit of 0.2% is found from 10ADU to 63000ADU in the high gain range of amplifier B.
The Poisson noise analysis clearly is wrong. A noise modulation/source must be causing this, but the origin is currently unknown.
From analysis of photon noise statistics and Fe-55 impacts, the following conversion factors were found:
The low gain mode has been modified by NOT staff to a gain
around 1.6 e
/ADU.
Note that there is a discrepany between the Poisson and Fe-55 gain. In this text, the Poisson-derived gain is used for ADU to electron conversion.
At an illumination level of 101.000e-, saturation is evident as a
decrease in noise in a flat field image, while at 90.000e-, no
saturation effects are seen using either amplifier.
As the high-gain is approx. 1.0 e
/ADU, the
entire high-gain dynamic range is below saturation.
Listed below is the highest illumination level where saturation effects have not been seen. ADU count is for low-gain.
In MPP-- mode, full well is about the same or slightly higher.
For amplifier B, MPP--, saturation was not seen at 101.000 e
, or
35000 ADU.
Using binned pixels, the full well is not much larger than for unbinned
pixels. In 2*2 binning, amplifier B, MPP+, 113000 e
was reached
without saturation.
Figure:
Gain versus exposure level measured from noise statistics for
amplifier B in high-gain. The gain apparently changes by 5% over the
dynamic range, but change is shown to be much smaller by the more
precise measurement plotted in figure 3.
Figure 3:
A plot of ADU per second versus total exposure time, corrected for
shutter delay. No significant linearity deviation is detected
over almost four decades in exposure range. Amplifier B in high gain was used.