Gonzalez: count rate vs. counts
Lee: count rate vs. counts
Comparison of linearity test data taken by J. Lee et al. on 2007 December 5 with data taken by A. Gonzalez et al. on 2007 November 22. The observing sequences were similar; the one used for the Gonzalez data is described in more detail on here.
The Lee sequence was only taken with lights on; there is no corresponding darks or lights-off exposures. Since we can only analyze the lights-on data, and I have therefore regenerated similar plots using the Gonzalez lights-on data only. This does not significantly affect the comparison here.
The Gonzalez sequence was obsevred at night during cloudy weather between about 21:15 and 00:22 MST. The Lee sequence was taken during morning daylight hours, from about 06:50 to 07:55.
The first plots show the count rates for averaged images as a function of exposure time, for both the Gonzalez and Lee sequences. Note the relatively smooth decline of the Gonzalez count rates with increasing exposure time (past the "jump" at t=10s, discussed elsewhere), compared to the undulating behavior of the Lee sequence.
|
Gonzalez: count rate vs. counts |
Lee: count rate vs. counts |
For my previous analysis of the Gonzalez sequence, I fit a slope to the counts per sec vs. ADU over some range of count levels. This cannot easily be done for the Lee sequence.
Gonzalez: relative nonlinearity analysis |
Lee: relative nonlinearity analysis |
Here are the detailed plots showing frame-by-frame count rates for the Gonzalez and Lee sequences, arranged along the x-axis in the chronological order in which the data were taken.
Gonzalez lights-on data 
Lee lights-on data |