EFFECT OF TEMPERATURE ON NOISE LEVELS

FOR AN UNMODIFIED TOUCAM

To test the effect of temperature on the noise level frames were recorded using a Toucam Pro with the lens cap on at room temperature and after 10minutes in a domestic freezer.

Settings were chosen to give a roughly symmetric noise histogram at room temperature without truncation at the low value end (This is difficult to achieve in practise with the Toucam as the brightness does not behave as normal and interacts strongly with the other controls leading to truncation of low values at normal settings. This is not the case for the Vesta Pro which behaves as one would expect with the brightness affecting just the bias level) The values chosen were brightness and gamma 100%, gain 78%, shutter 1/25sec, frame rate 5fps. (A gamma of 0 would have been preferable as this gives best linearity but the histogram was severely truncated at the low end with this setting. Clearly the gamma setting affects the zero bias as well)

The mean level drops significantly at the lower temperature but this appears to be mainly just a depression of the zero level, not an actual reduction in noise, as is apparent from the width of the histogram. Note that the colour shifted as the zero for three colours were depressed by different amounts.