9.4.1 Conversion factor vs. level

findgain was used to determine the conversion factor (CF) and the read-out noise (RON) for the 34 pairs of images. The bias images dfsc1922 and 1925 (using the same section: [1:1060,300:700]) were used. These are from night 6 (whereas the dome flats used were from night 8). The bias images from night 6 are the ones which resembles the mean bias image the most.

findgain was used at the section [480:0600,160:280] (center of CCD) and [910:1030,080:200] (right hand side of CCD). It was suspected, that the CCD would behave strange in the right hand side. The result is shown in Figure . It it seen that apparently CF increases with level. However, this can not be the case since the CCD found to be linear in the direct test, cf. Sect.  below. More probable a malfunction in the read-out electronics caused RON to vary with the exposure level.

 

A direct way to calculate the RON in ADU is the fact, that the standard deviation on a bias image is the RON in ADU. The standard deviation of 5 raw bias images (dfsc1921-1925) was found to be 2.25 ADU (using imstat with an upper limit of 150 to cut cosmic-ray-events).

If we use only flat field images with levels above 1000 ADU and without severe cosmic-ray-events, we get CF = 1.95 e^-/ADU and RON = 2.25 ADU. These are the values that have been used hereafter.