The following files contain sets of .ps plots that show the readout bias-reset noise in the J-camera dawn flat sequence on 6/8/97. Each file has 3 frames, except for the last, which has 4. Scan and frame numbers are on the plots. For each frame there are 3 plots: a histogram of the whole frame, a plot of the row mean over x for each y = 0-255 in the frame, and a contour plot of the frame.
The effect is most clear and quantifiable in the row histograms (the middle plot of each set of three) but one might want to look at the contour plots (the 3rd of the 3) first to get oriented. Darks have been subtracted as indicated in the titles.
The 13 frames extend from the first of the morning series until the last. The first (0940001) row histogram shows sinusoidal ripple with a period of around 32 pixels, amplitude 20 DN or so, superimposed on a roughly sawtooth drop-off (readout bias-reset noise) with the quad frequency (ie, 2X per plot), amplitude about 25 or 30 in the first frame. As the sky brightens the amplitude of the sawtooth grows to around 300 DN, and the time constant shortens, while the sinusoid is roughly constant. Finally for the last few frames (median DN 6300 in the last), the amplitude of the bias reset sawtooth drops again.
| Scan 094 |
| Scan 095 |
| Scan 096 |
| Scan 097 |