## Info

8.3.4.1 Check the Bias Frames for Noise and Interference

Load the nine bias frames in the bias frame set. Measure the mean pixel value, Spv, and the standard deviation of the pixel value, gpv , in each bias frame. Tabulate the values as shown in Table 8.3.

• Tip: In AIP4Win, measure the mean pixel value and standard deviation using the Pixel Tool. Set the outer radius to 25 pixels, and the sample shape to square. Click Get Statistics and read the mean value. Measure the same part of each bias frame.

Make a median bias frame and save the resulting image as BIAS-MED.FTS.

•Tip: Use the Median Image Tool in AIP4Win to generate the median bias frame.

Examine each of the bias frames and the median bias frame. An ideal bias frame has a salt-and-pepper scattering of pixel values.

The readout noise in an ideal bias should vary randomly about an average value (the bias value) with a Gaussian distribution. The standard deviation of the variation is the readout noise. Apply a linear scaling that strongly accentuates the noise and interference present; then make a histogram of each bias frame. The histogram should show a Gaussian distribution.

•Tip: In AIP4Win, use Enhance, Brightness Scaling to set the linear stretch thresholds just below the lowest pixel value in the image and just above the highest pixel value in it. Examine the image histogram to see if the distribution of pixel values is reasonably similar to Gaussian. On a logarithmic plot, a Gaussian distribution appears as an inverted parabola.

On visual examination, bias frames often show variations from top to bottom and left to right, patterns of various kinds, as well as random and pseudo-random variations in the intensity of the lines and columns. In each of the bias frames, look for the following:

• Wavy patterns with sinusoidal or repetitive structure, either across the columns (relatively high frequency) or down the

Table 8.4 Sample Transfer Curve and Linearity Data

Flat Set