Image Noises

Virtually every image produced with digital cameras or scanners has noise constituents. This noise appears due to interference in the useful signal of the light-sensitive sensor. You can’t avoid this, but we can evaluate this parameter by calculating the amount of noise. Noise can be either random or correlated, so we have two test steps in this section. Why calculate the noise? Because the signal-to-noise ratio value will show us how well the scanner’s electronics is protected against interference of all kinds.

Random Noise

Refer to our scanner testing methodology for details about how the measurements are taken. The diagram below shows the dependence of the random noise on the reflection power of the grayscale sectors of the KODAK IT8.7/2 Q-60R2 target.

The signal-to-noise ratio, which is the ratio of the median of a grayscale sector to the deviation, should be regarded as “bigger is better? That is, the bigger the SNR, the higher the scanner’s noise tolerance is. The delta SNR parameter is the total of the measurements. The next diagram compares the Scanjet 8200 to scanners I’ve tested before:

As you see, the Scanjet 8200 is worse than the modern scanner models from Epson in this test, but by a small margin.

Correlated Noise

Correlated forms of noise are the most annoying ?they show themselves in the scanned image as regular “patterns?(usually horizontal or vertical stripes). The predominance of such noise greatly reduces the signal-to-noise coefficient, which is calculated as the ratio of the median to the deviation. The calculated coefficients are put into the next diagram:

The Scanjet 8200 is worse than any Epson scanner in terms of resistance to correlated noise, as you can clearly see in the diagram. It means that the electronics of the Scanjet 8200 is less tolerant to interference. In some cases correlated noise is discernable in scans made at the maximum optical resolution. Moreover, there are color aberrations which are visible on high-contrast areas of the target (black text on white background, raster lines and others).