Slide Scanner Repair Manual Diagnostics and Troubleshooting
149
Dark Noise II Test
The dark noise test analyses the statistical and fourier noise patterns in a scanned image of
the dark calibration target in the scanner. The dark calibration target is the back of the
scanner’s carriage. For the purposes of the test, the system is calibrated to a dark set point
of 7 dn for the SS35 Plus (4 dn for the SS35ES) in the units of the A/D converter. The ideal
expectation is that a fully corrected scan of the same target will yield a uniform response at the
dark set point, with some small amount of noise. The individual components of the test are
described in the following paragraphs, in the context of the stand-alone version, called Noise
IIw power subvi. The integrated version performs the same tests.
Dark Uniformity Response (Streaks)
Part 1 of the test constructs a column average response for each color channel of each
pixel in the tested area of interest.
In the scanner functional tester, the column average is done on a 64 line sample.
In the stand-alone version, this fixed pattern dark response is plotted vs. pixel number as
shown in Figure 5-15.
Note: The full scanner tester only checks 1000 pixels of the 4000 pixel CCD, but the
stand-alone version can be programmed to test a larger number of pixels.
Figure 5-15. Fixed dark pattern response vs. pixel number
Based on this data, the test checks that the Mean Fixed Dark Response, or the average of
the dark response for all the dark data tested, is within a small range of the targeted dark set
point. The data tested should have been calibrated to exactly the targeted dark set point.
Large deviations are indicative of an optical or electronics problem (or firmware operation).
This is the Mean Fixed Dark - Dark Set Point test. There are different tolerances for
SprintScan 35ES and SprintScan Plus because the SS35ES is tested at 10-bits and the SS35
Plus is tested at 12-bits. The test generates three results, one for the red, green and blue. All
data and tolerances are done in units of the A/D converters least significant bits.
To Next Page
Loading...
Need help?
General
