On a scanner, the sensor is most likely a
single row of pixels or perhaps two or three rows. The scan head moves
the sensor line across the image to be scanned. A single dust spec then is present as a
column in all rows of pixels as the sensor is moved. In a camera, the sensor is both rows and
columns and all have to be functional. A single dust spec blocks one or
at most a few pixels. The camera processor sends an impulse down
each row and across each column and assembles the image by combining the
information it gets. If a single row or column amplifier ( the electronic
gadget that ‘reads’ the row or column ) is failing, then the entire
row or column for that pixel amplifier is messed up. In the G9 case, it
sounds like only a single color amplifier in a particular row or column is on
the fritz since the entire strip is blue. I’d guess that the fix is replacing
the sensor – don’t know what Canon would charge for that. Cheers, James From: I had a Polaroid 35mm scanner that
produced a rainbow line across my scans at the same location. I contacted
Polaroid and they said send it in. But by that time I had blown enough
air through the scanner that the line disappeared. Just a speck of dust
diffracting the light. But i could never figure out why it was a line and not
just a spot. |