>Actually you are not limited to the 'fairly linear' part of
>the scan - Nikon and I guess other manufacturers linearise
>their output so that they can use more of the scan (and also
>because even the 'fairly linear' part of the scan isn't
>really sufficiently linear.)
>
> Guy
>
Hi again,
I agree. No part of a sine wave is really linear. Print one out and
then try to run a ruler along it... You may decide that accepting 30%
of the horizontal space (time) as being relatively linear is really
quite generous.
One can linearize more of it to some extent, but again there are
limits. One usually linearizes to preserve the linear geometry of the
image by varying the pixel clock so that the pixels near the center
of the scan (where the mirror is moving fastest but accelerating
least) are shorter in time compared to those at the edges of the
scan, with the idea that each pixel can then represent the same area
of the specimen.
However, as both the dose to the specimen (bleaching) and the number
of photons emitted (and collected) during a pixel depend on the pixel
time, one begins to have rather non-linear (doughnut-shaped, in 2D)
bleaching if you use too much of the sinusoidal scan pattern. In
addition, the S/N of the image on axis becomes noticeably worse than
that around the periphery (where more photons are collected, making
the Poisson noise less).
In any case, you must also adjust the gain of the signal collection
system (or the range of the digitizer) so that the longer pixels
don't also look brighter. This can be done but quantification becomes
tricky as so many other factors cause the signal to vary between on-
and off-axis (apodization, curvature of field, chromatic
magnification error etc.).
Cheers,
Jim P.
--
**********************************************
Prof. James B. Pawley, Ph. 608-263-3147
Room 223, Zoology Research Building,
FAX 608-265-5315
1117 Johnson Ave., Madison, WI, 53706
[log in to unmask]
3D Microscopy of Living Cells Course, June 13-25, 2009, UBC, Vancouver Canada
Info: http://www.3dcourse.ubc.ca/ Applications due by March 15, 2009
"If it ain't diffraction, it must be statistics." Anon.
|
