We have been working with thick (living) specimens. In these samples we
lose confocal signal as a function of focal depth. We believe this is due
to light scatter, and when looking at SNARF fluorescence we lose both the
580 and 640 emission wavelengths in parallel (the emission ratio stays
constant as a function of focal distance into tissue). The details are in
an article that came out in Am. J. Physiol (cell physiol) last December
(Chu, Brownell and Montrose are the authors but I can't remember any more
detailed citation stuff).
 
FINALLY, here is my question. Can the point spread function be determined
accurately enough to account for this assymetric loss of signal during
deconvolution? Has anyone tried? Is there any hint from the accuracy of
the PSF determinations (we lose about 3% of the signal per micrometer
depth)? It strikes me that without accurate accounting of a known signal
loss, your deconvolution in this case would be doomed to failure.
 
 
Chip Montrose
Johns Hopkins University