The problem you are seeing is probably due to autofluorescence. Cultured cells
tend to have significantly increased autofluorescence. The fluorescence
spectrum of AF is such that the signal is highly correlated in (essentially) all
channels. Since the spectrum is distinct from fluorescein, PE, etc., proper
compensation of these fluors will leave autofluorescent cells as having a
diagonal characteristic.
There is little to be done about this; if you have a channel to spare, you can
do autofluorescence compensation. I.e., if you are measuring only FITC, then
use the PE to compensate the autofluorescence in the FITC channel (as well as
the Cy5PE channel; then you can do 2-color immunofluorescence with AF
compensation). However, if you don't have a spare channel, or if you need to do
PE staining, then you won't get rid of the "diagonal".
mr