[log in to unmask] writes:

> 4.23. Interference Contrast and Confocal

> Interference contrast is a very useful parameter in microscopy and
> it can be combined with fluorescence. However, because the
> microscope system was designed for light to traverse through two
> interference filters, when this optical system is applied to a
> confocal microscope there is distortion in the fluorescence
> signals. The fluorescent light traverses the interference contrast
> filter and excites the sample, and then the emitted fluorescence
> travels back down through the same interference contrast filter and
> back through the scan head.  The resulting image shows a duplication
> of very small particles (0.17 μ m, PSF beads) and a distortion of
> larger particles. PSF beads show two spots and 0.5 μm beads show an
> egg shaped image instead of a round image. The same distortion that
> is observed on beads will occur on biological structures in cells (
> see Fig.  15). For optimum resolution of data that will be
> deconvoluted later, it is recommended to remove the interference
> filters when acquiring an image.

On my first reading of this I thought by interference contrast filter
Robert was referring to the polariser. On a second reading I realise
that it refers to the DIC prism. I wrote this extended reply before
realising that we are saying the same thing but I am posting this
anyway as a second description might help people understand what is
going on and why this happens. 

DIC works by sheering the two polarizations relative to each
other with the condenser prism. The beams then pass through slightly
different sections of the sample, and are recombined with the second
(objective) prism. This produces an image of relative phase shift
between the two beams.

In epi-fluorescence the excitation beam passes through the DIC
(objective) prism and is split into two beams, offset relative to one
another. The fluorescence from these two regions is them shifted back
as the emission passes back through the (objective) DIC prism. This
produces a double image shifted by the sheer in the DIC prism. The
sheer tends to be a fraction of the resolution, say 1/3rd but varies
with lens, manufacture etc... In conventional wide field this is
generally not noticeable. On a properly set up confocal this leads to
a pronounced broadening of the PSF in the sheer direction, at 45
degrees to the x and y sample axis.

As Robert says, the take home message is it is best to remove any DIC
optics before taking confocal images.

Ian