Re: DIC condensers

Posted by James Pawley on
URL: http://confocal-microscopy-list.275.s1.nabble.com/DIC-condensers-tp7583008p7583016.html

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Dear Jeff,

I agree with Guy about the Kohler but there is
also the matter of wavelength. DIC prisms do work
with "white" light but they are usually designed
to work best at a particular wavelength (often
the green line of the Hg source, 546nm). I don't
know the emission spectrum of your LED (or how
hot you have your tungsten-halogen, as this
affects it spectrum, IR filter? UV filter?) and I
don't know the response of your detector, but if
seeing the greatest possible contrast is
important, you might optimize the wavelength of
the operation as well as being sure that Kohler
is working properly

Another thing to keep in mind is that DIC depends
on keeping two light paths separate by having
their polarization be at 90 degrees to each
other. Problems can stem from the fact that the
degree to which light is reflected when passing
through a glass-air surface varies with the pol
axis. This is not a major problem on low-NA
lenses because the incidence angle is near normal
and reflection is low (especially on modern
coated optics). However, on high-NA lenses, rays
near the edge of the aperture impinge at angles
where the differential reflection of the two ray
bundles becomes significant. This can be seen by
looking at the BFP using the phase telescope
(Bertrand lens?) with only the polarizer and
analyzer in place and adjusted for maximum
extinction (no DIC prisms and a clear glass
specimen, but an oiled condenser open to the same
NA as the objective).

If the problem is significant, you will see what
is called The Maltese Cross, which comes about
because 4 symmetrically located, blurry,
more-or-less circular lighter blobs occur in the
NA regions where particular ray-bundles are
depleted. Of course, this also happens when you
are using the same optics with the DIC prisms in
place.  The result is that the two bundles are no
longer quite independent and this reduces the
contrast. Shinya Inoue developed a corrector lens
that Nikon used to sell to reduce this problem,
but I don't know if it was ever upgraded to the
newer scopes (maybe someone from Nikon can tell
us?). And while we are on the topic of pol, DIC
is best performed using Pol (strain-free)
objectives because strain-birefringence can also
mess up pol performance. I am not clear as to
whether your 1.45 lens meets this requirement.

Finally, if you interested in seeing individual
microtubules, you will need more than eyeballs
and good optics.  MTs are very small and don't
produce much contrast to start off with. They
were initially only made visible by virtue of
video-rate electronic contrast enhancement that
involved not only using a very bright light
source (to increase signal levels and reduce the
relative effect of Poisson Noise in the detector,
and to reduce this further, one could
exponentially average the image electronically
over many video frames. Usually one used a
mercury source, but if you do this, be sure to
include a UV filter (and perhaps an
"interference-green" to isolate the 546nm line)
so you don't damage your polarizer.). The
resulting high "average brightness" was then
electronically subtracted off so that one could
expand the contrast of the signal that remained.

This procedure invariably produced a very blotchy
image. Small imperfections in the optics (dust,
bubbles, scratches) create low contrast and
generally out-of-focus features referred to as
mottle. The contrast of mottle is so low that it
is only visible after the sort of contrast
enhancement just described. What made DIC so
suitable for viewing small features by
video-enhancment was its very narrow depth of
field. By slightly changing the focus, one could
form an image in which the ONLY features recorded
where those due to mottle. This image could then
be averaged, stored and subtracted from the live
image (often this difference image was also
averaged to further reduce noise). The result was
essentially the image contrast caused only by the
"now-in-focus" MT, a feature that might have had
only 1% contrast before enhancement.

Good luck,

Jim Pawley

>Jeff,
>
> If you want to get decent contrast in DIC
>you need a high NA condenser with a high NA
>objective.  Of course the plate must be correct
>for the objective, and Köhler illumination must
>be correctly set up.  My guess is that that
>wasn't true for your LED source, and you weren't
>filling the BFP.  Don't take this personally but
>the widespread use of fluorescence has made many
>microscopists become slack about setting up
>Köhler illumination!
>
> Guy
>
>Guy Cox, Honorary Associate Professor
>School of Medical Sciences
>
>Australian Centre for Microscopy and Microanalysis,
>Madsen, F09, University of Sydney, NSW 2006
>
>-----Original Message-----
>From: Confocal Microscopy List
>[mailto:[hidden email]] On
>Behalf Of Jeff Spector
>Sent: Saturday, 29 November 2014 1:22 PM
>To: [hidden email]
>Subject: DIC condensers
>
>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>Post images on http://www.imgur.com and include the link in your posting.
>*****
>
>Greetings,
>    This isn't exactly a "confocal" questions but
>I know a lot of "micoscopy gurus" live on this
>list so I thought it a good place to ask this. I
>have a colleague who is trying to image
>individual (i.e. small and diffraction
>limited) microtubules in a flow chamber by using
>DIC. They are currently using a 100x 1.45 oil
>Objective, but only a .52 LWD condenser. They
>were using a solid state light source but
>couldn't get good image so we switched to a lamp
>for illumination and the images are much better,
>and we can now see the microtbules but there
>still isn't a lot of contrast.  My question is,
>is it worth it to go to a high NA (perhaps oil
>immersion) condenser, and can anyone think of
>why the lamp would give a better DIC image than
>a solid state light source?
>thanks in advance for the help...
>  -Jeff

--
               ****************************************
James and Christine Pawley, 5446 Burley Place (PO
Box 2348), Sechelt, BC, Canada, V0N3A0,
Phone [phone], email <[hidden email]>
NEW! NEW! AND DIFFERENT Cell (when I remember to turn it on!) 1-604-989-6146