DIC prism, fluorescence and focus shift

[Stanislav Vitha]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hallo,
I would appreciate a pointer to a paper or a book chapter that provides the
theoretical background/explanation for the fact that a DIC prism will degrade
resolution of the fluorescence image. My search in archives of this list brought
up quite a few discussions on this topic, but I was looking for a more
theoretical treatment of the issue.

I have a little puzzle with our Fluoview FV1000 system  (for those not familiar
with Olympus setup: these is a single DIC prism behind the objective; the
condenser turret then has several different DIC prisms, specific to each
objective - but that is not relevant for epifluorescnece).
A expected, having the DIC prism behind the objective degrades the confocal
image with all objectives, but surprisingly the degradation is much worse with
our 20x/0.85 oil than is with the 60x/1.2 WI or 100x/1.4 oil objectives. I
expected the opposite, the amount of shear introduced by the prism (a
fraction of a micrometer) should have been more noticeable with higher
resolution objectives.  It makes it very easy to demonstrate the effect of DIC
prism on confocal image quality to new users during training, but I would sure
like to know what is happening.
 
The second puzzle is that inserting the DIC prism in the optical path causes a
shift in focus (at least several micrometers, very noticeable). In my naive
thinking, I assumed that inserting a flat piece of glas at normal orientation into
the infininty space would not cause any focus shift, at least for imaging
sample area close to the center of the optical axis.
Any ideas?  

I am in discussion with Olympus and thay are very receptive, but I thought I
would ask here as well.

 Thanks in advance!

Stan Vitha
Microscopy and Imaging Center
Texas A&M University
College Station, TX

[Johannes Helm]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Stanislav.


You might have a look at

http://micro.magnet.fsu.edu/primer/techniques/dic/dichome.html

Depending on what your concept of "theoretical" is, this might be the
source of information you are looking for - or not.

The wording "behind" the objective in your text is a little bit un-clear.
What you mean, I suppose, is the following:

In a classical transmitted light configuration, you will have the light
source, a collector lens, the field diaphragm, then the DIC polarizer, the
condenser DIC prism - also known as the "main DIC prism" -, the condenser
with its front lens, the condenser immersion medium (air, water, oil, or
glycerine),  the microscope slide, the specimen in its mounting medium,
the cover slip, the immersion medium for the objective - e.g. air, water,
oil, glycerine -, the objective, the objective DIC prism - also known as
"auxiliary DIC prism" -, the DIC analyzer and finally the observation or
imaging optics.

What you describe as the "DIC prism behind the objective" most probably is
the auxiliary DIC prism, i.e. the DIC prism on the image side of the
objective. While a number of microscope producers have this auxiliary DIC
prism mounted so that you can adjust it for optimizing your contrast image
("DIC slider"), in the Olympus systems an adjustment is done in a slightly
different way; Olympus provides a lambda/4 plate as an additional element
between DIC polarizer and condenser DIC prism. The user then adjusts the
orientation of the DIC polarizer in order to change the contrast settings.
Also, optimal DIC observation sometimes may request a rotatable specimen
stage centered in the optical axis of the microscope.

The name DIC, "D"IC, "D"ifferential Interference Contrast, delivers the
explanation to your question. The DIC prism, also known as "Nomarski
prism", named after its inventor Jerzy Nomarski, provides a "d"ifferential
splitting of the ray paths of the light in the microscope. If you do
confocal laser scanning microscopy and have a well polarized laser beam -
normally, a laser beam generated by a "typical CLSM laser" will be
polarized per se, but it might nevertheless be sent through the DIC
analyzer (which is nothing else but a polarizer) -, then that beam will be
split into two beams, which are laterally displaced "differentially". The
lateral separation will hence be small, smaller than the lateral
resolution limit of the objective when used in the wide field observation
mode. Other types of interference contrast microscopes - e.g. Jamin
Lebedeff - will generate two beams with a macroscopic separation; then one
beam, the reference beam, will bypass the entire microscope on a ray path
outside the microscope frame, but this is not of any interest here.
What hence happens when you try to do CLSM with a fluorescent specimen -
fluorescent light assumed to be un-polarized - and have a DIC prism in the
ray path is that you will experience what are a kind of super imposed
"double images" if you have a sufficiently large zooming factor so that
you approach the resolution limit with your pixelization. The story is a
little bit more complicated in case of reflected light CLS microscopy
because of possible polarization conservation effects, but let's keep this
out of the discussion here for simplicity.

So, in order to avoid these effects, take out the secondary DIC prism from
the ray path when CLSMing.

NOW, there are some microscope objective holders, e.g. on a BX51WI used by
cell physiologists, e.g. Olympus item code WI-SRE2, in which a secondary
DIC prism, e.g. WI-DICTHRA2, is inserted so that you cannot easilry remove
it from the ray path. There does, however, exist a version of WI-SRE2,
named "FV10-SRE", if I recall correctly, where you have a slider, which
will permit you to easily insert or extract the DIC prism from the ray
path.

Unfortunately, this FV10-SRE seemingly is not sold a million times per
month or so, hence it is not really cheap. But it might help.

Best wishes,

Johannes



--
P. Johannes Helm, M.Sc. PhD
Seniorengineer
CMBN
University of Oslo
Institute of Basic Medical Science
Department of Anatomy
Postboks 1105 - Blindern
NO-0317 Oslo

Voice: [phone]
Fax: [phone]

WWW: folk.uio.no/jhelm

[Stanislav Vitha]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Johannes,
Thank you for your detailed explanation. I should clarify that our confocal
system has an inverted IX-81 microscope.
By "behind the objective" I meant on the opposite side of the objective than
the specimen is (the convention, I believe, is that when taking pictures, the
object is in front of the lens, and photographer is behind the lens. The image
is formed behind the lens, which is also the side where we find the back focal
plane).

Since I am collecting confocal epifluorescnce image simultaneously with
recording a DIC image on the transmitted detector, the degradation of
fluorescence images relates to the objective DIC prism. The condenser DIC
prism is not in the optical path of the epifluorescence imaging setup.  

I prefer to take the DIC prism out of the optical path when doing fluorescence
imaging, but there are situations where I need DIC + fluorescence
simultaneously.
 
My main question was not why I get degradation of resolution in the
fluorescence channel (that I expected),  but rather why is the degradation
worse for a lower-resolution objective  (20x/0.85 oil imm versus 100x/1.4 oil
imm). The "image doubling" due to the shear of the objective DIC prism should
be more noticeable with the high-resolution lens.

Stan Vitha
   
On Thu, 7 Apr 2011 22:42:00 +0200, P. Johannes Helm
<[hidden email]> wrote:

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Dear Stanislav.
>
>
>You might have a look at
>
>http://micro.magnet.fsu.edu/primer/techniques/dic/dichome.html
>
>Depending on what your concept of "theoretical" is, this might be the
>source of information you are looking for - or not.
>
>The wording "behind" the objective in your text is a little bit un-clear.
>What you mean, I suppose, is the following:
>
>In a classical transmitted light configuration, you will have the light
>source, a collector lens, the field diaphragm, then the DIC polarizer, the
>condenser DIC prism - also known as the "main DIC prism" -, the condenser
>with its front lens, the condenser immersion medium (air, water, oil, or
>glycerine),  the microscope slide, the specimen in its mounting medium,
>the cover slip, the immersion medium for the objective - e.g. air, water,
>oil, glycerine -, the objective, the objective DIC prism - also known as
>"auxiliary DIC prism" -, the DIC analyzer and finally the observation or
>imaging optics.
>
>What you describe as the "DIC prism behind the objective" most probably is
>the auxiliary DIC prism, i.e. the DIC prism on the image side of the
>objective. While a number of microscope producers have this auxiliary DIC
>prism mounted so that you can adjust it for optimizing your contrast image
>("DIC slider"), in the Olympus systems an adjustment is done in a slightly
>different way; Olympus provides a lambda/4 plate as an additional element
>between DIC polarizer and condenser DIC prism. The user then adjusts the
>orientation of the DIC polarizer in order to change the contrast settings.
>Also, optimal DIC observation sometimes may request a rotatable specimen
>stage centered in the optical axis of the microscope.
>
>The name DIC, "D"IC, "D"ifferential Interference Contrast, delivers the
>explanation to your question. The DIC prism, also known as "Nomarski
>prism", named after its inventor Jerzy Nomarski, provides a "d"ifferential
>splitting of the ray paths of the light in the microscope. If you do
>confocal laser scanning microscopy and have a well polarized laser beam -
>normally, a laser beam generated by a "typical CLSM laser" will be
>polarized per se, but it might nevertheless be sent through the DIC
>analyzer (which is nothing else but a polarizer) -, then that beam will be
>split into two beams, which are laterally displaced "differentially". The
>lateral separation will hence be small, smaller than the lateral
>resolution limit of the objective when used in the wide field observation
>mode. Other types of interference contrast microscopes - e.g. Jamin
>Lebedeff - will generate two beams with a macroscopic separation; then one
>beam, the reference beam, will bypass the entire microscope on a ray path
>outside the microscope frame, but this is not of any interest here.
>What hence happens when you try to do CLSM with a fluorescent specimen -
>fluorescent light assumed to be un-polarized - and have a DIC prism in the
>ray path is that you will experience what are a kind of super imposed
>"double images" if you have a sufficiently large zooming factor so that
>you approach the resolution limit with your pixelization. The story is a
>little bit more complicated in case of reflected light CLS microscopy
>because of possible polarization conservation effects, but let's keep this
>out of the discussion here for simplicity.
>
>So, in order to avoid these effects, take out the secondary DIC prism from
>the ray path when CLSMing.
>
>NOW, there are some microscope objective holders, e.g. on a BX51WI used

by

>cell physiologists, e.g. Olympus item code WI-SRE2, in which a secondary
>DIC prism, e.g. WI-DICTHRA2, is inserted so that you cannot easilry remove
>it from the ray path. There does, however, exist a version of WI-SRE2,
>named "FV10-SRE", if I recall correctly, where you have a slider, which
>will permit you to easily insert or extract the DIC prism from the ray
>path.
>
>Unfortunately, this FV10-SRE seemingly is not sold a million times per
>month or so, hence it is not really cheap. But it might help.
>
>Best wishes,
>
>Johannes
>
>
>
>--
>P. Johannes Helm, M.Sc. PhD
>Seniorengineer
>CMBN
>University of Oslo
>Institute of Basic Medical Science
>Department of Anatomy
>Postboks 1105 - Blindern
>NO-0317 Oslo
>
>Voice: [phone]
>Fax: [phone]
>
>WWW: folk.uio.no/jhelm

[Johannes Helm]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi, again, Stanislav,

the "double" image, I think, might be prominent in any type of lens as
soon as you approach the resolution limit with your pixelization, in other
words, as soon as you "nyquist the image". Of course, the construction of
the prism will also have a considerable influence on this matter.

Best,
Johannes

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Dear Johannes,
> Thank you for your detailed explanation. I should clarify that our
> confocal
> system has an inverted IX-81 microscope.
> By "behind the objective" I meant on the opposite side of the objective
> than
> the specimen is (the convention, I believe, is that when taking pictures,
> the
> object is in front of the lens, and photographer is behind the lens. The
> image
> is formed behind the lens, which is also the side where we find the back
> focal
> plane).
>
> Since I am collecting confocal epifluorescnce image simultaneously with
> recording a DIC image on the transmitted detector, the degradation of
> fluorescence images relates to the objective DIC prism. The condenser DIC
> prism is not in the optical path of the epifluorescence imaging setup.
>
> I prefer to take the DIC prism out of the optical path when doing
> fluorescence
> imaging, but there are situations where I need DIC + fluorescence
> simultaneously.
>
> My main question was not why I get degradation of resolution in the
> fluorescence channel (that I expected),  but rather why is the degradation
> worse for a lower-resolution objective  (20x/0.85 oil imm versus 100x/1.4
> oil
> imm). The "image doubling" due to the shear of the objective DIC prism
> should
> be more noticeable with the high-resolution lens.
>
> Stan Vitha
>
> On Thu, 7 Apr 2011 22:42:00 +0200, P. Johannes Helm
> <[hidden email]> wrote:
>
>>*****
>>To join, leave or search the confocal microscopy listserv, go to:
>>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>*****
>>
>>Dear Stanislav.
>>
>>
>>You might have a look at
>>
>>http://micro.magnet.fsu.edu/primer/techniques/dic/dichome.html
>>
>>Depending on what your concept of "theoretical" is, this might be the
>>source of information you are looking for - or not.
>>
>>The wording "behind" the objective in your text is a little bit un-clear.
>>What you mean, I suppose, is the following:
>>
>>In a classical transmitted light configuration, you will have the light
>>source, a collector lens, the field diaphragm, then the DIC polarizer,
>> the
>>condenser DIC prism - also known as the "main DIC prism" -, the condenser
>>with its front lens, the condenser immersion medium (air, water, oil, or
>>glycerine),  the microscope slide, the specimen in its mounting medium,
>>the cover slip, the immersion medium for the objective - e.g. air, water,
>>oil, glycerine -, the objective, the objective DIC prism - also known as
>>"auxiliary DIC prism" -, the DIC analyzer and finally the observation or
>>imaging optics.
>>
>>What you describe as the "DIC prism behind the objective" most probably
>> is
>>the auxiliary DIC prism, i.e. the DIC prism on the image side of the
>>objective. While a number of microscope producers have this auxiliary DIC
>>prism mounted so that you can adjust it for optimizing your contrast
>> image
>>("DIC slider"), in the Olympus systems an adjustment is done in a
>> slightly
>>different way; Olympus provides a lambda/4 plate as an additional element
>>between DIC polarizer and condenser DIC prism. The user then adjusts the
>>orientation of the DIC polarizer in order to change the contrast
>> settings.
>>Also, optimal DIC observation sometimes may request a rotatable specimen
>>stage centered in the optical axis of the microscope.
>>
>>The name DIC, "D"IC, "D"ifferential Interference Contrast, delivers the
>>explanation to your question. The DIC prism, also known as "Nomarski
>>prism", named after its inventor Jerzy Nomarski, provides a
>> "d"ifferential
>>splitting of the ray paths of the light in the microscope. If you do
>>confocal laser scanning microscopy and have a well polarized laser beam -
>>normally, a laser beam generated by a "typical CLSM laser" will be
>>polarized per se, but it might nevertheless be sent through the DIC
>>analyzer (which is nothing else but a polarizer) -, then that beam will
>> be
>>split into two beams, which are laterally displaced "differentially". The
>>lateral separation will hence be small, smaller than the lateral
>>resolution limit of the objective when used in the wide field observation
>>mode. Other types of interference contrast microscopes - e.g. Jamin
>>Lebedeff - will generate two beams with a macroscopic separation; then
>> one
>>beam, the reference beam, will bypass the entire microscope on a ray path
>>outside the microscope frame, but this is not of any interest here.
>>What hence happens when you try to do CLSM with a fluorescent specimen -
>>fluorescent light assumed to be un-polarized - and have a DIC prism in
>> the
>>ray path is that you will experience what are a kind of super imposed
>>"double images" if you have a sufficiently large zooming factor so that
>>you approach the resolution limit with your pixelization. The story is a
>>little bit more complicated in case of reflected light CLS microscopy
>>because of possible polarization conservation effects, but let's keep
>> this
>>out of the discussion here for simplicity.
>>
>>So, in order to avoid these effects, take out the secondary DIC prism
>> from
>>the ray path when CLSMing.
>>
>>NOW, there are some microscope objective holders, e.g. on a BX51WI used
> by
>>cell physiologists, e.g. Olympus item code WI-SRE2, in which a secondary
>>DIC prism, e.g. WI-DICTHRA2, is inserted so that you cannot easilry
>> remove
>>it from the ray path. There does, however, exist a version of WI-SRE2,
>>named "FV10-SRE", if I recall correctly, where you have a slider, which
>>will permit you to easily insert or extract the DIC prism from the ray
>>path.
>>
>>Unfortunately, this FV10-SRE seemingly is not sold a million times per
>>month or so, hence it is not really cheap. But it might help.
>>
>>Best wishes,
>>
>>Johannes
>>
>>
>>
>>--
>>P. Johannes Helm, M.Sc. PhD
>>Seniorengineer
>>CMBN
>>University of Oslo
>>Institute of Basic Medical Science
>>Department of Anatomy
>>Postboks 1105 - Blindern
>>NO-0317 Oslo
>>
>>Voice: [phone]
>>Fax: [phone]
>>
>>WWW: folk.uio.no/jhelm
>

--
P. Johannes Helm

Voice: ([phone] (office)
Fax: [phone] (office)

[Guy Cox-2]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

This is just guessing, but the 20x NA 0.85 is very low mag for its NA.
This may mean that the light rays are going through a greater width of
the Wollaston prism.  

                                          Guy

Sponsor my next half-marathon on May 15th
There's a special reason - find it out at
http://www.everydayhero.com.au/Guy_Cox_4846
______________________________________________
Associate Professor Guy Cox, MA, DPhil(Oxon)
Australian Centre for Microscopy & Microanalysis,
Madsen Building F09, University of Sydney, NSW 2006

Phone [phone]     Fax [phone]
             Mobile [phone]
______________________________________________
      http://www.guycox.net


-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]]
On Behalf Of Stanislav Vitha
Sent: Friday, 8 April 2011 5:28 PM
To: [hidden email]
Subject: Re: DIC prism, fluorescence and focus shift

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Johannes,
Thank you for your detailed explanation. I should clarify that our
confocal
system has an inverted IX-81 microscope.
By "behind the objective" I meant on the opposite side of the objective
than
the specimen is (the convention, I believe, is that when taking
pictures, the
object is in front of the lens, and photographer is behind the lens. The
image
is formed behind the lens, which is also the side where we find the back
focal
plane).

Since I am collecting confocal epifluorescnce image simultaneously with
recording a DIC image on the transmitted detector, the degradation of
fluorescence images relates to the objective DIC prism. The condenser
DIC
prism is not in the optical path of the epifluorescence imaging setup.  

I prefer to take the DIC prism out of the optical path when doing
fluorescence
imaging, but there are situations where I need DIC + fluorescence
simultaneously.
 
My main question was not why I get degradation of resolution in the
fluorescence channel (that I expected),  but rather why is the
degradation
worse for a lower-resolution objective  (20x/0.85 oil imm versus
100x/1.4 oil
imm). The "image doubling" due to the shear of the objective DIC prism
should
be more noticeable with the high-resolution lens.

Stan Vitha
   
On Thu, 7 Apr 2011 22:42:00 +0200, P. Johannes Helm
<[hidden email]> wrote:

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Dear Stanislav.
>
>
>You might have a look at
>
>http://micro.magnet.fsu.edu/primer/techniques/dic/dichome.html
>
>Depending on what your concept of "theoretical" is, this might be the
>source of information you are looking for - or not.
>
>The wording "behind" the objective in your text is a little bit

un-clear.
>What you mean, I suppose, is the following:
>
>In a classical transmitted light configuration, you will have the light
>source, a collector lens, the field diaphragm, then the DIC polarizer,
the
>condenser DIC prism - also known as the "main DIC prism" -, the
condenser
>with its front lens, the condenser immersion medium (air, water, oil,
or
>glycerine),  the microscope slide, the specimen in its mounting medium,
>the cover slip, the immersion medium for the objective - e.g. air,
water,
>oil, glycerine -, the objective, the objective DIC prism - also known
as
>"auxiliary DIC prism" -, the DIC analyzer and finally the observation
or
>imaging optics.
>
>What you describe as the "DIC prism behind the objective" most probably
is
>the auxiliary DIC prism, i.e. the DIC prism on the image side of the
>objective. While a number of microscope producers have this auxiliary
DIC
>prism mounted so that you can adjust it for optimizing your contrast
image
>("DIC slider"), in the Olympus systems an adjustment is done in a
slightly
>different way; Olympus provides a lambda/4 plate as an additional
element
>between DIC polarizer and condenser DIC prism. The user then adjusts
the
>orientation of the DIC polarizer in order to change the contrast
settings.
>Also, optimal DIC observation sometimes may request a rotatable
specimen
>stage centered in the optical axis of the microscope.
>
>The name DIC, "D"IC, "D"ifferential Interference Contrast, delivers the
>explanation to your question. The DIC prism, also known as "Nomarski
>prism", named after its inventor Jerzy Nomarski, provides a
"d"ifferential
>splitting of the ray paths of the light in the microscope. If you do
>confocal laser scanning microscopy and have a well polarized laser beam
-
>normally, a laser beam generated by a "typical CLSM laser" will be
>polarized per se, but it might nevertheless be sent through the DIC
>analyzer (which is nothing else but a polarizer) -, then that beam will
be
>split into two beams, which are laterally displaced "differentially".
The
>lateral separation will hence be small, smaller than the lateral
>resolution limit of the objective when used in the wide field
observation
>mode. Other types of interference contrast microscopes - e.g. Jamin
>Lebedeff - will generate two beams with a macroscopic separation; then
one
>beam, the reference beam, will bypass the entire microscope on a ray
path
>outside the microscope frame, but this is not of any interest here.
>What hence happens when you try to do CLSM with a fluorescent specimen
-
>fluorescent light assumed to be un-polarized - and have a DIC prism in
the
>ray path is that you will experience what are a kind of super imposed
>"double images" if you have a sufficiently large zooming factor so that
>you approach the resolution limit with your pixelization. The story is
a
>little bit more complicated in case of reflected light CLS microscopy
>because of possible polarization conservation effects, but let's keep
this
>out of the discussion here for simplicity.
>
>So, in order to avoid these effects, take out the secondary DIC prism
from
>the ray path when CLSMing.
>
>NOW, there are some microscope objective holders, e.g. on a BX51WI used

by
>cell physiologists, e.g. Olympus item code WI-SRE2, in which a
secondary
>DIC prism, e.g. WI-DICTHRA2, is inserted so that you cannot easilry
remove

>it from the ray path. There does, however, exist a version of WI-SRE2,
>named "FV10-SRE", if I recall correctly, where you have a slider, which
>will permit you to easily insert or extract the DIC prism from the ray
>path.
>
>Unfortunately, this FV10-SRE seemingly is not sold a million times per
>month or so, hence it is not really cheap. But it might help.
>
>Best wishes,
>
>Johannes
>
>
>
>--
>P. Johannes Helm, M.Sc. PhD
>Seniorengineer
>CMBN
>University of Oslo
>Institute of Basic Medical Science
>Department of Anatomy
>Postboks 1105 - Blindern
>NO-0317 Oslo
>
>Voice: [phone]
>Fax: [phone]
>
>WWW: folk.uio.no/jhelm

[Shalin Mehta]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hello Stan,

On Fri, Apr 8, 2011 at 3:28 PM, Stanislav Vitha <[hidden email]> wrote:
>
> My main question was not why I get degradation of resolution in the
> fluorescence channel (that I expected),  but rather why is the degradation
> worse for a lower-resolution objective  (20x/0.85 oil imm versus 100x/1.4 oil
> imm). The "image doubling" due to the shear of the objective DIC prism should
> be more noticeable with the high-resolution lens

I happened to have gone through this exercise when measuring the shear
in our DIC setup.
We tried to recap what we understood in the appendix of our paper
(http://www.mshalin.com/blog/wp-content/uploads/2010/09/mehtaao2010-samplefreecalibrationdic.pdf).

Here is an attempt at understanding the specific case that you described:

Nomarski prism is an angular polarizing beam-splitting devices. i.e.,
it splits the incoming light into two polarizations that travel at
different angles. The angular split caused by the Nomarski prism
(placed in the back focal plane) amounts to spatial shear in the
specimen plane. The relationship between angular shear (a) and spatial
shear (s) is rather simple: s=tan(a)*f. Where f is the focal length of
the objective. This fact can be seen easily from the geometry of fig.
11(d) in above paper. Now the image doubling is caused when the shear
is large with respect to the resolution. To judge the doubling, we
should normalize both sides of the above equation by lambda/NA. Then,
we have s/(lambda/NA) = tan(a) * f/(lambda/NA). Therefore, when
s/(lambda/NA) or s*NA is large (for given wavelength) we will have
image doubling.

In the Olympus setup there is only one DIC prism on objective side,
i.e., the angular shear is the same for all objectives. Therefore, the
objectives for which f*NA is large will give rise to larger s*NA and
hence image doubling.

The product of focal length and NA is large for 20X 0.85 than for
100X1.4. I compute the product of focal length and NA below.

Objective's magnification (M) = tube length (ft)/ objective's focal length (f).
Therefore, f=ft/M.
Therefore, f*NA = ft*(NA/M).
For, 20X 0.85: f*NA= 0.0425*ft.
For, 100x 1.4: f*NA= 0.014*ft.

So the 'image doubling'  for 20X 0.85 is almost 3 times more than  for 100x 1.4.

Now, it is interesting to see how this relates with Guy's intuitive observation.

NA of the objective = radius of the objective BFP (d)/ objective focal
length (f).
So, r=f*NA. As we have seen above,  the size of the BFP (ie., f*NA)
for 20X 0.85 is 3 times larger than for 100x1.4. So as Guy says, rays
collected by 20X 0.85 pass through larger (3 times larger) surface of
the Wollaston prism.

Best
Shalin

http://mshalin.com

Research Associate
Bioimaging Lab, Block-E3A, #7-10
Div of Bioengineering, NUS Singapore 117574
(M) [phone]

[Stanislav Vitha]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Shalin (and Guy),
thank you so much for the explanation and the article, it is exactly what
I was looking for. It now makes sense.

I have to disclose that there may be one additional factor that
contributes to image degradation with the 20x objective (and has
something to do with the large diameter of the BFP) - I looked at the
DIC prism and discovered that there is an oil droplet close to he edge of
the prism. Generous application of immersion oil by our users must have
caused the immersion oil to run down, seep through the thread and drip
on the prism. So I think the oil droplet would degrade image more
severely for objectives with large BFP.  

Now off to do some cleaning.


Stan

Dr. Stanislav Vitha      
Microscopy and Imaging Center
Texas A&M University
BSBW 119
College Station, TX 77843-2257

http://microscopy.tamu.edu


On Sun, 10 Apr 2011 22:31:06 +0800, Shalin Mehta
<[hidden email]> wrote:

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Hello Stan,
>
>On Fri, Apr 8, 2011 at 3:28 PM, Stanislav Vitha <[hidden email]>
wrote:
>>
>> My main question was not why I get degradation of resolution in the
>> fluorescence channel (that I expected),  but rather why is the
degradation
>> worse for a lower-resolution objective  (20x/0.85 oil imm versus
100x/1.4 oil
>> imm). The "image doubling" due to the shear of the objective DIC
prism should
>> be more noticeable with the high-resolution lens
>
>I happened to have gone through this exercise when measuring the
shear
>in our DIC setup.
>We tried to recap what we understood in the appendix of our paper
>(http://www.mshalin.com/blog/wp-
content/uploads/2010/09/mehtaao2010-samplefreecalibrationdic.pdf).
>
>Here is an attempt at understanding the specific case that you
described:
>
>Nomarski prism is an angular polarizing beam-splitting devices. i.e.,
>it splits the incoming light into two polarizations that travel at
>different angles. The angular split caused by the Nomarski prism
>(placed in the back focal plane) amounts to spatial shear in the
>specimen plane. The relationship between angular shear (a) and spatial
>shear (s) is rather simple: s=tan(a)*f. Where f is the focal length of
>the objective. This fact can be seen easily from the geometry of fig.
>11(d) in above paper. Now the image doubling is caused when the
shear
>is large with respect to the resolution. To judge the doubling, we
>should normalize both sides of the above equation by lambda/NA.
Then,

>we have s/(lambda/NA) = tan(a) * f/(lambda/NA). Therefore, when
>s/(lambda/NA) or s*NA is large (for given wavelength) we will have
>image doubling.
>
>In the Olympus setup there is only one DIC prism on objective side,
>i.e., the angular shear is the same for all objectives. Therefore, the
>objectives for which f*NA is large will give rise to larger s*NA and
>hence image doubling.
>
>The product of focal length and NA is large for 20X 0.85 than for
>100X1.4. I compute the product of focal length and NA below.
>
>Objective's magnification (M) = tube length (ft)/ objective's focal

length (f).
>Therefore, f=ft/M.
>Therefore, f*NA = ft*(NA/M).
>For, 20X 0.85: f*NA= 0.0425*ft.
>For, 100x 1.4: f*NA= 0.014*ft.
>
>So the 'image doubling'  for 20X 0.85 is almost 3 times more than  for
100x 1.4.
>
>Now, it is interesting to see how this relates with Guy's intuitive
observation.

>
>NA of the objective = radius of the objective BFP (d)/ objective focal
>length (f).
>So, r=f*NA. As we have seen above,  the size of the BFP (ie., f*NA)
>for 20X 0.85 is 3 times larger than for 100x1.4. So as Guy says, rays
>collected by 20X 0.85 pass through larger (3 times larger) surface of
>the Wollaston prism.
>
>Best
>Shalin
>
>http://mshalin.com
>
>Research Associate
>Bioimaging Lab, Block-E3A, #7-10
>Div of Bioengineering, NUS Singapore 117574
>(M) [phone]