Confocal Microscopy List

Nikon 60x/1.49NA lens tube lens distance to camera

Classic

List

Threaded

8 messages Options

Lu Yan

Nikon 60x/1.49NA lens tube lens distance to camera

*****
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 listers,

I hope some from the list could enlighten me. We recently purchased a Nikon 60X/1.49Oil TIRF lens, and I have been playing with it a bit, and found that when imaging fluorescent beads, it creates 'star-like' spots at the edge of the lens, which is somewhat expected since high NA and mag lenses do not behave good when the FOV is large.

However, what surprised me was other interesting findings. I found a collection of Zemax model files online for Nikon's various objective lens designs, one of which is a 60x/1.49Oil. So I took the file and ran some simple simulations:

* Surfaces: the objective+tube lens (set 150 mm away from the last surface of the objective lens).
* Settings:
* I put the object on the coverslip;
* I set the coverslip thickness of 0.170 mm, with refractive index nd= 1.523, and abbe number Vd=55, and I set the oil refractive index nd = 1.518, and vd=41 (got from Thorlabs website), with varying thickness;
* I also set the distance between the tube lens and the image plane to be a variable so that I can run the optimization for smallest RMS wavefront aberration or spot radius (reference to the Centroid) on the image plane;
* I ran the optimization for the smallest RMS wavefront aberration, with above two variables, i.e. oil thickness (related to the working distance), and image plane distance to the tube lens.

The optimal results I got are:

* When optimizing for RMS wavefront aberration:
* Oil thickness=0.139 mm, which is close to the working distance of 0.12mm.
* Image (or camera) plane distance = 137.08mm!
* Spot RMS radius on the image plane = 129.8 um!
* When optimizing for RMS spot radius:
*
Oil thickness=0.141mm.
*
Image (or camera) plane distance = 130.63mm!
*
Spot RMS radius on the image plane = 79.5 um!

So my questions are:

1. Do these results suggest the Zemax model was not accurate? And the tube lens distance is supposed to be one focal length away from the camera for best imaging quality?
2. For Zemax modeling of microscopy imaging, which one should be a more practical figure of merit when optimizing the design, the RMS spot radius or RMS wavefront aberration? What properties/performance parameters would you suggest I look for when doing the zemax modeling?
3. As the sample is deviating from the coverslip, should I expect seeing more aberration or less aberration at the edge of the FOV?

For your information, I also simulated when the sample/object is 0.01mm from the coverslip, with embedded medium nd=1.472, and Vd=41. The corresponding results are:

* When optimizing for RMS wavefront aberration:
* Oil thickness=0.119mm.
* Image (or camera) plane distance = 177.16mm!
* Spot RMS radius on the image plane = 80.1 um!
* When optimizing for RMS spot radius:
* Oil thickness=0.141mm.
* Image (or camera) plane distance = 177.16mm
* Spot RMS radius on the image plane = 80.8 um
* SAME AS RMS WAVEFRONT RESULTS.

This bugged me for several days. The required image/camera plane became 177 mm, from ~140 mm. Was this simply because the inaccurate Zemax model? Should we believe that the objective lenses are designed to have the best imaging quality when the camera is placed one focal length from the tube lens? How sensitive practically speaking, is the image quality to the distance between the sample and the coverslip?

Any input from the list is welcome, and thank you all in advance.

Have a nice weekend,
Lu

Zdenek Svindrych-2

Re: Nikon 60x/1.49NA lens tube lens distance to camera

*****
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.
*****

Hi Lu,
clearly there is something wrong with your model, 80 um spot radius in the
image plane is way too much, I would expect less than 20 um.

The lens prescriptions usually come from patents, so the actual production
lens might be slightly different. Anyway, I would check the original patent
to see if your model is correct and at what conditions (coverslip, oil,
wavelengths, tubelens) the lens is supposed to work OK.

https://patents.google.com/patent/US7046451B2/en

Also you may check the zemax files associated with this paper:

https://arxiv.org/abs/1507.04037

where the authors did some modelling of high-NA lenses.

As to what parameters are important for modeling... what is your
application? For TIRF microscopy there will be design criteria related to
the TIRF illumination; when it comes to resolution you should put more
weight on the high-NA rays, as quite big part of your signal actually comes
from supercritical angle fluorescence. For multiphoton imaging you might
want to optimize for Strehl ratio (the peak intensity in the focus), etc...

Btw, our old Olympus 60x / 1.45 NA TIRFM lens also shows quite strong
aberrations near the edges of the FOV.

Hope it helps. Best, zdenek

--
Zdenek Svindrych, Ph.D.
W.M. Keck Center for Cellular Imaging (PLSB 003)
Department of Biology,University of Virginia
409 McCormick Rd, Charlottesville, VA-22904
http://www.kcci.virginia.edu/
tel: 434-982-4869

---------- Původní e-mail ----------
Od: Yan, Lu <[hidden email]>
Komu: [hidden email]
Datum: 13. 10. 2017 21:16:24
Předmět: Nikon 60x/1.49NA lens tube lens distance to camera
"*****
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 listers,

I hope some from the list could enlighten me. We recently purchased a Nikon
60X/1.49Oil TIRF lens, and I have been playing with it a bit, and found that
when imaging fluorescent beads, it creates 'star-like' spots at the edge of
the lens, which is somewhat expected since high NA and mag lenses do not
behave good when the FOV is large.

However, what surprised me was other interesting findings. I found a
collection of Zemax model files online for Nikon's various objective lens
designs, one of which is a 60x/1.49Oil. So I took the file and ran some
simple simulations:

* Surfaces: the objective+tube lens (set 150 mm away from the last surface
of the objective lens).
* Settings:
* I put the object on the coverslip;
* I set the coverslip thickness of 0.170 mm, with refractive index nd=
1.523, and abbe number Vd=55, and I set the oil refractive index nd = 1.518,
and vd=41 (got from Thorlabs website), with varying thickness;
* I also set the distance between the tube lens and the image plane to be a
variable so that I can run the optimization for smallest RMS wavefront
aberration or spot radius (reference to the Centroid) on the image plane;
* I ran the optimization for the smallest RMS wavefront aberration, with
above two variables, i.e. oil thickness (related to the working distance),
and image plane distance to the tube lens.

The optimal results I got are:

* When optimizing for RMS wavefront aberration:
* Oil thickness=0.139 mm, which is close to the working distance of 0.12mm.
* Image (or camera) plane distance = 137.08mm!
* Spot RMS radius on the image plane = 129.8 um!
* When optimizing for RMS spot radius:
*
Oil thickness=0.141mm.
*
Image (or camera) plane distance = 130.63mm!
*
Spot RMS radius on the image plane = 79.5 um!

So my questions are:

1. Do these results suggest the Zemax model was not accurate? And the tube
lens distance is supposed to be one focal length away from the camera for
best imaging quality?
2. For Zemax modeling of microscopy imaging, which one should be a more
practical figure of merit when optimizing the design, the RMS spot radius or
RMS wavefront aberration? What properties/performance parameters would you
suggest I look for when doing the zemax modeling?
3. As the sample is deviating from the coverslip, should I expect seeing
more aberration or less aberration at the edge of the FOV?

For your information, I also simulated when the sample/object is 0.01mm from
the coverslip, with embedded medium nd=1.472, and Vd=41. The corresponding
results are:

* When optimizing for RMS wavefront aberration:
* Oil thickness=0.119mm.
* Image (or camera) plane distance = 177.16mm!
* Spot RMS radius on the image plane = 80.1 um!
* When optimizing for RMS spot radius:
* Oil thickness=0.141mm.
* Image (or camera) plane distance = 177.16mm
* Spot RMS radius on the image plane = 80.8 um
* SAME AS RMS WAVEFRONT RESULTS.

This bugged me for several days. The required image/camera plane became 177
mm, from ~140 mm. Was this simply because the inaccurate Zemax model? Should
we believe that the objective lenses are designed to have the best imaging
quality when the camera is placed one focal length from the tube lens? How
sensitive practically speaking, is the image quality to the distance between
the sample and the coverslip?

Any input from the list is welcome, and thank you all in advance.

Have a nice weekend,
Lu
"

John Oreopoulos

Re: Nikon 60x/1.49NA lens tube lens distance to camera

In reply to this post by Lu Yan

*****
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 Lu,
>
> There's a lot to unpack there, so let me just ask a simple question first: How are you modeling the object to be imaged by the objective in Zemax?
>
> You quoted and RMS spot size of something like 80 um, but a rough calculation based on the NA of the objective and assuming a green emission wavelength of 525 nm, I calculate the Rayleigh resolution as 176nm (lambda/2NA). So, the diameter of the Airy disk (which has some correspondence to the RMS spot size) should be about 21 um under ideal conditions. Hard to see how the objective would yield a spot size that is roughly 4x bigger than what it should be (the diffraction limit), even for an object at the centre of the field of view. I think there would be very many unhappy customers if the objectives really behaved like that. So I take this to mean there is something not quite right with how you've set up the full ray tracing model, and hence my question about the light emitting object.
>
> Maybe take a step backwards and make sure you get sensible answers by modeling the objective and tube lenses as thin paraxial ideal lenses first. Then you'll know if the object model is correct.
>

John Oreopoulos

On 2017-10-13, at 9:13 PM, Yan, Lu wrote:

> *****
> 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 listers,
>
>
> I hope some from the list could enlighten me. We recently purchased a Nikon 60X/1.49Oil TIRF lens, and I have been playing with it a bit, and found that when imaging fluorescent beads, it creates 'star-like' spots at the edge of the lens, which is somewhat expected since high NA and mag lenses do not behave good when the FOV is large.
>
>
> However, what surprised me was other interesting findings. I found a collection of Zemax model files online for Nikon's various objective lens designs, one of which is a 60x/1.49Oil. So I took the file and ran some simple simulations:
>
>
> * Surfaces: the objective+tube lens (set 150 mm away from the last surface of the objective lens).
> * Settings:
> * I put the object on the coverslip;
> * I set the coverslip thickness of 0.170 mm, with refractive index nd= 1.523, and abbe number Vd=55, and I set the oil refractive index nd = 1.518, and vd=41 (got from Thorlabs website), with varying thickness;
> * I also set the distance between the tube lens and the image plane to be a variable so that I can run the optimization for smallest RMS wavefront aberration or spot radius (reference to the Centroid) on the image plane;
> * I ran the optimization for the smallest RMS wavefront aberration, with above two variables, i.e. oil thickness (related to the working distance), and image plane distance to the tube lens.
>
> The optimal results I got are:
>
> * When optimizing for RMS wavefront aberration:
> * Oil thickness=0.139 mm, which is close to the working distance of 0.12mm.
> * Image (or camera) plane distance = 137.08mm!
> * Spot RMS radius on the image plane = 129.8 um!
> * When optimizing for RMS spot radius:
> *
> Oil thickness=0.141mm.
> *
> Image (or camera) plane distance = 130.63mm!
> *
> Spot RMS radius on the image plane = 79.5 um!
>
> So my questions are:
>
> 1. Do these results suggest the Zemax model was not accurate? And the tube lens distance is supposed to be one focal length away from the camera for best imaging quality?
> 2. For Zemax modeling of microscopy imaging, which one should be a more practical figure of merit when optimizing the design, the RMS spot radius or RMS wavefront aberration? What properties/performance parameters would you suggest I look for when doing the zemax modeling?
> 3. As the sample is deviating from the coverslip, should I expect seeing more aberration or less aberration at the edge of the FOV?
>
> For your information, I also simulated when the sample/object is 0.01mm from the coverslip, with embedded medium nd=1.472, and Vd=41. The corresponding results are:
>
> * When optimizing for RMS wavefront aberration:
> * Oil thickness=0.119mm.
> * Image (or camera) plane distance = 177.16mm!
> * Spot RMS radius on the image plane = 80.1 um!
> * When optimizing for RMS spot radius:
> * Oil thickness=0.141mm.
> * Image (or camera) plane distance = 177.16mm
> * Spot RMS radius on the image plane = 80.8 um
> * SAME AS RMS WAVEFRONT RESULTS.
>
> This bugged me for several days. The required image/camera plane became 177 mm, from ~140 mm. Was this simply because the inaccurate Zemax model? Should we believe that the objective lenses are designed to have the best imaging quality when the camera is placed one focal length from the tube lens? How sensitive practically speaking, is the image quality to the distance between the sample and the coverslip?
>
> Any input from the list is welcome, and thank you all in advance.
>
> Have a nice weekend,
> Lu

John Oreopoulos

Re: Nikon 60x/1.49NA lens tube lens distance to camera

In reply to this post by Zdenek Svindrych-2

*****
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.
*****

Zdenek,

That's a good paper recommendation. Have not seen that one. Lu may also like to take a look at chapter 12 of this book (which I just remembered now):

https://www.spiedigitallibrary.org/ebooks/TT/Optical-Design-of-Microscopes/eISBN-9780819480965/10.1117/3.855480?SSO=1

where a very detailed analysis of the Zemax model for the same TIRF objective is described. It also details what figures of merit might be of interest from the perspective of microscopy.

Cheers again,

John Oreopoulos

On 2017-10-13, at 10:14 PM, [hidden email] wrote:

> *****
> 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.
> *****
>
> Hi Lu,
> clearly there is something wrong with your model, 80 um spot radius in the
> image plane is way too much, I would expect less than 20 um.
>
> The lens prescriptions usually come from patents, so the actual production
> lens might be slightly different. Anyway, I would check the original patent
> to see if your model is correct and at what conditions (coverslip, oil,
> wavelengths, tubelens) the lens is supposed to work OK.
>
> https://patents.google.com/patent/US7046451B2/en
>
> Also you may check the zemax files associated with this paper:
>
> https://arxiv.org/abs/1507.04037
>
>
> where the authors did some modelling of high-NA lenses.
>
> As to what parameters are important for modeling... what is your
> application? For TIRF microscopy there will be design criteria related to
> the TIRF illumination; when it comes to resolution you should put more
> weight on the high-NA rays, as quite big part of your signal actually comes
> from supercritical angle fluorescence. For multiphoton imaging you might
> want to optimize for Strehl ratio (the peak intensity in the focus), etc...
>
> Btw, our old Olympus 60x / 1.45 NA TIRFM lens also shows quite strong
> aberrations near the edges of the FOV.
>
> Hope it helps. Best, zdenek
>
>
> --
> Zdenek Svindrych, Ph.D.
> W.M. Keck Center for Cellular Imaging (PLSB 003)
> Department of Biology,University of Virginia
> 409 McCormick Rd, Charlottesville, VA-22904
> http://www.kcci.virginia.edu/
> tel: 434-982-4869
>
> ---------- Původní e-mail ----------
> Od: Yan, Lu <[hidden email]>
> Komu: [hidden email]
> Datum: 13. 10. 2017 21:16:24
> Předmět: Nikon 60x/1.49NA lens tube lens distance to camera
> "*****
> 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 listers,
>
>
> I hope some from the list could enlighten me. We recently purchased a Nikon
> 60X/1.49Oil TIRF lens, and I have been playing with it a bit, and found that
> when imaging fluorescent beads, it creates 'star-like' spots at the edge of
> the lens, which is somewhat expected since high NA and mag lenses do not
> behave good when the FOV is large.
>
>
> However, what surprised me was other interesting findings. I found a
> collection of Zemax model files online for Nikon's various objective lens
> designs, one of which is a 60x/1.49Oil. So I took the file and ran some
> simple simulations:
>
>
> * Surfaces: the objective+tube lens (set 150 mm away from the last surface
> of the objective lens).
> * Settings:
> * I put the object on the coverslip;
> * I set the coverslip thickness of 0.170 mm, with refractive index nd=
> 1.523, and abbe number Vd=55, and I set the oil refractive index nd = 1.518,
> and vd=41 (got from Thorlabs website), with varying thickness;
> * I also set the distance between the tube lens and the image plane to be a
> variable so that I can run the optimization for smallest RMS wavefront
> aberration or spot radius (reference to the Centroid) on the image plane;
> * I ran the optimization for the smallest RMS wavefront aberration, with
> above two variables, i.e. oil thickness (related to the working distance),
> and image plane distance to the tube lens.
>
> The optimal results I got are:
>
> * When optimizing for RMS wavefront aberration:
> * Oil thickness=0.139 mm, which is close to the working distance of 0.12mm.
> * Image (or camera) plane distance = 137.08mm!
> * Spot RMS radius on the image plane = 129.8 um!
> * When optimizing for RMS spot radius:
> *
> Oil thickness=0.141mm.
> *
> Image (or camera) plane distance = 130.63mm!
> *
> Spot RMS radius on the image plane = 79.5 um!
>
> So my questions are:
>
> 1. Do these results suggest the Zemax model was not accurate? And the tube
> lens distance is supposed to be one focal length away from the camera for
> best imaging quality?
> 2. For Zemax modeling of microscopy imaging, which one should be a more
> practical figure of merit when optimizing the design, the RMS spot radius or
> RMS wavefront aberration? What properties/performance parameters would you
> suggest I look for when doing the zemax modeling?
> 3. As the sample is deviating from the coverslip, should I expect seeing
> more aberration or less aberration at the edge of the FOV?
>
> For your information, I also simulated when the sample/object is 0.01mm from
> the coverslip, with embedded medium nd=1.472, and Vd=41. The corresponding
> results are:
>
> * When optimizing for RMS wavefront aberration:
> * Oil thickness=0.119mm.
> * Image (or camera) plane distance = 177.16mm!
> * Spot RMS radius on the image plane = 80.1 um!
> * When optimizing for RMS spot radius:
> * Oil thickness=0.141mm.
> * Image (or camera) plane distance = 177.16mm
> * Spot RMS radius on the image plane = 80.8 um
> * SAME AS RMS WAVEFRONT RESULTS.
>
> This bugged me for several days. The required image/camera plane became 177
> mm, from ~140 mm. Was this simply because the inaccurate Zemax model? Should
> we believe that the objective lenses are designed to have the best imaging
> quality when the camera is placed one focal length from the tube lens? How
> sensitive practically speaking, is the image quality to the distance between
> the sample and the coverslip?
>
> Any input from the list is welcome, and thank you all in advance.
>
> Have a nice weekend,
> Lu
> "

Lu Yan

Re: Nikon 60x/1.49NA lens tube lens distance to camera

In reply to this post by Zdenek Svindrych-2

*****
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.
*****

Hi zdenek,

Thanks very much for your reply. The files that I got were actually from the authors of the archive paper you included:

https://figshare.com/articles/Zemax_optical_design_files_of_microscope_objectives_tube_lenses_and_Fourier_imaging_setups/1481270

I thought the model was correct considering the amount of work they did, and when the oil and coverslip indices matched (1.523@589.3 nm), the lens did yield close to diffraction limited performance in terms of spot size, i.e. I got RMS spot radius of 7.3 um, but again, the 'optimal' tube lens distance (to the camera) was 172.4 mm. This gave me some confidence on the modeling.

Weird things started happening when there was some practical mismatch between the oil/coverslip, and coverslip/mounting medium, and I got the those results I listed in my original post, which confused me a lot. Especially the spot size is way larger than that of index-matched case.

For the Zemax simulation, can you suggest a practical way of simulating the FOV of the scope? Is the 'Image Simulation' module a good way of doing it? What kind of object image should I use? There are really some confusing parameter settings there... (like Field Height & Field etc).

Thanks again,

Lu

Zemax (optical design) files of microscope objectives, tube lenses, and Fourier imaging setups<https://figshare.com/articles/Zemax_optical_design_files_of_microscope_objectives_tube_lenses_and_Fourier_imaging_setups/1481270>
figshare.com
This fileset contains Zemax files related to the manuscript entitled “Comparative analysis of imaging configurations and objectives for Fourier microscopy” by Jonathan A. Kurvits, Mingming Jiang, and Rashid Zia. Specifically, it includes multi-configuration Zemax files that allow for the comparison of different Fourier microscopy imaging configurations as well as Zemax lens files and lens catalogs for microscope objectives and tube lenses, as inferred from the following published patents: 1) U.S. Patent # 5,517,360; assignee: Olympus; objective description: 60x, 1.4 NA Plan Apo. 2) U.S. Patent # 5,659,425; assignee: Olympus; objective description: 100x, 1.65 NA Apo. 3) U.S. Patent # 6,504,563; assignee: Zeiss; objective description: 100x, 1.45 NA TIRF. 4) U.S. Patent # 6,519,092; assignee: Nikon; objective description: 60x, 1.4 NA Plan Apo. 5) U.S. Patent # 6,519,092; assignee: Nikon; objective description: 100x, 1.4 NA Plan Apo. 6) U.S. Patent # 7,046,451; assignee: Nikon; objective description: 60x, 1.5

------------------------------------------
Lu Yan
Nanostructured Fibers and Nonlinear Optics Laboratory
Electrical and Computer Engineering
Boston University
8 Saint Mary's Steet, Rm. 505, Boston, MA 02215
+1.617.353.0286 (office)
[hidden email]
------------------------------------------

________________________________
From: Confocal Microscopy List <[hidden email]> on behalf of [hidden email] <[hidden email]>
Sent: Friday, October 13, 2017 10:14 PM
To: [hidden email]
Subject: Re: Nikon 60x/1.49NA lens tube lens distance to camera

*****
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.
*****

Hi Lu,
clearly there is something wrong with your model, 80 um spot radius in the
image plane is way too much, I would expect less than 20 um.

The lens prescriptions usually come from patents, so the actual production
lens might be slightly different. Anyway, I would check the original patent
to see if your model is correct and at what conditions (coverslip, oil,
wavelengths, tubelens) the lens is supposed to work OK.

https://patents.google.com/patent/US7046451B2/en
Immersion microscope objective lens - Google Patents<https://patents.google.com/patent/US7046451B2/en>
patents.google.com
To provide an immersion microscope objective lens capable of making a numerical aperture NA larger than 1.45 even if an ordinary oil is used, the objective lens ...

Also you may check the zemax files associated with this paper:

https://arxiv.org/abs/1507.04037
Title: Comparative analysis of imaging configurations and ...<https://arxiv.org/abs/1507.04037>
arxiv.org
Abstract: Fourier microscopy is becoming an increasingly important tool for the analysis of optical nanostructures and quantum emitters. However, achieving ...

where the authors did some modelling of high-NA lenses.

As to what parameters are important for modeling... what is your
application? For TIRF microscopy there will be design criteria related to
the TIRF illumination; when it comes to resolution you should put more
weight on the high-NA rays, as quite big part of your signal actually comes
from supercritical angle fluorescence. For multiphoton imaging you might
want to optimize for Strehl ratio (the peak intensity in the focus), etc...

Btw, our old Olympus 60x / 1.45 NA TIRFM lens also shows quite strong
aberrations near the edges of the FOV.

Hope it helps. Best, zdenek

--
Zdenek Svindrych, Ph.D.
W.M. Keck Center for Cellular Imaging (PLSB 003)
Department of Biology,University of Virginia
409 McCormick Rd, Charlottesville, VA-22904
http://www.kcci.virginia.edu/
W.M. Keck Center for Cellular Imaging, U.Va.<http://www.kcci.virginia.edu/>
www.kcci.virginia.edu
The primary goal of the W.M. Keck Center for Cellular Imaging is to provide a state of the art optical imaging facility to enhance both the research and teaching ...

tel: 434-982-4869

---------- Původní e-mail ----------
Od: Yan, Lu <[hidden email]>
Komu: [hidden email]
Datum: 13. 10. 2017 21:16:24
Předmět: Nikon 60x/1.49NA lens tube lens distance to camera
"*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
LISTSERV 16.0 - CONFOCALMICROSCOPY List at LISTS.UMN.EDU<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>
lists.umn.edu
[hidden email]: listserv archives. confocalmicroscopy

Post images on http://www.imgur.com and include the link in your posting.
[http://upload.wikimedia.org/wikipedia/commons/thumb/e/e9/Imgur_logo.svg/150px-Imgur_logo.svg.png]<http://www.imgur.com/>

Imgur: The most awesome images on the Internet<http://www.imgur.com/>
www.imgur.com
Imgur is the best place to share and enjoy the most awesome images on the Internet. Every day, millions of people use Imgur to be entertained and inspired by funny ...

*****

Dear listers,

I hope some from the list could enlighten me. We recently purchased a Nikon
60X/1.49Oil TIRF lens, and I have been playing with it a bit, and found that
when imaging fluorescent beads, it creates 'star-like' spots at the edge of
the lens, which is somewhat expected since high NA and mag lenses do not
behave good when the FOV is large.

However, what surprised me was other interesting findings. I found a
collection of Zemax model files online for Nikon's various objective lens
designs, one of which is a 60x/1.49Oil. So I took the file and ran some
simple simulations:

* Surfaces: the objective+tube lens (set 150 mm away from the last surface
of the objective lens).
* Settings:
* I put the object on the coverslip;
* I set the coverslip thickness of 0.170 mm, with refractive index nd=
1.523, and abbe number Vd=55, and I set the oil refractive index nd = 1.518,
and vd=41 (got from Thorlabs website), with varying thickness;
* I also set the distance between the tube lens and the image plane to be a
variable so that I can run the optimization for smallest RMS wavefront
aberration or spot radius (reference to the Centroid) on the image plane;
* I ran the optimization for the smallest RMS wavefront aberration, with
above two variables, i.e. oil thickness (related to the working distance),
and image plane distance to the tube lens.

The optimal results I got are:

* When optimizing for RMS wavefront aberration:
* Oil thickness=0.139 mm, which is close to the working distance of 0.12mm.
* Image (or camera) plane distance = 137.08mm!
* Spot RMS radius on the image plane = 129.8 um!
* When optimizing for RMS spot radius:
*
Oil thickness=0.141mm.
*
Image (or camera) plane distance = 130.63mm!
*
Spot RMS radius on the image plane = 79.5 um!

So my questions are:

1. Do these results suggest the Zemax model was not accurate? And the tube
lens distance is supposed to be one focal length away from the camera for
best imaging quality?
2. For Zemax modeling of microscopy imaging, which one should be a more
practical figure of merit when optimizing the design, the RMS spot radius or
RMS wavefront aberration? What properties/performance parameters would you
suggest I look for when doing the zemax modeling?
3. As the sample is deviating from the coverslip, should I expect seeing
more aberration or less aberration at the edge of the FOV?

For your information, I also simulated when the sample/object is 0.01mm from
the coverslip, with embedded medium nd=1.472, and Vd=41. The corresponding
results are:

* When optimizing for RMS wavefront aberration:
* Oil thickness=0.119mm.
* Image (or camera) plane distance = 177.16mm!
* Spot RMS radius on the image plane = 80.1 um!
* When optimizing for RMS spot radius:
* Oil thickness=0.141mm.
* Image (or camera) plane distance = 177.16mm
* Spot RMS radius on the image plane = 80.8 um
* SAME AS RMS WAVEFRONT RESULTS.

This bugged me for several days. The required image/camera plane became 177
mm, from ~140 mm. Was this simply because the inaccurate Zemax model? Should
we believe that the objective lenses are designed to have the best imaging
quality when the camera is placed one focal length from the tube lens? How
sensitive practically speaking, is the image quality to the distance between
the sample and the coverslip?

Any input from the list is welcome, and thank you all in advance.

Have a nice weekend,
Lu
"

Lu Yan

Re: Nikon 60x/1.49NA lens tube lens distance to camera

In reply to this post by John Oreopoulos

*****
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.
*****

Hi John,

I used the model I got from here:

https://figshare.com/articles/Zemax_optical_design_files_of_microscope_objectives_tube_lenses_and_Fourier_imaging_setups/1481270

Zemax (optical design) files of microscope objectives, tube lenses, and Fourier imaging setups<https://figshare.com/articles/Zemax_optical_design_files_of_microscope_objectives_tube_lenses_and_Fourier_imaging_setups/1481270>
figshare.com
This fileset contains Zemax files related to the manuscript entitled “Comparative analysis of imaging configurations and objectives for Fourier microscopy” by Jonathan A. Kurvits, Mingming Jiang, and Rashid Zia. Specifically, it includes multi-configuration Zemax files that allow for the comparison of different Fourier microscopy imaging configurations as well as Zemax lens files and lens catalogs for microscope objectives and tube lenses, as inferred from the following published patents: 1) U.S. Patent # 5,517,360; assignee: Olympus; objective description: 60x, 1.4 NA Plan Apo. 2) U.S. Patent # 5,659,425; assignee: Olympus; objective description: 100x, 1.65 NA Apo. 3) U.S. Patent # 6,504,563; assignee: Zeiss; objective description: 100x, 1.45 NA TIRF. 4) U.S. Patent # 6,519,092; assignee: Nikon; objective description: 60x, 1.4 NA Plan Apo. 5) U.S. Patent # 6,519,092; assignee: Nikon; objective description: 100x, 1.4 NA Plan Apo. 6) U.S. Patent # 7,046,451; assignee: Nikon; objective description: 60x, 1.5

, i.e. from the paper that Zdenek included in his reply.

When I had everything perfectly matched (refractive index), I got a RMS spot radius of 7.2 um at 690 nm, which is close to the diffraction limit. It starts becoming unreasonable when the index of the oil does not match that of the coverslip. I had hard time imagining this slight mismatch will cause so much deviation, if the modeling was correct.

Thanks,

Lu

------------------------------------------
Lu Yan
Nanostructured Fibers and Nonlinear Optics Laboratory
Electrical and Computer Engineering
Boston University
8 Saint Mary's Steet, Rm. 505, Boston, MA 02215
+1.617.353.0286 (office)
[hidden email]
------------------------------------------

________________________________
From: Confocal Microscopy List <[hidden email]> on behalf of John Oreopoulos <[hidden email]>
Sent: Friday, October 13, 2017 10:53 PM
To: [hidden email]
Subject: Re: Nikon 60x/1.49NA lens tube lens distance to camera

*****
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 Lu,
>
> There's a lot to unpack there, so let me just ask a simple question first: How are you modeling the object to be imaged by the objective in Zemax?
>
> You quoted and RMS spot size of something like 80 um, but a rough calculation based on the NA of the objective and assuming a green emission wavelength of 525 nm, I calculate the Rayleigh resolution as 176nm (lambda/2NA). So, the diameter of the Airy disk (which has some correspondence to the RMS spot size) should be about 21 um under ideal conditions. Hard to see how the objective would yield a spot size that is roughly 4x bigger than what it should be (the diffraction limit), even for an object at the centre of the field of view. I think there would be very many unhappy customers if the objectives really behaved like that. So I take this to mean there is something not quite right with how you've set up the full ray tracing model, and hence my question about the light emitting object.
>
> Maybe take a step backwards and make sure you get sensible answers by modeling the objective and tube lenses as thin paraxial ideal lenses first. Then you'll know if the object model is correct.
>

John Oreopoulos

On 2017-10-13, at 9:13 PM, Yan, Lu wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
LISTSERV 16.0 - CONFOCALMICROSCOPY List at LISTS.UMN.EDU<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>
lists.umn.edu
[hidden email]: listserv archives. confocalmicroscopy

> Post images on http://www.imgur.com and include the link in your posting.
[http://upload.wikimedia.org/wikipedia/commons/thumb/e/e9/Imgur_logo.svg/150px-Imgur_logo.svg.png]<http://www.imgur.com/>

Imgur: The most awesome images on the Internet<http://www.imgur.com/>
www.imgur.com
Imgur is the best place to share and enjoy the most awesome images on the Internet. Every day, millions of people use Imgur to be entertained and inspired by funny ...

> *****
>
> Dear listers,
>
>
> I hope some from the list could enlighten me. We recently purchased a Nikon 60X/1.49Oil TIRF lens, and I have been playing with it a bit, and found that when imaging fluorescent beads, it creates 'star-like' spots at the edge of the lens, which is somewhat expected since high NA and mag lenses do not behave good when the FOV is large.
>
>
> However, what surprised me was other interesting findings. I found a collection of Zemax model files online for Nikon's various objective lens designs, one of which is a 60x/1.49Oil. So I took the file and ran some simple simulations:
>
>
> * Surfaces: the objective+tube lens (set 150 mm away from the last surface of the objective lens).
> * Settings:
> * I put the object on the coverslip;
> * I set the coverslip thickness of 0.170 mm, with refractive index nd= 1.523, and abbe number Vd=55, and I set the oil refractive index nd = 1.518, and vd=41 (got from Thorlabs website), with varying thickness;
> * I also set the distance between the tube lens and the image plane to be a variable so that I can run the optimization for smallest RMS wavefront aberration or spot radius (reference to the Centroid) on the image plane;
> * I ran the optimization for the smallest RMS wavefront aberration, with above two variables, i.e. oil thickness (related to the working distance), and image plane distance to the tube lens.
>
> The optimal results I got are:
>
> * When optimizing for RMS wavefront aberration:
> * Oil thickness=0.139 mm, which is close to the working distance of 0.12mm.
> * Image (or camera) plane distance = 137.08mm!
> * Spot RMS radius on the image plane = 129.8 um!
> * When optimizing for RMS spot radius:
> *
> Oil thickness=0.141mm.
> *
> Image (or camera) plane distance = 130.63mm!
> *
> Spot RMS radius on the image plane = 79.5 um!
>
> So my questions are:
>
> 1. Do these results suggest the Zemax model was not accurate? And the tube lens distance is supposed to be one focal length away from the camera for best imaging quality?
> 2. For Zemax modeling of microscopy imaging, which one should be a more practical figure of merit when optimizing the design, the RMS spot radius or RMS wavefront aberration? What properties/performance parameters would you suggest I look for when doing the zemax modeling?
> 3. As the sample is deviating from the coverslip, should I expect seeing more aberration or less aberration at the edge of the FOV?
>
> For your information, I also simulated when the sample/object is 0.01mm from the coverslip, with embedded medium nd=1.472, and Vd=41. The corresponding results are:
>
> * When optimizing for RMS wavefront aberration:
> * Oil thickness=0.119mm.
> * Image (or camera) plane distance = 177.16mm!
> * Spot RMS radius on the image plane = 80.1 um!
> * When optimizing for RMS spot radius:
> * Oil thickness=0.141mm.
> * Image (or camera) plane distance = 177.16mm
> * Spot RMS radius on the image plane = 80.8 um
> * SAME AS RMS WAVEFRONT RESULTS.
>
> This bugged me for several days. The required image/camera plane became 177 mm, from ~140 mm. Was this simply because the inaccurate Zemax model? Should we believe that the objective lenses are designed to have the best imaging quality when the camera is placed one focal length from the tube lens? How sensitive practically speaking, is the image quality to the distance between the sample and the coverslip?
>
> Any input from the list is welcome, and thank you all in advance.
>
> Have a nice weekend,
> Lu

Rowlands, Christopher J

Re: Nikon 60x/1.49NA lens tube lens distance to camera

*****
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.
*****

That objective has a correction collar - do you know which lens group it moves, and does moving that group correct for the aberration? I'm not very surprised that the aberrations are so bad when you change the refractive index of the oil - 1.49 is a very high NA, and without correction you'll likely get pronounced spherical aberration.

Chris

-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Yan, Lu
Sent: 16 October 2017 21:38
To: [hidden email]
Subject: Re: Nikon 60x/1.49NA lens tube lens distance to camera

*****
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.
*****

Hi John,

I used the model I got from here:

https://figshare.com/articles/Zemax_optical_design_files_of_microscope_objectives_tube_lenses_and_Fourier_imaging_setups/1481270

Zemax (optical design) files of microscope objectives, tube lenses, and Fourier imaging setups<https://figshare.com/articles/Zemax_optical_design_files_of_microscope_objectives_tube_lenses_and_Fourier_imaging_setups/1481270>
figshare.com
This fileset contains Zemax files related to the manuscript entitled "Comparative analysis of imaging configurations and objectives for Fourier microscopy" by Jonathan A. Kurvits, Mingming Jiang, and Rashid Zia. Specifically, it includes multi-configuration Zemax files that allow for the comparison of different Fourier microscopy imaging configurations as well as Zemax lens files and lens catalogs for microscope objectives and tube lenses, as inferred from the following published patents: 1) U.S. Patent # 5,517,360; assignee: Olympus; objective description: 60x, 1.4 NA Plan Apo. 2) U.S. Patent # 5,659,425; assignee: Olympus; objective description: 100x, 1.65 NA Apo. 3) U.S. Patent # 6,504,563; assignee: Zeiss; objective description: 100x, 1.45 NA TIRF. 4) U.S. Patent # 6,519,092; assignee: Nikon; objective description: 60x, 1.4 NA Plan Apo. 5) U.S. Patent # 6,519,092; assignee: Nikon; objective description: 100x, 1.4 NA Plan Apo. 6) U.S. Patent # 7,046,451; assignee: Nikon; objective description: 60x, 1.5

, i.e. from the paper that Zdenek included in his reply.

When I had everything perfectly matched (refractive index), I got a RMS spot radius of 7.2 um at 690 nm, which is close to the diffraction limit. It starts becoming unreasonable when the index of the oil does not match that of the coverslip. I had hard time imagining this slight mismatch will cause so much deviation, if the modeling was correct.

Thanks,

Lu

------------------------------------------
Lu Yan
Nanostructured Fibers and Nonlinear Optics Laboratory Electrical and Computer Engineering Boston University
8 Saint Mary's Steet, Rm. 505, Boston, MA 02215
+1.617.353.0286 (office)
[hidden email]
------------------------------------------

________________________________
From: Confocal Microscopy List <[hidden email]> on behalf of John Oreopoulos <[hidden email]>
Sent: Friday, October 13, 2017 10:53 PM
To: [hidden email]
Subject: Re: Nikon 60x/1.49NA lens tube lens distance to camera

*****
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 Lu,
>
> There's a lot to unpack there, so let me just ask a simple question first: How are you modeling the object to be imaged by the objective in Zemax?
>
> You quoted and RMS spot size of something like 80 um, but a rough calculation based on the NA of the objective and assuming a green emission wavelength of 525 nm, I calculate the Rayleigh resolution as 176nm (lambda/2NA). So, the diameter of the Airy disk (which has some correspondence to the RMS spot size) should be about 21 um under ideal conditions. Hard to see how the objective would yield a spot size that is roughly 4x bigger than what it should be (the diffraction limit), even for an object at the centre of the field of view. I think there would be very many unhappy customers if the objectives really behaved like that. So I take this to mean there is something not quite right with how you've set up the full ray tracing model, and hence my question about the light emitting object.
>
> Maybe take a step backwards and make sure you get sensible answers by modeling the objective and tube lenses as thin paraxial ideal lenses first. Then you'll know if the object model is correct.
>

John Oreopoulos

On 2017-10-13, at 9:13 PM, Yan, Lu wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
LISTSERV 16.0 - CONFOCALMICROSCOPY List at LISTS.UMN.EDU<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>
lists.umn.edu
[hidden email]: listserv archives. confocalmicroscopy

> Post images on http://www.imgur.com and include the link in your posting.
[http://upload.wikimedia.org/wikipedia/commons/thumb/e/e9/Imgur_logo.svg/150px-Imgur_logo.svg.png]<http://www.imgur.com/>

Imgur: The most awesome images on the Internet<http://www.imgur.com/> www.imgur.com Imgur is the best place to share and enjoy the most awesome images on the Internet. Every day, millions of people use Imgur to be entertained and inspired by funny ...

James Pawley

Re: Nikon 60x/1.49NA lens tube lens distance to camera

*****
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.
*****

Two points.

1) Very small changes (2µm) in coverslip thickness make a clearly evident asymmetry in the z-azis thru-focus response of NA 1.2, Zeiss C-Apo water ( Handbook,p 408 Fig 20-3). This would be much worse at NA 1.49.

2) The NA 1.49 TIRF lens is a special-purpose lens in which it is not clear that the high-NA ray paths are to be used for anything but illumination. It is not clear that these rays are even designed to provide SA-corrected imaging ray paths (and remember SA correction (like chromatic correction) is really only correct at certain wavelengths). Another feature of is being a TIRF lens is that is is designed to work best when imaging a planar surface with a known coverslip thickness AND oil thickness. At no other z-position will there be any fluorescence to image.

Of a practical level, careful measurements of PSF by Rimas Juskaitis in Chapter 11 of the same Handbook, showed that even the “best” direct-from-factory NA 1.4 oil lens did not provide phase-corrected rays beyond NA1.35. (i.e., it is not uncommon for very good lenses not to be perfect). Keep in mind that a definition of “Diffraction limited” is that all “accepted" rays must converge to a point in the image plane with less than 1/4 wave of phase difference (some would say 1/10 wave). Although one can usually make the spherical surfaces round with almost the correct accuracy, errors in component thickness, mounting and centring mount up. As a result, the ratio of the brightness of the peak to that of the first bright ring is seldom equal to the 25:1 ratio you might expect from the math. (i.e., They are not really “diffraction limited.”)

(http://www.astro.caltech.edu/~lah/ay105/pdf/Fundamental-Optics.pdf (Figure 1.28).

I confess to knowing nothing about the details of Zemax, but I can ask you if the model values put in for this objective are accurate in position to less than 1/4wavelength? 1/10th wavelength? What about the RI number?

In practice the manufacturers make dozens of more-or-less identical component lenses and then very patient workers (usually women!) sit in dark rooms patiently trying one combination then another until some level of perfection (i.e., some arrangements that cancels out at least some of the errors) is reached. This is level is set not so much by mathematics as by our willingness to pay for the resulting lens.

Can’t imagine that ZeMax can take all this into account.

Happy calculating!

JP

James and Christine Pawley, 5446 Burley Place, Box 2348, Sechelt BC, Canada, V0N3A0 [hidden email]<mailto:[hidden email]>, Phone 1-604-885-0840, cell 1-604-989-6146

On Oct 16, 2017, at 1:45 PM, Rowlands, Christopher J <[hidden email]<mailto:[hidden email]>> wrote:

*****
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.
*****

That objective has a correction collar - do you know which lens group it moves, and does moving that group correct for the aberration? I'm not very surprised that the aberrations are so bad when you change the refractive index of the oil - 1.49 is a very high NA, and without correction you'll likely get pronounced spherical aberration.

Chris

-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Yan, Lu
Sent: 16 October 2017 21:38
To: [hidden email]
Subject: Re: Nikon 60x/1.49NA lens tube lens distance to camera

*****
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.
*****

Hi John,

I used the model I got from here:

https://figshare.com/articles/Zemax_optical_design_files_of_microscope_objectives_tube_lenses_and_Fourier_imaging_setups/1481270

Zemax (optical design) files of microscope objectives, tube lenses, and Fourier imaging setups<https://figshare.com/articles/Zemax_optical_design_files_of_microscope_objectives_tube_lenses_and_Fourier_imaging_setups/1481270>
figshare.com
This fileset contains Zemax files related to the manuscript entitled "Comparative analysis of imaging configurations and objectives for Fourier microscopy" by Jonathan A. Kurvits, Mingming Jiang, and Rashid Zia. Specifically, it includes multi-configuration Zemax files that allow for the comparison of different Fourier microscopy imaging configurations as well as Zemax lens files and lens catalogs for microscope objectives and tube lenses, as inferred from the following published patents: 1) U.S. Patent # 5,517,360; assignee: Olympus; objective description: 60x, 1.4 NA Plan Apo. 2) U.S. Patent # 5,659,425; assignee: Olympus; objective description: 100x, 1.65 NA Apo. 3) U.S. Patent # 6,504,563; assignee: Zeiss; objective description: 100x, 1.45 NA TIRF. 4) U.S. Patent # 6,519,092; assignee: Nikon; objective description: 60x, 1.4 NA Plan Apo. 5) U.S. Patent # 6,519,092; assignee: Nikon; objective description: 100x, 1.4 NA Plan Apo. 6) U.S. Patent # 7,046,451; assignee: Nikon; objective description: 60x, 1.5

, i.e. from the paper that Zdenek included in his reply.

When I had everything perfectly matched (refractive index), I got a RMS spot radius of 7.2 um at 690 nm, which is close to the diffraction limit. It starts becoming unreasonable when the index of the oil does not match that of the coverslip. I had hard time imagining this slight mismatch will cause so much deviation, if the modeling was correct.

Thanks,

Lu

------------------------------------------
Lu Yan
Nanostructured Fibers and Nonlinear Optics Laboratory Electrical and Computer Engineering Boston University
8 Saint Mary's Steet, Rm. 505, Boston, MA 02215
+1.617.353.0286 (office)
[hidden email]
------------------------------------------

________________________________
From: Confocal Microscopy List <[hidden email]> on behalf of John Oreopoulos <[hidden email]>
Sent: Friday, October 13, 2017 10:53 PM
To: [hidden email]
Subject: Re: Nikon 60x/1.49NA lens tube lens distance to camera

*****
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 Lu,

There's a lot to unpack there, so let me just ask a simple question first: How are you modeling the object to be imaged by the objective in Zemax?

You quoted and RMS spot size of something like 80 um, but a rough calculation based on the NA of the objective and assuming a green emission wavelength of 525 nm, I calculate the Rayleigh resolution as 176nm (lambda/2NA). So, the diameter of the Airy disk (which has some correspondence to the RMS spot size) should be about 21 um under ideal conditions. Hard to see how the objective would yield a spot size that is roughly 4x bigger than what it should be (the diffraction limit), even for an object at the centre of the field of view. I think there would be very many unhappy customers if the objectives really behaved like that. So I take this to mean there is something not quite right with how you've set up the full ray tracing model, and hence my question about the light emitting object.

Maybe take a step backwards and make sure you get sensible answers by modeling the objective and tube lenses as thin paraxial ideal lenses first. Then you'll know if the object model is correct.

John Oreopoulos

On 2017-10-13, at 9:13 PM, Yan, Lu wrote:

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
LISTSERV 16.0 - CONFOCALMICROSCOPY List at LISTS.UMN.EDU<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>
lists.umn.edu
[hidden email]: listserv archives. confocalmicroscopy

Post images on http://www.imgur.com and include the link in your posting.
[http://upload.wikimedia.org/wikipedia/commons/thumb/e/e9/Imgur_logo.svg/150px-Imgur_logo.svg.png]<http://www.imgur.com/>

Imgur: The most awesome images on the Internet<http://www.imgur.com/> www.imgur.com Imgur is the best place to share and enjoy the most awesome images on the Internet. Every day, millions of people use Imgur to be entertained and inspired by funny ...

*****

Dear listers,

I hope some from the list could enlighten me. We recently purchased a Nikon 60X/1.49Oil TIRF lens, and I have been playing with it a bit, and found that when imaging fluorescent beads, it creates 'star-like' spots at the edge of the lens, which is somewhat expected since high NA and mag lenses do not behave good when the FOV is large.

However, what surprised me was other interesting findings. I found a collection of Zemax model files online for Nikon's various objective lens designs, one of which is a 60x/1.49Oil. So I took the file and ran some simple simulations:

* Surfaces: the objective+tube lens (set 150 mm away from the last surface of the objective lens).
* Settings:
* I put the object on the coverslip;
* I set the coverslip thickness of 0.170 mm, with refractive index nd= 1.523, and abbe number Vd=55, and I set the oil refractive index nd = 1.518, and vd=41 (got from Thorlabs website), with varying thickness;
* I also set the distance between the tube lens and the image plane to be a variable so that I can run the optimization for smallest RMS wavefront aberration or spot radius (reference to the Centroid) on the image plane;
* I ran the optimization for the smallest RMS wavefront aberration, with above two variables, i.e. oil thickness (related to the working distance), and image plane distance to the tube lens.

The optimal results I got are:

* When optimizing for RMS wavefront aberration:
* Oil thickness=0.139 mm, which is close to the working distance of 0.12mm.
* Image (or camera) plane distance = 137.08mm!
* Spot RMS radius on the image plane = 129.8 um!
* When optimizing for RMS spot radius:
*
Oil thickness=0.141mm.
*
Image (or camera) plane distance = 130.63mm!
*
Spot RMS radius on the image plane = 79.5 um!

So my questions are:

1. Do these results suggest the Zemax model was not accurate? And the tube lens distance is supposed to be one focal length away from the camera for best imaging quality?
2. For Zemax modeling of microscopy imaging, which one should be a more practical figure of merit when optimizing the design, the RMS spot radius or RMS wavefront aberration? What properties/performance parameters would you suggest I look for when doing the zemax modeling?
3. As the sample is deviating from the coverslip, should I expect seeing more aberration or less aberration at the edge of the FOV?

For your information, I also simulated when the sample/object is 0.01mm from the coverslip, with embedded medium nd=1.472, and Vd=41. The corresponding results are:

* When optimizing for RMS wavefront aberration:
* Oil thickness=0.119mm.
* Image (or camera) plane distance = 177.16mm!
* Spot RMS radius on the image plane = 80.1 um!
* When optimizing for RMS spot radius:
* Oil thickness=0.141mm.
* Image (or camera) plane distance = 177.16mm
* Spot RMS radius on the image plane = 80.8 um
* SAME AS RMS WAVEFRONT RESULTS.

This bugged me for several days. The required image/camera plane became 177 mm, from ~140 mm. Was this simply because the inaccurate Zemax model? Should we believe that the objective lenses are designed to have the best imaging quality when the camera is placed one focal length from the tube lens? How sensitive practically speaking, is the image quality to the distance between the sample and the coverslip?

Any input from the list is welcome, and thank you all in advance.

Have a nice weekend,
Lu