OPD definition
laurent.pares
path difference (OPD) :
should it be measured along a real ray ( normal to the wavefront
surface)
or
should it be measured along a path normal to the reference sphere.
One could find a scheme on
http://www.sinopt.com/software/usrguide54/evaluate/raytrace.html
laurent.pares
path difference (OPD) :
should it be measured along a real ray ( normal to the wavefront
surface)
or
should it be measured along a path normal to the reference sphere.
One could find a scheme on
http://www.sinopt.com/software1/usrguide54/evaluate/raytrace.htm
Acme Optics
This is how I understand it.
>
The difference between the optical path length of a general ray and
the optical path length of the chief ray (one that goes through the
center of the stop surface), referenced to their intersections with
the reference sphere. The reference sphere is a sphere who's center is
located where the chief ray intersects the image surface and who's
radius is equal to the distance from the reference sphere center to
the location of the reference ray and the exit pupil. The exit pupil
is located where a close differential chief ray, referenced to the
reference ray either intersects the reference ray or makes it's
closest approach to the reference ray (in the case of pupil abbe
ration).
One can then also then fit this wavefront, defined by a grid of
general rays, to a zernike polynomial (provided you have a circular
aperture) and then remove the tilt term (LOS) and the focus term
(defocus) if you want.
See the documentation in the free Roadrunner manual
(www.acmeoptics.com)
Jim Klein
>
>
>
Repeating Decimal
wrote on 6/3/02 1:25 AM:
If you were to take the Feynman approach, you would take an arbitrary path.
Find the pathlength along that path Lp. Convert that to a phase
phip=Lp*2*pi/lamda. Then add up the amplitudes exp(i*phip) over ALL possible
paths p.
For most paths these amplitudes will cancel out and add up to zero. There
will be a net contribution from paths near the true path. It is this
effective phase that contributes to the optical path measured in phase.
For large dimensions, this means that the pathlength should be measured
along the geometrical path.
Bill
Bob May
rays going through the different parts of the optics. Normally it is
referenced to the chief ray which is the center ray of the system.
--
Bob May
Imagine the terrorist's fun when they realize that their 72 "nubile virgins"
are all lesbians and cranky from it being that time of the month!
Ted
Shack the same question a few years ago and he was specific on that point.
Ted Thall
"laurent.pares" <lauren...@free.fr> wrote in message
news:3CFB26DC...@free.fr...
Ted
asked Dr. Roland Shack the same question a few years ago and he was specific
on that point
Ted Thall
"laurent.pares" <lauren...@free.fr> wrote in message
news:3CFB26DC...@free.fr...
Steve Eckhardt
>
>I would like to know what is the accepted definition for the optical
>path difference (OPD) :
>
> should it be measured along a real ray ( normal to the wavefront
>surface)
>>or
>> should it be measured along a path normal to the reference sphere.
For meridional rays, the answer is straightforward, but for skew rays that do
not intersect the chief ray in the image plane, there is a bit of a problem.
The primary reference to consult for this is "The Wave Aberration Associated
with Skew Rays" by H.H. Hopkins in Proc. Phys. Soc. B, vol. 65, pp 934-942
(1952). Unfortunately, there is an error in the focal shift formula. This was
detected and corrected by Berge Tatian of Itek. His article is "A comment on
the wave aberration formula of H. H. Hopkins" in Optica Acta vol 19, pp. 79-83
(1972). I believe that Berge's work is used in all modern lens design
programs.
--
Best regards,
Steve Eckhardt
skeck...@mmm.com
Opinions expressed herein are my own and may not represent those of my employer.
Jim Klein
>In article <3CFB27DF...@free.fr>, lauren...@free.fr says...
>>
>>I would like to know what is the accepted definition for the optical
>>path difference (OPD) :
>>
>> should it be measured along a real ray ( normal to the wavefront
>>surface)
>>>or
>>> should it be measured along a path normal to the reference sphere.
>
>For meridional rays, the answer is straightforward, but for skew rays that do
>not intersect the chief ray in the image plane, there is a bit of a problem.
>The primary reference to consult for this is "The Wave Aberration Associated
>with Skew Rays" by H.H. Hopkins in Proc. Phys. Soc. B, vol. 65, pp 934-942
>(1952). Unfortunately, there is an error in the focal shift formula. This was
>detected and corrected by Berge Tatian of Itek. His article is "A comment on
>the wave aberration formula of H. H. Hopkins" in Optica Acta vol 19, pp. 79-83
>(1972). I believe that Berge's work is used in all modern lens design
>programs.
When a ray needs to intersect the chief ray to define a point and it
almost intersects it but not quite, I find the point of closest
approach and use it. Seems to work fine and the OPDs agree with CODE-V
so I'm happy.
Jim Klein
Jim Klein
Oh, I also provide a switch to force the use of the paraxial pupil
position or to force the use of an infinite reference sphere as in
OSLO.
Jim
Douglas Sinclair
> position or to force the use of an infinite reference sphere as in
> OSLO.
>
> Jim
>
OSLO is not at infinity. Please see the reference cited in the original
post.
-Doug
Acme Optics
Hi Doug,
My mistake. I thought there was a switch to force the use of an
infinite radius reference sphere in OSLO. Maybe it was in CODE-V. I
heard that one of the codes had it, so I added it to.:-)
Sorry for the mistake.
Jim
Douglas Sinclair
The real answer to your question is that there isn't any accepted definition
for the OPD. The scheme that you cite produces a reasonable starting point,
but it depends on the selection of a reference ray that itself is determined
by ad hoc methods (there is no particular reason why the reference ray
should be the chief ray). In typical practice, the OPD is used to
characterize a wavefront by a single number, which is inherently impossible.
If you are computing a spot diagram, you can use the phase length along each
of the rays traced to find the position of a reference sphere that minimizes
the variance of the wavefront. This is what is done in OSLO, for example.
But it requires tracing several rays, not just a single reference ray.
Moreover, the value obtained for the OPD will depend on the radius of the
reference sphere, since an aberrated wave changes its shape as it
propagates. If you let the reference sphere have an infinite radius, you
find the nice property that the transverse ray aberrations are proportional
to the wavefront differentials, but this nicety is mathematical and not
necessarily physical. If you are concerned about physics, you have to be
more specific about what you want.
The good news about OPD is that if there isn't any, the exact definition
doesn't matter.
-Doug
Steve Eckhardt
dsin...@tampabay.rr.com says...
Doug, et al.,
To quote from Berge Tatian's article, "Note that other authors measure
the focal shift term along the radius of the reference sphere rather than along
the ray [3,4]. The resulting focal shift formula [5] is neither as simple as
Hopkins' contribution nor as exact. Worse from a theoretical point of view,
the aberration as so defined depends upon the position of the exit pupil,
although as a practical matter this dependence is usually not very
significant. All of these discrepancies become more pronounced as the exit
pupil approaches the image." I'm one of the many designers that can vouch for
this!
Tatian gives a simple, exact formula for the OPD that is independent of
the reference sphere. It is based on the mid-point of the shortest join (mpsj)
of the ray in question and the reference ray. Jim Klein appears to be doing
the right thing!
Jim Klein
<snip>
> Jim Klein appears to be doing the right thing!
Since I have made it my policy, from the very inception of KDP and now
Roadrunner to copy, as best I can, every single worthy feature of
every other optical design program; I can't help but get something
right once in a while. :-)
Jim Klein
Acme Optics
Doug Sinclair
"Steve Eckhardt" <skeck...@mmm.com.deletethis> wrote in message
news:ae2m0v$r11$1...@magnum.mmm.com...
Steve,
You're talking about something different from my point. OSLO measures the
difference along the ray, as Tatian recommends, which you can see from the
discussion cited in the laurent.peres posting, which should actually be
http://www.sinopt.com/software1/usrguide54/evaluate/raytrace.htm
(The discussion of OPD is at the end of the page). The point I was trying to
make is that the Hopkins-Tatian definition normalizes the physics out of the
problem. You can set up an interferometer to show how the wavefront changes
as the radius of the reference sphere changes; it's a real effect.
-Doug
Acme Optics
There may be quite a few descriptions of what OPD is and there are
various formulae which describe the removal of excess TILT, FOCUS or
both, from OPD.
Unfortunately, if the final resulting answer which is gotten from
(supply the name of your favorite optical design or analysis program
here) does not agree closely with the answer which comes from modeling
the same system with CODE-V, then it really doesn't matter how
rational the argument is for belief in it, as the correct answer.
If it does not agree with CODE-V, anyone who be a complete fool to
believe it was still correct. Duh!
The folks from Acme Optics, Focus Software, the creators of ACCOS-V,
the BEAM I, II, III, IV guys and yes, even OSLO and Dr. Dilworth are
held to the standards and answers given by CODE-V.
Why? Because ORA has earned the overall trust of the optical
engineering community on Planet Earth through hard work, great
advertising and absolutly wonderful user support.
Now, that said, the OPD map, predicted by a code, better look a lot
like the wavefront measured with that Zygo. :-)
Jim Klein
3a
dauphins, etc) à partir d'un avion, dans un but touristique et
scientifique.
Nous recherchons, avec nos petits moyens financiers, l'achat de lunettes
infrarouge ou autre materiel qui nous permettrait d'améliorer nos
recherches.
Si vous avez des informations à ce sujet, merci de nous laisser un
message sur notre mail
3...@addi.org
R. Varriano
OpticsNotes
I am French, I make research of marine animals (Whales, dolphins, etc)
starting from a plane, with a tourist and scientific aim. We seek, with our
small financial means, the purchase of glasses infra-red or different
materiel which would enable us to improve our research. If you have
information on this subject, thank you to leave us a message on our mall
http://babelfish.altavista.com/
-------------------------------------------------------------------------
REPLY
Can you tell us more about the application? What wavelengths, lens sizes,
cost limitations etc.?
Are you only asking about glass, or are you looking for a viewer, or
detector plane?.
----------------------------------------------------------------------------
------
Pouvez-vous nous dire davantage au sujet de l'application? Quelles longueurs
d'onde, objectif classe, les limitations etc. de coût? Êtes vous posant des
questions seulement sur le verre, ou êtes vous recherchant une visionneuse,
ou avion de détecteur?.
Bruce
http://www.OpticsNotes.Com
Optics and Photonics Tutorials, References and Resources
"3a" <3...@addi.org> wrote in message news:3D472546...@addi.org...