Montecarlo optical ray tracing
algog
I'm reproducing on Zemax-EE a reflector with a UV bandpass filter holder
in the Non-Sequential mode. The filter holder allows transmission of
up to 32% of the emitted radiation (mesured at th immediate
neighbourhood of the system). The lamp is a cylinder-volume like source
and is enclosed between the reflector and the filter holder.
The whole system would rather be a heater than an emitter!
Anyway, a problem arises when I place an aborbing detector 50 cm away
from the system (and even less!!!), because the program abruptly
interrupts the Montecarlo ray tracing routine with the message: "Not
enough intersections allocated to finish ray trace", regardless of the
pixel number along the two sides of the detector!
I can't understand why it should! Don't you think it should complete
anyway its routine (I don't care how many "few rays" I collect, since
I'm just reproducing an alredy existing arrangement with the "correct
geometry"). The detector (50 x 50 cm^2) sees the source under the
correct solid angle!!!
Moreover, if I remove the filter holder (no obscuration at all!) the ray
tracing works correctly!
Would you please help me? Where am I failing?
Thanks in advance
Algog
Mike Leveille
You need to increase the maximum number of intersections per ray value. To
do this click on the 'GEN' button on the toolbar then select the
'non-sequential' tab. You will need to adjust the value(s) in at least the
first edit box if not also the second. Try it out until your rays trace
without that error message.
Mike
"algog" <tung...@jumpy.it> wrote in message
news:c2rs3o$t6$1...@fata.cs.interbusiness.it...
algog
Sincerely
Algog
Alex
windows.
This prevent too many ray tracing as you will get while changing NSC
properties as you've advised but still will be able to get result.
Ciao, Alex
West Coast Engineering
How come we never have questions like this about CODE-V?
Because they have real user support and it goes way beyond insulting
the registered user and telling them to look in the manual .
The reason you don't see questions like this concerning Enterprise is
that I answer all questions myself, providing useful answers.
There is no excuse why an optical design program user must ask for
help from a use group rather than getting useful help from the vendor
who sold them the software in the first place.
The argument, "they are no longer under user support" is pure BS.
If I sell something to another human being, I owe it to them to not
leave them stranded. To do anything else, it is simply IMMORAL.
Ken Moore owes his livelyhood to these people. The least he can do is
answer their email.
I hope I never see the ZEMAX business model as a good thing.
WCE
elliot burke
This is not a rude answer, but a useful way to teach someone how to learn.
The question at the top of this thread is answered in the manual. It will
take more than just a few minutes of study to be proficient with non
sequential raytracing.
The other questions, that tech support should not be answering, should be
asked here, as they are not specific to a particular lens design code.
example:
what glasses should I use in a double Gauss objective?
what good are aspheres?
can lens design do something about my tendency to over-react?
regards-
Elliot B.
"West Coast Engineering" <westcoaste...@westcoastengineering.com>
wrote in message news:bv84501oo6nn63oml...@4ax.com...
Helmut Wabnig
>...........
>The other questions, that tech support should not be answering, should be
>asked here, as they are not specific to a particular lens design code.
>example:
>what glasses should I use in a double Gauss objective?
>what good are aspheres?
Exactly the kind of questions I expect to find answers
within an expensive CAD program.
Interactive content sensitive help, like Windows.
:-)
w.
Detector195
> RTFM answers 99% of the questions that tech support should be answering.
> This is not a rude answer, but a useful way to teach someone how to learn.
This is condescending. People learn by being treated with respect.
Maybe a point in the documentation needs to be explained differently.
Maybe the RTFM response is just a gatekeeper to drive away users who
are not persistent enough to demand real support.
The RTFM response might be applicable to high quality software with an
intelligible user interface and organized, readable documentation.
This is not the state of affairs with commercial optics design
software. Most of these programs have the look and feel of being
written by an expert optical engineer who happens to be an amateur
programmer. A text window ringed by a zillion tiny buttons with
unintelligible icons is not a GUI. Critical settings hidden in modal
dialogs that prevent you from seeing your entire design at once is
bush league. And so forth.
If a vendor is receiving too many naieve questions, the solution is
not to diss the (paying) customer, but to improve the software. It is
easy to miss important information in manuals. Write better manuals.
Hire real programmers.
In my view, writing Bad Software and then charging for support is
simply burdening the user with the cost of quality.
Sorry, I'm not attacking you, but this is a pet peeve of mine.
West Coast Engineering
expainitory.
If I buy a ~$1000.00 or > peice of software, I expect to have all of
my questions answered in a friendly and helpful way.
I got none of this when I purchased ZEMIN.
Branches of one aerospace company were told they were not welcome to
buy ZEMIN because the payment did not make it in 30 days.
So ZEMIN hangs their users out to dry because a buyer was on a coffee
break.
You need to be stupid to use ZEMIN. It is a piece of crap.
WCE
Mark W. Lund, PhD
West Coast Engineering wrote:
>
> How come we never have questions like this about CODE-V?
>
Maybe because if I bought Code-V and my boss found out that I was
asking silly questions on the internet about it he would doubt my
ability to perform and hire an expert instead?
Or maybe the Code-V users are way to proud to admit that they
can't use the dumb thing.
(The last time I used Code-V personally I had one of the punch
cards in the wrong order and I got a few pages of garbage out:-)
--
----------------------------------------------------------------------
Mark W. Lund, PhD ** Custom Battery Chargers
CEO ** Bulk Cells and Custom Battery Packs
PowerStream Technology ** Custom Power Supplies
140 S. Mountainway Drive ** DC/DC Converters
Orem Utah 84058 ** Custom UPS
http://www.PowerStream.com ** Custom power management electronics
Alternate e-mail address las...@netscape.net
Richard Cline
Engineering <westcoaste...@westcoastengineering.com> wrote:
Just to keep the record straight, I received excellent service from
Zemax. They were prompt in updating my hardware key. They were very
helpful in answering questions. It is quite possible that you managed
to insult them before asking any questions.
Dick
elliot burke
Are you saying that a $1000 piece of software should include enough time
from support personal to teach you not only how to run the software, but
explain the engineering involved?
Would this be true if learning the engineering involved several years at
grad school?
So for $1000 you'd expect a graduate education?
If not, then where would the responsibility of the software vendor end?
IMHO, the answer is simple, and I repeat, RTFM.
Not everything will be in the manual, but if you are too lazy to study the
program by reading the manual and working the examples, you'll never learn
how to use it.
The position of Zemax seems to be far more lenient than I would be. On many
occasions they have told me the answers to things that were in the manual
(and where). Eventually I got to the point where I was too embarrassed to
call them with a question before checking. They were without exception
polite and responsive to my questions.
Windows help files are much better than PDF files for help, because of the
indexing, context sensitivity and powerful search function. If Zemax help
was in the form of Windows help this would be a great thing.
Anybody else want to join with me in petitioning them for this change?
regards-
Elliot B.
"West Coast Engineering" <westcoaste...@westcoastengineering.com>
wrote in message news:fs6750pamcppnnski...@4ax.com...
West Coast Engineering
>
>
>West Coast Engineering wrote:
>>
>> How come we never have questions like this about CODE-V?
>>
>
>Maybe because if I bought Code-V and my boss found out that I was
>asking silly questions on the internet about it he would doubt my
>ability to perform and hire an expert instead?
>
>Or maybe the Code-V users are way to proud to admit that they
>can't use the dumb thing.
>
>(The last time I used Code-V personally I had one of the punch
>cards in the wrong order and I got a few pages of garbage out:-)
Today it is easier to use than ZEMAX. The GUI is at least as good if
not better and for the "old hands" we did not loose the command
interface.
When you call CODE-V support, they stay on the phone with you, every
time, until the question is resolved. The question can be as simple as
where to look up proceedure or as complex as how to perform a near
field propagation through a system.
WCE
West Coast Engineering
>In article <fs6750pamcppnnski...@4ax.com>, West Coast
>Engineering <westcoaste...@westcoastengineering.com> wrote:
>
>Just to keep the record straight, I received excellent service from
>Zemax. They were prompt in updating my hardware key. They were very
>helpful in answering questions. It is quite possible that you managed
>to insult them before asking any questions.
I waited until the ZEMIN author insulted me pretty good and hung me
out to dry on a proposal (there was a bug in the telecentric ray
aiming in ZEMIN which he refused to fix or supply a short term work
around for) before I threw the key and manual in the trash can and
completely switched to CODE-V.
I wrote my own code to bail out the proposal.
WCE
To put my distaste for insulting software user support into
prospective, I lead the revolt at Hughes Aircraft which lost Pat
Hennessey his control of HEXAGON because he was treating the designers
like shit. That little exercise probably sensitized me to the kind of
support I got from ZEMIN and preordained my reaction to it.
West Coast Engineering
At ORA, the user gets to ask the stupid question in the privacy of
their office on the phone. The user support staff does not tell the
user they think the question is either smart or stupid. The user
support staff makes sure the user has the answer needed so that the
user solves the problem at hand and looks good to his customers and
boss.
It does not get any better than that.
WCE
Helpful person
>
> How come we never have questions like this about CODE-V?
>
> Because they have real user support and it goes way beyond insulting
> the registered user and telling them to look in the manual .
>
> The reason you don't see questions like this concerning Enterprise is
> that I answer all questions myself, providing useful answers.
>
> There is no excuse why an optical design program user must ask for
> help from a use group rather than getting useful help from the vendor
> who sold them the software in the first place.
>
> The argument, "they are no longer under user support" is pure BS.
>
> If I sell something to another human being, I owe it to them to not
> leave them stranded. To do anything else, it is simply IMMORAL.
>
> Ken Moore owes his livelyhood to these people. The least he can do is
> answer their email.
>
> I hope I never see the ZEMAX business model as a good thing.
>
> WCE
I generally agree with you. However, I can understand a software
provider requiring a support contract, as over the life of a user,
support can be very expensive. In addition, I agree that it is not
only bad business but insulting when the support personnel talk down
to the user whom they assume are not educated in the methods of lens
design. (I was furious when the support person from a well known
company did this to me, when I am certain that I have more knowledge
and experience with various design programs and methods of
optimization than they.)
I do not maintain a support contract with that certain software
company, as I don't need it. However, I consider it unconscionable
when they are not interested in hearing about bugs. No matter how
trivial or important, the attitude is "buy support and we will
consider fixing the bug".
Laurent d'Mascarenhas
"West Coast Engineering" wrote
> At ORA, the users get to ask the stupid question in the privacy of
> their office on the phone.
Yeah, sure, you mean the privacy of their cubicle.
For their manager, colleagues or ennemies to overhear in their own adjacent
cubicles.
Assuming of course that phone calls are not already monitored or that their
management has not implemented a key logger in case they communicate with
ORA by e-mail..
> It does not get any better than that.
Indeed
> WCE
Now , seriously, are you a real, bona-fide Code V / ORA / user /
subscriber?
There is a slight difference in cost between ORA and FS * (or say Oslo) , as
I assume you know.
Not everyone can afford ORA, even they are truly superior in every respect,
a point of general agreement.
So you get what you pay for, or at least what you can afford.
* FS = Focus Soft, clever trick to avoid spelling out the 5 letter word.
Also, just to satisfy my curiosity, are you using your own software
Enterprise at work?
If no, why not?
Laurent d'Mascarenhas, Secrétariat
Athénée Somnifère de St Rémy-sur-Deule
31416 Somme-Inférieure
Jamie Carter
key of a stray version (to read files from outside sources), I tried
it and gave it back. It always invokes distaste to me and most the
other professional designers that tried it at LMC.
I am, however a serious, bona fide ('Oh Brother, Where Art Thou'
reference?) Code V user with my own serious, bona fide license for
personal use (yes, I am paying for it myself). I am also a serious,
bona fide Enterprise user with a license for its use. I have developed
designs with Enterprise using special "user requested" operands that
would have been much more difficult in Code V. That's one of the most
important and seriously overlooked aspects of licensing optical design
software. If the developer/publisher is not responsive to the
designer, then their software isn't worth the effort to load not, to
mention to train on. WCE excels at this and deserves respect for
service and commitment to the art of optical design.
You are correct when you state that you get what you pay for. Anyone
buying serious, bona fide, complex optical systems will save a mere
pittance on design software and blow a whole bundle on designer costs
(additional time required to complete a potential inferior design) and
production costs (making the inferior design) if they don't use the
best tool for job. The software costs isn't even in the noise for
serious, bona fide, optical designers at serious, bona fide,
technology firms.
Unfortunately, most real designers are managed by technical types that
have such weak skills that they were promoted to allow someone else to
be hired that could to the technical job. Then these guys impress
their management on how they are saving money in the engineering
department which is always seen as a cost center to the MBAs. Rarely
do you find a company that recognizes that the money spent in the
design and manufacture engineering can save a bundle or make a
competitive edge in the market place.
The one-seat shops that buy the cheap solution, mostly represent
individuals that "do" lens design on the side or every now and then or
maybe use it to "comp" lenses in production. For them, these solutions
are adequate and they get what they pay for. But it's a fallacy to
mistake these users for the guys that develop optical solutions for
complex systems every day of the week. And these guys generally use
the best tools.
In fact, the real pro's don't settle for just one software tool. The
serious, bona fide, optical engineer will model the design in more
than one code and search for inconsistencies. This is the only shot
that one gets at design assurance before committing big $$$ to
implementation and test. Since no two codes have exactly the same
methods for representing and modeling the optics, this process will
identify problems in the design that are easy to overlook. This also
forces the designer to re-enter and re-evaluate all of the
assumptions, operating conditions and model parameters and thus forces
an audit of the design. I know that this simple step, although
requiring a couple of days to complete, has saved me from serious
mistakes and embarrassment on more than one occasion. It also saves
much more time than it has cost and I won't mention the financial
aspects. As a designer and engineer, I know that I must fight tunnel
vision regarding the design goals and I must take the time to ensure
accuracy in my own designs. I have yet to find another person who can
take this responsibility away from me with any real assurance - not
even a peer review can get to the level of detail that I am referring
to.
In the privacy of my cubicle, I have no sense of shame to ask my
colleagues (that might or might not be able to listen to my
conversations on the phone) if I have problem or mis-understanding of
the optical model or code. Anyone who does have "shame" may suffer
from an exaggerated ego. Even the Code V support folks will tell you
that no one person (even at ORA) is intimately familiar with all of
the things that Code V can do. I have used it for well over twenty
years and I know that I am not. I still call 1 800 2getORA about once
a month and still ask "stupid" questions. The old RTFM cliche has no
meaning here. None of the design codes has a serious, bona fide
reference that explains how to do complex tasks. These manuals list
all of the commands or options or dialog boxes from a program function
perspective. They all seriusly lack a comprehensive guide for "how to"
approach a design goal (other than the Cooke triplet design tutorial).
If I had a designer working on my projects and he didn't call when he
stuck in a problem, I would fire him much faster than the guy that
calls for help when he needs it and works past the issue.
The help of another designer gives the most important guidance
available. Both ORA and WCE feature this type of support. And that
completes the circle regarding why user support is the single most
important aspect of the optical design software that you buy.
James A Carter III
"Laurent d'Mascarenhas" <Athénée.Somnifère@Saint.Rémy-sur-Deule.Net> wrote in message news:<105crj4...@corp.supernews.com>...
West Coast Engineering
wrote:
>
>"West Coast Engineering" wrote
>
>> At ORA, the users get to ask the stupid question in the privacy of
>> their office on the phone.
>
>Yeah, sure, you mean the privacy of their cubicle.
>For their manager, colleagues or ennemies to overhear in their own adjacent
>cubicles.
>Assuming of course that phone calls are not already monitored or that their
>management has not implemented a key logger in case they communicate with
>ORA by e-mail..
>
Hi,
You either need a new company, a new country or a new therapist. Maybe
all three.
WCE
Detector195
> The one-seat shops that buy the cheap solution, mostly represent
> individuals that "do" lens design on the side or every now and then or
> maybe use it to "comp" lenses in production. For them, these solutions
> are adequate and they get what they pay for. But it's a fallacy to
> mistake these users for the guys that develop optical solutions for
> complex systems every day of the week. And these guys generally use
> the best tools.
No disagreement here. And I doubt that any mistake is being made. If
you are doing lens design at this level, then you know what tool is
best for you. But you still can't presume that you know what is the
best tool for everybody.
An analogy can be made with software development. Most programmers
have come to grips with the fact that we will not all kowtow to C++.
There are often more practical choices for smaller shops, scientists,
and entrepreneurs. I guard my copy of Visual Basic 5 with my life. And
if this does not infuriate the pro's enough, I also use ExpressPCB for
electronic design and Excel for numerical computation. Adding
Enterprise to this collection of tools, I am fully equipped to do
small scale optoelectronic instrumentation development with virtually
zero overhead.
> In fact, the real pro's don't settle for just one software tool.
I can understand where this is important for a contract designer or
aerospace shop. My design risks are simply much lower, and in fact,
the greater risks generally involve system integration issues or
simply misjudging the needs of the customer.
We are not all doing complex multi-element lens design. Optics is a
big tent. There is a place in the world for multiple tools covering a
range of price levels. I am not assuming that the software I inherited
6+ years ago is the best tool, which is why I am keenly interested in
this discussion.
I have definitely begun to appreciate the strengths of Enterprise.
Also, I noticed that among the "slick GUI" programs, Optalix seems
very well thought out. Has anybody else had experience with it?
Laurent d'Mascarenhas
Thank you for a nice informative answer.
Ld'M in person.
Laurent d'Mascarenhas
"West Coast Engineering" wrote
>
> You either need a new company,
too late, retired.
> a new country
tried several, none 100% satisfactory. Too late to try yet another one.
> or a new therapist.
A confirmed case of the pot calling the kettle black (or vice versa,
whatever).
Besides you must have been living under a rock.
An office? in this age of cubicles? LOL as my daughter would write!
Keeping anything secret, confidential or private... LOL again.
Ever heard of total information awareness?
It's not because you are paranoid that they are not out there to get you.
Now lie down on this couch, Mr. WCE and tell me more about this optical
software that you hate so much.
I mean how long have you had this obsession? A looong time, I see.
OK, well listen if you are taking this medication regularly it will most
certainly disappear.
I do hope so because you *can* talk about optical design matters
knowledgeably and intelligently.
How about our next appointment? Same time next week, OK. And please, please,
start your medication immediately.
Ld'M
> WCE
Jamie Carter
> jaca...@onebox.com (Jamie Carter) wrote in message news:<4b78e1a9.04031...@posting.google.com>...
>
> > The one-seat shops that buy the cheap solution, mostly represent
> > individuals that "do" lens design on the side or every now and then or
> > maybe use it to "comp" lenses in production. For them, these solutions
> > are adequate and they get what they pay for. But it's a fallacy to
> > mistake these users for the guys that develop optical solutions for
> > complex systems every day of the week. And these guys generally use
> > the best tools.
>
> No disagreement here. And I doubt that any mistake is being made. If
> you are doing lens design at this level, then you know what tool is
> best for you. But you still can't presume that you know what is the
> best tool for everybody.
Detector195,
By your own admission, you use several different types of applications
to develop "small scale optoelectronic instrumentation." If you don't
fall into the category described in my missive given above, then I
must have trouble understanding what your do. Certainly, you do not
spend even a majority of yout time doing optical design in a
geometrical optics code. In fact, occasional use seems to be a good
description of your needs. I believe I was quite explicit in
acknowledging that your choice is quite reasonable.
Maybe you could explain where I have "presumed" that I know what is
the best tool for everyone? In fact, maybe you could identify anywhere
that I stated what this "best" tool might be?
>
> An analogy can be made with software development. Most programmers
> have come to grips with the fact that we will not all kowtow to C++.
> There are often more practical choices for smaller shops, scientists,
> and entrepreneurs. I guard my copy of Visual Basic 5 with my life. And
> if this does not infuriate the pro's enough, I also use ExpressPCB for
> electronic design and Excel for numerical computation. Adding
> Enterprise to this collection of tools, I am fully equipped to do
> small scale optoelectronic instrumentation development with virtually
> zero overhead.
>
> > In fact, the real pro's don't settle for just one software tool.
>
> I can understand where this is important for a contract designer or
> aerospace shop. My design risks are simply much lower, and in fact,
> the greater risks generally involve system integration issues or
> simply misjudging the needs of the customer.
>
> We are not all doing complex multi-element lens design. Optics is a
> big tent. There is a place in the world for multiple tools covering a
> range of price levels. I am not assuming that the software I inherited
> 6+ years ago is the best tool, which is why I am keenly interested in
> this discussion.
>
> I have definitely begun to appreciate the strengths of Enterprise.
> Also, I noticed that among the "slick GUI" programs, Optalix seems
> very well thought out. Has anybody else had experience with it?
I am very happy for you that your design risks are very low. If you
are not doing multi-element design, why not use Excel? It's certainly
capable of analyzing and even optimizing 1 or 2 element lenses. In
fact, I use it for first order layouts and predesign (3rd order
solutions) before tackling the individual lens groups in a distributed
path.
BTW, I _was_ addressing persons engaged in "doing complex
multi-element lens design" in an environment that gives significant
design risks. VEry often, these people represent the "real pro's" in
optical design. These are also the people from whom you will learn
about your problems and needs. Obviously, you are aware of this as you
are "keenly interested in this discussion." If you can get over the
need to criticize their opinions, chances are that you might learn
even more.
Kind regards,
Jamie
Steve Eckhardt
Secrétar...@Institut.Auvergnat.de.Métaphysique.Org says...
>A confirmed case of the pot calling the kettle black (or vice versa,
>whatever).
>
>Besides you must have been living under a rock.
>An office? in this age of cubicles? LOL as my daughter would write!
>Keeping anything secret, confidential or private... LOL again.
>Ever heard of total information awareness?
>
>It's not because you are paranoid that they are not out there to get you.
>
>Now lie down on this couch, Mr. WCE and tell me more about this optical
>software that you hate so much.
>I mean how long have you had this obsession? A looong time, I see.
>OK, well listen if you are taking this medication regularly it will most
>certainly disappear.
>
>I do hope so because you *can* talk about optical design matters
>knowledgeably and intelligently.
>
>How about our next appointment? Same time next week, OK. And please, please,
>start your medication immediately.
>
> Ld'M
Best laugh I've had from this newsgroup in a long time! Thanks.
--
Best regards,
Steve Eckhardt
skeckhardt at mmm dot com
Detector195
> Maybe you could explain where I have "presumed" that I know what is
> the best tool for everyone? In fact, maybe you could identify anywhere
> that I stated what this "best" tool might be?
If I mistook your comments, sorry. This is not a formal debate, and we
all do our best to draw reasonable ineferences, but 100% accuracy is
not guaranteed. Besides, I think we actually agree with one another.
> I am very happy for you that your design risks are very low. If you
> are not doing multi-element design, why not use Excel? It's certainly
> capable of analyzing and even optimizing 1 or 2 element lenses. In
> fact, I use it for first order layouts and predesign (3rd order
> solutions) before tackling the individual lens groups in a distributed
> path.
Point well taken. I use Excel for predesign too -- zeroth order
solutions in my case ;-)
My designs are just interesting enough for a design program to be
useful. Seeing is believing, so a quick look at the layout drawing
with some traced rays catches a lot of my numerous conceptual
mistakes. Also, these drawings help me explain my designs to my
colleagues. The stock lens libraries are handy, as is optimization. It
is useful to fully optimize a design, just to make sure I am not
giving up too much performance by using stock elements.
Cheers,
Detector195
Huygens
> Point well taken. I use Excel for predesign too -- zeroth order
> solutions in my case ;-)
>
> My designs are just interesting enough for a design program to be
> useful. Seeing is believing, so a quick look at the layout drawing
> with some traced rays catches a lot of my numerous conceptual
> mistakes. Also, these drawings help me explain my designs to my
> colleagues. The stock lens libraries are handy, as is optimization. It
> is useful to fully optimize a design, just to make sure I am not
> giving up too much performance by using stock elements.
As a small side note to your zeroth order comment, you might like to
look at the PreDesigner freeware - written for that purpose
- purely thin lens
- choice of key conjugate, aperture and field variables
- drawing & printout
- sliders for altering key parameters etc.
You can download it from www.winlens.de
Jan