Home
SearchSearch MemberlistMemberlist RegisterRegister ProfileProfile Log in to check your private messagesLog in to check your private messages Log inLog in

How to design a lens... CLP#100
View previous topic :: View next topic  


PostPosted: Tue Jan 07, 2020 11:36 am    Post subject: How to design a lens... CLP#100 Reply with quote

If anyone is interested in how lenses are designed as well as some history of lens evolution then this show may be of interest. It's the most "technical" show we've ever done (by some margin) and so it's aimed fairly and squarely at those people that people that want to know how a lens really works, which I suspect may apply to a few people here in MFLenses...

#100 Jason Lane To Zemax!
Our second anniversary and our 100th show. To mark the occasion professional lens designer, Jason Lane has returned and this time you get to see the software that Karl, Johnny and Simon were raving about in Karl’s final show, Episode #64. So here’s something of a first for us, a podcast designed to be watched on Youtube, rather than just listened to as an audio file. Having moved away from highly technical shows, here’s one for those people that want to get down and dirty at the lens element scale as Jason gives an insight into the world of lens design.

https://youtu.be/3yfFcP6k58M



PostPosted: Tue Jan 07, 2020 6:02 pm    Post subject: Reply with quote

Like 1


PostPosted: Tue Jan 07, 2020 10:41 pm    Post subject: Reply with quote

A fascinating session that I couldn't imagine working without the graphics.
Thanks for all the work.
Friends


PostPosted: Wed Jan 08, 2020 5:58 am    Post subject: Reply with quote

Like 1 Like 1 Like 1 Thank you!


PostPosted: Wed Jan 08, 2020 6:04 pm    Post subject: Reply with quote

Excellent stuff. A fitting podcast for your 100 not out Like 1 .


PostPosted: Thu Jan 09, 2020 10:20 am    Post subject: Reply with quote

Thank you all, glad you enjoyed it!


PostPosted: Thu Jan 09, 2020 10:35 am    Post subject: Reply with quote

itsfozzy wrote:
Thank you all, glad you enjoyed it!


Simon, if you get the chance when Jason comes back for part II, could you ask him what has been holding the periscope lens design back. They never got out of pocket point and shoot cameras in the 2000s, I think, due to image quality limits, but I'm not sure. They've now started to go into high end smartphones, providing optical zooming - perhaps the last barrier to being taken seriously as hobbyist cameras. Perhaps. Will smartphone manufacturers hit the same problem (if there is one), or have they overcame whatever was limiting the designs use?

Only if you get the chance.

Cheers man.


PostPosted: Thu Jan 09, 2020 10:42 am    Post subject: Reply with quote

Sciolist wrote:
itsfozzy wrote:
Thank you all, glad you enjoyed it!


Simon, if you get the chance when Jason comes back for part II, could you ask him what has been holding the periscope lens design back. They never got out of pocket point and shoot cameras in the 2000s, I think, due to image quality limits, but I'm not sure. They've now started to go into high end smartphones, providing optical zooming - perhaps the last barrier to being taken seriously as hobbyist cameras. Perhaps.

Only if you get the chance.

Cheers man.


I've heard of them regarding new phone cameras, but it's new to me with regard to conventional cameras. I've had a quick look for info, but I've not found anything useful. Do you have any links to illuminate your question as this will help me raise the question? (assuming I remember).


PostPosted: Thu Jan 09, 2020 11:02 am    Post subject: Reply with quote

itsfozzy wrote:

I've heard of them regarding new phone cameras, but it's new to me with regard to conventional cameras. I've had a quick look for info, but I've not found anything useful. Do you have any links to illuminate your question as this will help me raise the question? (assuming I remember).


I suppose they may be most well know in the very chic, Sony Cyber-shot T series of cameras that ran from 2004 -'12. They were in other cameras of a similar style at the same time, Fuji comes to mind. But they were around before that. I have a memory of a Minolta Dimage having one I think. I could be wrong on that one. I've read that the Periscope lens was the last major development in lens design, so, if that's actually true, it would perhaps be interesting if Jason is still able to tie it back to the designs from the late 19th century as he did with the Double-Gauss.

I'm not sure how useful a list of cameras will be, but... scroll down to the T series for model numbers, then you can find many reviews of them on Imaging-resource, dpreview and such. Like you, I've not found anything on the technical side.

https://en.wikipedia.org/wiki/List_of_Sony_Cyber-shot_cameras

Regards.


PostPosted: Thu Jan 09, 2020 1:20 pm    Post subject: Reply with quote

somewhat loquacious presentation, but very interesting points towards the end about the old Kodak box cameras as well as on intentional vignetting. presumably digital in-camera-correction makes it even easier to skimp on optics design.

Looking forward to more.

p.


PostPosted: Thu Jan 09, 2020 2:17 pm    Post subject: Reply with quote

itsfozzy wrote:
Sciolist wrote:
itsfozzy wrote:
Thank you all, glad you enjoyed it!


Simon, if you get the chance when Jason comes back for part II, could you ask him what has been holding the periscope lens design back. They never got out of pocket point and shoot cameras in the 2000s, I think, due to image quality limits, but I'm not sure. They've now started to go into high end smartphones, providing optical zooming - perhaps the last barrier to being taken seriously as hobbyist cameras. Perhaps.

Only if you get the chance.

Cheers man.


I've heard of them regarding new phone cameras, but it's new to me with regard to conventional cameras. I've had a quick look for info, but I've not found anything useful. Do you have any links to illuminate your question as this will help me raise the question? (assuming I remember).


The technical term is 'folded optics'. The Minolta Dimage X is an early example, here they show the lens diagram: https://www.dpreview.com/reviews/minoltadimagex
While the image quality is not great, there is nothing there to indicate that it is an inherent problem of the folded optical path, however we don't know if the prism is phase-coated or not. That could or could not be part of the problem. Without a phase-correction coating, there are light interference artifacts and a loss of contrast, but I'm not sure when these were invented. Actually without being an expert, I can't even say for sure that such a simple 45-degree prism requires phase coating, as porro prisms are unaffected. On the other hand, I'm sure there is a good reason as to why a simple mirror would be inadequate.

I don't think there is any major technical limitation on image quality, or a reason for why we see them becoming more prevalent in smartphones again today for any other reason than there is a market demand for it. See here, where they briefly discuss the merits and downside of an experimental 5mm thick series of folded optics, vs a conventional compact camera with 38mm lens: https://www.vision-systems.com/factory/robotics/article/16738922/folded-optics-makes-cameras-smaller

"In the laboratory, the engineers compared a 5-mm-thick, eight-fold imager optimized to focus on objects 2.5 m away with a conventional high-resolution, compact camera lens with 38-mm focal length. “At best focus,” says Tremblay, “the resolution, color, and image quality are similar between the two.” However, one drawback with the folded camera was its limited depth of focus.

“With a numerical aperture of 0.7 and approximately 90% obscured aperture, the depth of focus and depth of field become extremely narrow,” says Tremblay. “The depth of focus for the design was approximately 10 µm, corresponding to a depth of field of only 24 mm at a 2.5-m conjugate point,” he says."

When you think about it, folded optics have been used since nearly forever, as nearly every pair of porro prism binoculars is an example of the method. Until recently developed phase coatings, these have always had better image quality than binoculars utilising an in-line optical path via roof prisms.
Another example is well... periscopes!

Here is an even crazier concept currently being researched in the U.S: https://scitechdaily.com/time-folded-optics-create-new-possibilities-for-imaging/
A similar/related concept is 4d imaging, which uses a series of video, time and stills to recreate frames computationally. These, I believe, all use folded optical arrangements.

Folded optics arrangements on phones gets a brief mention on Dxo: https://www.dxomark.com/multi-camera-smartphones-benefits-and-challenges/


PostPosted: Thu Jan 09, 2020 8:12 pm    Post subject: Reply with quote

Great resource. Thanks for sharing.


PostPosted: Thu Jan 09, 2020 8:31 pm    Post subject: Reply with quote

Teemō wrote:
… The technical term is 'folded optics'. …



Great post Teemo Like 1 .


PostPosted: Thu Jan 09, 2020 8:45 pm    Post subject: Reply with quote

The folded optics for gaining more zoom room in smartphone package crossed my mind too. I left it at that plus anamorphic could equal pretty nice results.

Thanks Teemō


PostPosted: Fri Jan 10, 2020 7:55 am    Post subject: Reply with quote

or simply an ordinary camera with interchangeable lenses and built-in telephone, but batteries that die within a day may not be so popular.

p.


PostPosted: Fri Jan 10, 2020 8:51 am    Post subject: Reply with quote

Like 1 Like 1


PostPosted: Sat Jan 11, 2020 11:57 am    Post subject: Reply with quote

Teemō wrote:
...

Here is an even crazier concept currently being researched in the U.S: https://scitechdaily.com/time-folded-optics-create-new-possibilities-for-imaging/


I just read the article on 'Time-Folded Optics'. Man, that was really interesting.


PostPosted: Sat Jan 11, 2020 12:50 pm    Post subject: Reply with quote

Is there an open source alternative to Zemax OpticStudio that offers GUI and Raytrace?


PostPosted: Sat Jan 11, 2020 7:46 pm    Post subject: Reply with quote

Enjoyed the whole episode, thanks for posting.


PostPosted: Tue Jan 14, 2020 8:08 am    Post subject: Reply with quote

Blazer0ne wrote:
Is there an open source alternative to Zemax OpticStudio that offers GUI and Raytrace?


I think WinLens Basic would be the best option:
http://www.qioptiq-shop.com/Optik-Software/Winlens-Optical-Design-Software/Free-Winlens-Basic/Winlens-Basic-Features/

But in the free version is no optimization as far as I know.
I often thought about installing it, but have not done it yet. At work I use Zemax, which is quite nice, with a lot of possibilities.


PostPosted: Tue Jan 14, 2020 1:24 pm    Post subject: Reply with quote

ZoneV wrote:
Blazer0ne wrote:
Is there an open source alternative to Zemax OpticStudio that offers GUI and Raytrace?


I think WinLens Basic would be the best option:



Thank you, fellow lens nerd!

This looks like a nice app to get the feet wet.


PostPosted: Tue Jan 14, 2020 1:32 pm    Post subject: Reply with quote

Blazer0ne wrote:


Thank you, fellow lens nerd!

This looks like a nice app to get the feet wet.


Yo are welcome!
Here is other food for software thoughts, but mixed with commercial, and some old ones.
http://www.optenso.com/links/links.html

Good thing about Qioptiq Winlens suite: If you like it much, you can buy for ~400 Euro the better version with optimization and more.
But such software is not good for all optical simulations, even Zemax has some drawbacks.


PostPosted: Thu Jan 16, 2020 4:56 pm    Post subject: Reply with quote

Blazer0ne wrote:
Is there an open source alternative to Zemax OpticStudio that offers GUI and Raytrace?


I have used Optical Ray Tracer for a few years. I believe development has stopped and the interface/interaction is a bit clunky but the ray tracing is accurate and you can measure focal length and spot-size manually by zooming in/out and comparing coordinates. Fun to play with!

https://arachnoid.com/OpticalRayTracer/

There is a basic tutorial in the documentation to familiarise you with an achromatic doublet: https://arachnoid.com/OpticalRayTracer/helpresources/HelpText.html

The help GUI is available in the program too. With Excel or another program you can make some spreadsheets and do additional calculations and then test your idea in the ray tracer. Idea

I haven't tried other legacy design programs. This one has a simple GUI and is a great introduction, anything more and you'll find yourself needing a textbook.


PostPosted: Thu Jan 16, 2020 11:22 pm    Post subject: Reply with quote

Teemō wrote:
Blazer0ne wrote:
Is there an open source alternative to Zemax OpticStudio that offers GUI and Raytrace?


I have used Optical Ray Tracer for a few years. I believe development has stopped and the interface/interaction is a bit clunky but the ray tracing is accurate and you can measure focal length and spot-size manually by zooming in/out and comparing coordinates. Fun to play with!

https://arachnoid.com/OpticalRayTracer/

There is a basic tutorial in the documentation to familiarise you with an achromatic doublet: https://arachnoid.com/OpticalRayTracer/helpresources/HelpText.html

The help GUI is available in the program too. With Excel or another program you can make some spreadsheets and do additional calculations and then test your idea in the ray tracer. Idea

I haven't tried other legacy design programs. This one has a simple GUI and is a great introduction, anything more and you'll find yourself needing a textbook.


Thanks, Java works for me! At first glance, the documentation appears to be a great resource for understanding the calculations required. There is even a link to the eclipse project.


PostPosted: Sun Jan 19, 2020 10:17 am    Post subject: Reply with quote

Open source, as all good things should be. Smile