Episode 755 A camera lens has many lens elements (pieces of glass). Why? There are many reasons. I try to give some insight by explaining one, field curvature. Be a Patron: / imsaiguy
Well said. Even my high school physics text book tell me one single convex lens can make a perfect image everywhere, like if your object is 2f from lens then you will get a perfect image on 2f.
Coma is Vietnamese for vampire. When I was in Saigon in 1972, I went to a movie called Coma. Vampire. Except this vampire was a woman who wore white and played piano in the forest on her time off.
There seems to be no limit to the number of technology fields that you are great at explaining/teaching. I would have loved to have you as a teacher/professor IMSAIGuy. Thank you for making these videos for us fine sir! Fred
Thank you for this. Just a historical note regarding the Cooke Triplet: it was designed and patented in 1893 (not the 1930's) by Dennis Taylor who was employed as chief engineer by T. Cooke & Sons of York, England. Its earlier provenance makes it the more remarkable.
Cooke Triplet is also basis for all zoom lenses: Moving center element in between outer elements changes its focal length. But also aberrations change and are harder to control and it took until 1930s to get first zoom lens into production.
Clicked on the video thinking that this is the start to my journey of making the most compact 1.4 primes to be ever made. I’m just gonna stop while I’m behind.
Yes, Lenses can be very complicated to design, on one of my many visits to the old Hasselblad Factory, I was told that the Zeiss Biogon 38mm 4.5 took a long time to design and calculate, the amount of A4 papers used for calculation was a stack of one meter high ! I had the Zeiss Mutar 2x teleconverter, after I told the folks at Hasselblad how good it was on the Zeiss Planar 110mm 2.0, they got very interested and wanted to learn more, later I was told that Hasselblad had bought a computer from Canada to calculate Lenses. Today all Lens makers use computer software to calculate Lenses, and Lenses today outperform all older Lenses.
Really interesting. I've been a photographer for decades, and I'm a little ashamed to say that however mind-boggling I always found lens techniques to be, I know way too little about the 'glass' that I use. Thanks and thumbs up.
Another trick they've been doing recently for mirrorless cameras is that since you've got to apply digital processing to the sensor image's anyway, some corrections can be done in software and the lens design can be optimized for other factors. If you squint this is a analogous to the curved film trick you mention in the video.
Very interesting! I'd also like to learn about aperture, for example, why a narrower aperture improves image quality, why resolution goes down when it improves image quality, and so on. Thank you very much!
Very cool. I've wondered about this, off and on, for decades and never got around to looking it up! One thing I *think* might be missing is that you can use materials with different indexes of refraction for each lens to tweak it even more.
yes, glass with different dispersion, to correct for color aberrations. classically referred to as crowns and flints. I talk about it a little: ruclips.net/video/PxqhA37bKtI/видео.html
Likewise. Interesting to note that the more glass elements you have in a lens, the worse the image quality. It goes something like this: more glass = less light (in terms of F stops.
At 06:50 Yes, the use a lot of elements for a single lens. More so if you consider that they use a very narrow band UV light source. Chromatic aberration is therefore limited, EDIT: Good to see that you mention it also.
"The Cooke triplet is a photographic lens designed and patented in 1893 by Dennis Taylor who was employed as chief engineer by T. Cooke & Sons of York".
I could teach you how to do lens design and loose all my viewers. They only care about RF. In silicon valley I can count the lens designers on one hand. My buddy in San Diego designed one of the mars rover lenses. He is quite proud.
That's really interesting how you explain how lenses is working, so most people can understand, or at least have an idea about it. I subscribe to your channel because you have a lot of good interesting content. Greetings from Kenneth (Denmark)
this explains pretty well tho why it took a (comparatively) long time for compact cameras to gain lenses that went wider than 35mm (equiv.) guess it's just that - harder and more expensive to make light bend around alot more at wider angles
I watched a video a few years ago about how lenses are made. They mentioned that someone was experimenting with using a curved sensor and that that was the future. He said that they would have to redesign the lenses to work with it and you'd only be able to use lenses designed for it. But as you know, companies like to get existing customers into newer models. If they're willing to sell all of their gear to move to something totally new they might just leave and go somewhere else. You have to make it easy to adapt. Look at mirrorless for example. You can adapt your old mount lenses to the new mount and still move to mirrorless. And mirrorless cameras have made designing lenses better because the focusing point is closer to the sensor. Or at least I think it would be easier. I shoot Nikon so I mostly know about Nikon lenses. I know that the z lenses are sharper corner to corner than the old f mount lenses. There's little to no chromatic aberration or focus breathing with the new lenses vs the old lenses. There's less vignette. I think mirrorless cameras made curved sensors non essential.
This is so cool! Thanks for sharing. I believe the optical coating is also important, isn’t it to ensure the light entering the lens is within its comfortable zone (generally using thin film reflection to bounce back the other ones)
I have an Astro-Physics apochromatic triplet, with the 3 lens elements in contact with each other and oil spaced to reduce internal reflections, and improve transmission. I love this thing! For as much as it costs, the field flattener accessory comes with quite a breathtaking price tag; but it was intended for medium format film photography at the focal plane. Fortunately, at the time I acquired it, only APS-C sized CCD devices were practical (affordable!) and the part of the focal plane I was using was "flat enough". Later I move to using a Riccardi-Honders astrograph, an f/3.8 optical system and quite a novel design. You might mistake it for a Schmidt-Cassagrain but it's actually more like a folded refractor with a rear silvered lens, and more importantly, I believe all spherical lens figures. And a nice, large and wide flat focal plane! Only suitable for astrophotography, but that's the problem I had. Never had an eyepiece in it. I asked the designer of that telescope why he didn't use carbon fiber or some other material with low coefficient of expansion due to temperature for the tube assembly, rather than aluminum? That was a deliberate choice, the dimensional change of the tube due to temperature offset the changes in the optical elements as they also cooled down. There is quite a bit of "art" to go along with the science. Difference between theory and practice and all that stuff you learn the hard way. Optics are cool! Really nice optics are very cool and expensive! And definitely one of those fields where there's no single "best" tool/design. It's always a tradeoff over a bunch of factors, including $$$ it seems.
Now consider the trade offs for something like euv. Where they reduced the stack to just 7 reflexion lenses because it is so hard to refract the light without loss
Thanks for the information. I'll pursue Omnivision. Optically, the setup is simple : a 20D lens is held in front of the eye being examined while a cell phone camera is positioned 30 mm or so away from the 20D lens to capture the aeriali image produced by the 20 D.. The cell phone illumination and auto focus do the job. Some people capture the image as video and look for the best frames. Thanks for your suggestions.
Nice, I just discovered this. Good video. Off the top of my head, two other famous cameras that use a curved film plane are Agfa Clack and Agfa Click. Of course these were just really cheap, essentially toy cameras with single element lenses (or at best an achromat) but pictures are sharp thanks to the curved plane. I still use my Click from time to time.
I find your videos helpful, because I'm trying to design an eyepiece that magnifies the image of a camera viewfinder. I want to mount the eyepiece ontop of the viewfinder. A simple telescope doesn't do the job, because the aperture is too small and it cuts off some parts of the virtual image.
So interesting ! I am searching for the specs of a Canon FD 85mm 1.8 lens in order to manufacture the front element. Do you have any clue where to find it ? I would be so grateful
You will not find any specs. They are secret. Most large camera companies have front elements for sale. especially for expensive pro lenses. Having a single element made would cost more than the lens is worth.
Thanks for the clear explanation of why we need so many layers of lens in cameras. I have a stupid question, I am wondering why our eyes can do it so good with only one lens? Is the lens in our eyes any special?
They look similar but are quite different. The Schmidt plate is at the aperture stop of the system. There is a Fourier transform from the aperture to the image. the image is in x-y space and the aperture is in angle space. The incoming wavefront will be phase corrected by the plate to correct for spherical abberation and I think astigmatism and coma. it is difficult to grind an aspheric mirror so the job is given to a thin piece of glass that can be warped, ground flat, and unwarped to result in the shape.
imsaiguy: Thank you for this informative video. I just have only a cursory interest in lens design. I have not heard "IMSAI" for many years, this was the beginnings of the micro-processing revolution started by Intel, Zilog etc. started in the mid-1970s.
I found this very informative and interesting, thank you. I recently did a BINDT CAT 1 IR thermography course, cat 1 being the starter level! I intended to dive deeper but are the lens makups very similar to photography lenses or is the thermal camera purely dependent on the sensor technology to achieve a desirable image? Guess I may need to do the Cat 2 course!!
Looks like curved camera image sensors are going to be a thing in the next decade or so from the recent porotypes that were shown off. Wonder how that will change lens design.
@@IMSAIGuy That's right! It reminded me of the microscopy course I took at uni, where we covered some of the abarations and limitations you have to deal with when working on optics. You might have heard this already, but I recently came across some information that sony might be working on creating curved digital imaging sensors! People speculate that could lead to some impressively small optical designs for future cameras (maybe in phones too)
I think I found a new wormhole to dive into. Any advice on how to go about doing more studying of this type of stuff? Are there any companies that'd make the lens elements to order?
I would get the book Modern Optical Engineering by Warren Smith if you are interested in lens design. Yes there are prototype houses for lens elements. It is way too expensive for DIY stuff. you can get some generic stuff at places like Edmund Optics
11:02 digital: It's not impossible It's just very cost prohibitive with the current manufacturing methods. Like swapping carbon nanotubes for the fiber plain. You'd have to grow the nanotubes on satellites.
well explained! could SLR or DSLR lens designers use the cell phone field flattener lens type to make a minimalist sise lens for full frame cameras? too hard to grind complex shape in bigger lenses?
Now they can use software to correct images and use much simpler glass arrangements. Saves weight and complexity. It might work in practice, but does it work in theory?
I would be interested in more videos like thsi, i have been trying to research building a triplet telescope for a while but i havnt found any triplet sets so the idea is to find out how to select individual elements, problem is nobody actually explains it well at all. Are there any open source tools for this? Or any actually useful resources?
I would get the book Modern Optical Engineering by Warren Smith if you are interested in lens design. CAD tools I've used: ZEMAX, OSLO, Code V, LightTools. once in a while MATLAB or Octave for specialized calculations and color theory. OLSO may have a trial version, you can also try WinLens3D
Hey there. It is my first time on your channel and I am quite interested in learning more. Namely, I am interested in the detailed process of how you reverse engineered this lens, what software do you use, and the like. I am currently trying to design simple LED lenses, which do not require image fidelity and only usually attempt to narrow the light beam from a beam of 120 degrees to a one of 30 degrees or so. I was experimenting with the aspheric lens equation and got somewhat stuck. I was using open source software and mathematical solvers like sage. If you find the time to talk more about lens design and, for example, what equations, curvatures, and different types of glass there is, I would love to watch this content. The field leveller glass, for example, trades width for depth to achieve the outcome but how does it do that and do the surface follow an equation. Are lens surfaces all mathematical and how are they obtained (I mean, how are the formulas obtained). At a minimum, I would like to know what computational tools and/or software you use. Thanks!
I used ZEMAX, OSLO, Code V, LightTools. once in a while MATLAB or Octave for specialized calculations and color theory. OLSO may have a trial version, you can also try WinLens3D
this video was really great I have watched all your lens-related videos and really like them can you make one about UV lens protectors what effect do they have? thanks also, why do reversed lenses work so well for macro?
Great video! Has anyone explained how the focal lenght of a camera lens works? Like, where do you "fit" say 800mm of distance in a 800mm lens? Because most often, you cant, most lenses aren't even as long as their focal lenght. (Unless ofc there is some glass trickery involved)
Excellent video, I would love to learn more about how to go about designing lenses. I am currently working on a project that uses Kodak disposable camera lenses and adapts them to other cameras.
You could try a program could OSLO. It's an old version of a program like Zemax. It lets you model up to 10 surfaces for free. So if you're doing it unprofessionally that would be a good starting point. I use Zemax and it's great, but the license is expensive.
@@andrewdavies2358 my boss used OSLO and I used Zemax. we would both run optimizations to see if either has missed anything. Didn't know about the free 10 surfaces. that barely lets you design a triplet, but certainly would be useful for the astronomy crowd.
This is cool. I've been writing a renderer (for fun) and want to add realistic camera models. This kind of reverse engineering would be an awesome starting point. And you could try different parameters in the model, and see what the effect is. I'm not modelling wave effects so much, and no quantum effects, but it would be fun to play with.
the big boy lens design CAD packages allow you to render an image through a design. 2D only but it gives you some insight. 3D is possible for non-sequential ray tracing but will not have many of the aberrations included.
@@IMSAIGuy the fun part for me is the learning and design/coding. :-) My code isn't intended for lens design. It's intended for rendering scenes (vfx). Being able to include a realistic lense helps match a simulated scene to a real one, and also adds aristic range
Where do I start if I want to measure a lens and figure out its parameters? I have a huge aerial photography lens, the first element is nearly 6 inches in diameter, and aside from focal length and aperture, I'm a bit in the dark. The front and rear halves unscrew from the shutter (no adjustable aperture, just open-close). Taking a giant lens spanner to the front, it's just two elements. The rear has 3 elemens, one single and two glued together. Since it's fixed focus as well, I'm pondering how to make it adjustable focus - the easy answer might be to mount the lens on a rack / bellows type movement, but I've been testing it out on the kitchen counter using a sheet of paper to gauge the focus and a ruler to see how much movement is required, and for the range of distances I'd like to use it at, I need over a meter of travel. The bellows would be bigger than the camera and require some support to keep out of the optical path. Just for fun, I calculated the hyperfocal distance and its around a mile.
that is a very complicated process. you need to measure the curvatures of each surface and the thicknesses and spaces, then you will need to know what glass type is used for each element (index and dispersion). these will all require special instruments. once you have these number they will not be accurate enough. you will need to put them in an optics design program to adjust them.
There are also physical limitations. DLRSs have the mirror s the lens is mounted typically at 45 mm from the sensor and the lens cannot go much deeper. Yet some lenses have focal lenghts as small as 10 mm. On telephoto the problem is the opposite. You want to keep the lens short, shorter than its focal length. Zoom lenses naturally add their complications.
Which the lens optical design software you used to reverse engineer the design.can u do a video on lens rehousing.there is not much material available about rehousing.or suggest books or online material you know about it.This was one great video.thx
lens cad program: Zemax mechanical books: Opto-Mechanical Systems Design by Paul R. Yoder, Jr. Fundamentals of Optomechanics (Optical Sciences and Applications of Light) by Daniel Vukobratovich
I would get the book Modern Optical Engineering by Warren Smith if you are interested in lens design. CAD tools I've used: ZEMAX, OSLO, Code V, LightTools. once in a while MATLAB or Octave for specialized calculations and color theory. OLSO may have a trial version, you can also try WinLens3D
Interesting! I know virtually nothing about optics besides what I vaguely remember from undergrad physics class. I suppose nowadays one can use computers to optimize lens design in ways that weren't possible in pre-computer days?? Must be even harder to design good zoom lenses then?
Before computers they used ladies with adding machines. The job title was 'computer'. Each lady would be given one ray to trace. Computers make it much easier to design (and poorly design). I've seen lots of 'kids' that get an optics cad program and think they can design lenses. It doesn't go well. Yes, zoom design is a real art.
WA lenses has more complicated design mainly due to existenece of a mirror in SLRs which mimposes higher distance from lens to film/matrix. Extremly wide lens for large format or rangefinder camera is build of 6 to 8 lenses.
the SLR back focus requirements usually adds about or or two elements. of course the large format lens does not need as high of an MTF which helps. The Hasselblad HCD 24mm f/4.8 Lens still has 14 elements. and the Voigtlander Nokton 21mm rangefinder is 13 elements.
@@IMSAIGuy my experiences with LF lenses are that theirs sharpness is on par with SLR (not only medium format lenses), while design is much simpler (6 or 7 lens Gauss design - Schneider Super Angulon). 35mm lenses are more complicated, probably due to the vignette canceling formula (LF WA lenses use a central gray filter).
I'm curious: Is that a mathematically accurate statement to say that a perfect lens doesn't exist? If that's the case, is it also true if you replace surfaces with a smoothly changing refractive index gradient?
Наука
This is so much better than most videos about how lenses work. The other ones are frustrating because they just leave you with more questions.
Well said. Even my high school physics text book tell me one single convex lens can make a perfect image everywhere, like if your object is 2f from lens then you will get a perfect image on 2f.
I agree!
Coma is Vietnamese for vampire. When I was in Saigon in 1972, I went to a movie called Coma. Vampire. Except this vampire was a woman who wore white and played piano in the forest on her time off.
There seems to be no limit to the number of technology fields that you are great at explaining/teaching. I would have loved to have you as a teacher/professor IMSAIGuy. Thank you for making these videos for us fine sir! Fred
Thank you for this. Just a historical note regarding the Cooke Triplet: it was designed and patented in 1893 (not the 1930's) by Dennis Taylor who was employed as chief engineer by T. Cooke & Sons of York, England. Its earlier provenance makes it the more remarkable.
Interestingly the Cooke Triplet was f/3.5. Still modern entry level DSLR lenses typically are f/3.5 (at the wide). Some traditions live long.
Cooke Triplet is also basis for all zoom lenses:
Moving center element in between outer elements changes its focal length.
But also aberrations change and are harder to control and it took until 1930s to get first zoom lens into production.
I noticed this too, and there are some other smaller errors. Also, he just misses so many points and sounds disorganized.
Clicked on the video thinking that this is the start to my journey of making the most compact 1.4 primes to be ever made. I’m just gonna stop while I’m behind.
Thank you. Yet ANOTHER field of interest that I have always had great interest in.
Your experience and expertise is seemingly endless.
Thanks.
Yes, Lenses can be very complicated to design, on one of my many visits to the old Hasselblad Factory, I was told that the Zeiss Biogon 38mm 4.5 took a long time to design and calculate, the amount of A4 papers used for calculation was a stack of one meter high ! I had the Zeiss Mutar 2x teleconverter, after I told the folks at Hasselblad how good it was on the Zeiss Planar 110mm 2.0, they got very interested and wanted to learn more, later I was told that Hasselblad had bought a computer from Canada to calculate Lenses. Today all Lens makers use computer software to calculate Lenses, and Lenses today outperform all older Lenses.
Really interesting. I've been a photographer for decades, and I'm a little ashamed to say that however mind-boggling I always found lens techniques to be, I know way too little about the 'glass' that I use. Thanks and thumbs up.
Very interesting explanation. Always wondered what a field flattener does to light. Thanks for sharing.
Thanks, Time has never been that fast. So enjoyable to watch and learn
Hey, just wanted to say thanks a ton for teaching in such an awesome way. You made those tough concepts a breeze to grasp.
Great to hear!
All the years I did 35mm never thought about this. Did alot of macro shootings. Very interesting material. Thanks.
I have been looking for this for sooooo long!
Fascinating... Well done. I really learned a great deal. Thanks for taking time to do this for us.
Incredibly educational. I'm new to the camera hobby. (3 years). And was curious to find an explanation. Thank you.
I just came to check what the elements and groups mean in lenses... Ended up watching the whole video and learnt so much!
This is the best vedio on optical engineering. No one ever just gave this 15 min lecture...yiu are a god to me man
Fascinating. Thanks for sharing
Another trick they've been doing recently for mirrorless cameras is that since you've got to apply digital processing to the sensor image's anyway, some corrections can be done in software and the lens design can be optimized for other factors. If you squint this is a analogous to the curved film trick you mention in the video.
I know they correct for distortion and cos4, not sure about focus
Thank you very much for all the information and details, very well explained, it would be great to find more videos like this.
Very interesting! I'd also like to learn about aperture, for example, why a narrower aperture improves image quality, why resolution goes down when it improves image quality, and so on. Thank you very much!
Fantastic explanation. Would love to see more!
Very cool. I've wondered about this, off and on, for decades and never got around to looking it up! One thing I *think* might be missing is that you can use materials with different indexes of refraction for each lens to tweak it even more.
yes, glass with different dispersion, to correct for color aberrations. classically referred to as crowns and flints. I talk about it a little: ruclips.net/video/PxqhA37bKtI/видео.html
Likewise. Interesting to note that the more glass elements you have in a lens, the worse the image quality. It goes something like this: more glass = less light (in terms of F stops.
Wow! Very cool! Always wondered why so many lenses!
This is absolutely amazing thank you! I am in awe
Thanks for your explanation!
At 06:50
Yes, the use a lot of elements for a single lens. More so if you consider that they use a very narrow band UV light source. Chromatic aberration is therefore limited,
EDIT: Good to see that you mention it also.
AMAZING! Please more videos like this!!
VERY GOOD AND QUICK LENS EDUCATION !
Super interesting, thanks for posting.
I like the aperture logo
Awesome explanation, thanks
this is phenomenal, thank you! Subscribed!
"The Cooke triplet is a photographic lens designed and patented in 1893 by Dennis Taylor who was employed as chief engineer by T. Cooke & Sons of York".
Fascinating stuff! Thank you! I wish there were more videos on RUclips about optics and optical design.
I could teach you how to do lens design and loose all my viewers. They only care about RF. In silicon valley I can count the lens designers on one hand. My buddy in San Diego designed one of the mars rover lenses. He is quite proud.
Fully agree that their should be more exellent youtoube video's like this about optics. Maybe a second channel for optic geeks only?
@@IMSAIGuy I’m one of those lens designers in Silicon Valley. There are more of us these days, but not that many good ones. 😁
Very Informative video, the kind of video I was looking for
That's really interesting how you explain how lenses is working, so most people can understand, or at least have an idea about it.
I subscribe to your channel because you have a lot of good interesting content.
Greetings from Kenneth (Denmark)
this explains pretty well tho why it took a (comparatively) long time for compact cameras to gain lenses that went wider than 35mm (equiv.)
guess it's just that - harder and more expensive to make light bend around alot more at wider angles
Thanks for the information!
I watched a video a few years ago about how lenses are made. They mentioned that someone was experimenting with using a curved sensor and that that was the future. He said that they would have to redesign the lenses to work with it and you'd only be able to use lenses designed for it. But as you know, companies like to get existing customers into newer models. If they're willing to sell all of their gear to move to something totally new they might just leave and go somewhere else. You have to make it easy to adapt. Look at mirrorless for example. You can adapt your old mount lenses to the new mount and still move to mirrorless. And mirrorless cameras have made designing lenses better because the focusing point is closer to the sensor. Or at least I think it would be easier. I shoot Nikon so I mostly know about Nikon lenses. I know that the z lenses are sharper corner to corner than the old f mount lenses. There's little to no chromatic aberration or focus breathing with the new lenses vs the old lenses. There's less vignette. I think mirrorless cameras made curved sensors non essential.
I really enjoyed this! Thank you very much :)
Amazing video! Would love to see a video about a curved image sensor would change this. Also about how the lens in our eye works.
Ahh this is next level stuff !! Thank you good sir
This is so cool! Thanks for sharing. I believe the optical coating is also important, isn’t it to ensure the light entering the lens is within its comfortable zone (generally using thin film reflection to bounce back the other ones)
coating does two things
1. increases the light transmission.
2. decreases stray light reflections
Very nice explanations :)
Wow ! Very interesting ! Thank you.
I have an Astro-Physics apochromatic triplet, with the 3 lens elements in contact with each other and oil spaced to reduce internal reflections, and improve transmission. I love this thing! For as much as it costs, the field flattener accessory comes with quite a breathtaking price tag; but it was intended for medium format film photography at the focal plane. Fortunately, at the time I acquired it, only APS-C sized CCD devices were practical (affordable!) and the part of the focal plane I was using was "flat enough".
Later I move to using a Riccardi-Honders astrograph, an f/3.8 optical system and quite a novel design. You might mistake it for a Schmidt-Cassagrain but it's actually more like a folded refractor with a rear silvered lens, and more importantly, I believe all spherical lens figures. And a nice, large and wide flat focal plane! Only suitable for astrophotography, but that's the problem I had. Never had an eyepiece in it. I asked the designer of that telescope why he didn't use carbon fiber or some other material with low coefficient of expansion due to temperature for the tube assembly, rather than aluminum? That was a deliberate choice, the dimensional change of the tube due to temperature offset the changes in the optical elements as they also cooled down. There is quite a bit of "art" to go along with the science. Difference between theory and practice and all that stuff you learn the hard way.
Optics are cool! Really nice optics are very cool and expensive! And definitely one of those fields where there's no single "best" tool/design. It's always a tradeoff over a bunch of factors, including $$$ it seems.
very few people on the planet can do a good job of optomechanical design.
Now consider the trade offs for something like euv. Where they reduced the stack to just 7 reflexion lenses because it is so hard to refract the light without loss
Thanks for the information. I'll pursue Omnivision. Optically, the setup is simple : a 20D lens is held in front of the eye being examined while a cell phone camera is positioned 30 mm or so away from the 20D lens to capture the aeriali image produced by the 20 D.. The cell phone illumination and auto focus do the job. Some people capture the image as video and look for the best frames. Thanks for your suggestions.
Thanks, awesome content.
Very good! Thanks!
Taylor, Taylor & Hobson was the name (Cooke Triplets)
Great video, very interesting!
Bend Digital Sensors can be made to simplify the lense construction.
11:45 they also use this for another effect, the fibers aren't actually straight thru, instead they're all twisted up to turn the image right side up.
Ive watched this video a lot. There's some smart people in the world man
I have only one word for you sir "Genius"
no genius, I did optics for 30 years.
Awesome, thanks!
Nice, I just discovered this. Good video.
Off the top of my head, two other famous cameras that use a curved film plane are Agfa Clack and Agfa Click. Of course these were just really cheap, essentially toy cameras with single element lenses (or at best an achromat) but pictures are sharp thanks to the curved plane. I still use my Click from time to time.
6:22 i mean a RC mirror set is pretty great and it inly has 2 elements.
Just wow
Very impressive
I find your videos helpful, because I'm trying to design an eyepiece that magnifies the image of a camera viewfinder. I want to mount the eyepiece ontop of the viewfinder. A simple telescope doesn't do the job, because the aperture is too small and it cuts off some parts of the virtual image.
So interesting ! I am searching for the specs of a Canon FD 85mm 1.8 lens in order to manufacture the front element. Do you have any clue where to find it ? I would be so grateful
You will not find any specs. They are secret. Most large camera companies have front elements for sale. especially for expensive pro lenses. Having a single element made would cost more than the lens is worth.
Good show great content
Thanks for the clear explanation of why we need so many layers of lens in cameras. I have a stupid question, I am wondering why our eyes can do it so good with only one lens? Is the lens in our eyes any special?
ruclips.net/video/wZ1dcruxYR0/видео.html
Very interesting, thanks. That field flattener looks a bit like a Schmidt plate. Am I on the right track?
They look similar but are quite different. The Schmidt plate is at the aperture stop of the system. There is a Fourier transform from the aperture to the image. the image is in x-y space and the aperture is in angle space. The incoming wavefront will be phase corrected by the plate to correct for spherical abberation and I think astigmatism and coma. it is difficult to grind an aspheric mirror so the job is given to a thin piece of glass that can be warped, ground flat, and unwarped to result in the shape.
@@IMSAIGuy Thanks for the explanation. I'ii have to try and find out a bit more about this topic.
imsaiguy: Thank you for this informative video. I just have only a cursory interest in lens design. I have not heard "IMSAI" for many years, this was the beginnings of the micro-processing revolution started by Intel, Zilog etc. started in the mid-1970s.
this is why I love youtube!
Wow, that was great
I found this very informative and interesting, thank you. I recently did a BINDT CAT 1 IR thermography course, cat 1 being the starter level! I intended to dive deeper but are the lens makups very similar to photography lenses or is the thermal camera purely dependent on the sensor technology to achieve a desirable image? Guess I may need to do the Cat 2 course!!
resolution of thermal cameras is low compared to film or CCD. so lenses are simple (one element). they are also made of exotic material (Germanium).
This was so fascinating! Thanks for making this video. :) I could watch a whole series on lenses.
Looks like curved camera image sensors are going to be a thing in the next decade or so from the recent porotypes that were shown off. Wonder how that will change lens design.
Thankyou! Could you suggest how one gets more coma in a simple cooke design?
more off axis astigmatism
Awesome job. Can you share the prescription you recovered from that last sectioned lens?
that was decades ago
it is a type of Zeiss Sonnar
@@IMSAIGuy Thank you, Sir!
I think something that's crazy about the lens design, is that you didn't even mention focussing/zooming yet
this was a very simplistic look and only really addressed one aberration (field flatness). Yes zoom lenses add way more complexity and lens elements.
@@IMSAIGuy That's right! It reminded me of the microscopy course I took at uni, where we covered some of the abarations and limitations you have to deal with when working on optics.
You might have heard this already, but I recently came across some information that sony might be working on creating curved digital imaging sensors! People speculate that could lead to some impressively small optical designs for future cameras (maybe in phones too)
I think I found a new wormhole to dive into. Any advice on how to go about doing more studying of this type of stuff? Are there any companies that'd make the lens elements to order?
I would get the book Modern Optical Engineering by Warren Smith if you are interested in lens design. Yes there are prototype houses for lens elements. It is way too expensive for DIY stuff. you can get some generic stuff at places like Edmund Optics
11:02 digital: It's not impossible
It's just very cost prohibitive with the current manufacturing methods.
Like swapping carbon nanotubes for the fiber plain. You'd have to grow the nanotubes on satellites.
well explained! could SLR or DSLR lens designers use the cell phone field flattener lens type to make a minimalist sise lens for full frame cameras? too hard to grind complex shape in bigger lenses?
Field flatners have been used for 100 years
Now they can use software to correct images and use much simpler glass arrangements. Saves weight and complexity.
It might work in practice, but does it work in theory?
Do gasses between glass elements act as their own element as well? ie Would different gasses between glass elements change things?
I have not heard of gases. I have seen liquid.
I would be interested in more videos like thsi, i have been trying to research building a triplet telescope for a while but i havnt found any triplet sets so the idea is to find out how to select individual elements, problem is nobody actually explains it well at all.
Are there any open source tools for this? Or any actually useful resources?
I would get the book Modern Optical Engineering by Warren Smith if you are interested in lens design.
CAD tools I've used: ZEMAX, OSLO, Code V, LightTools. once in a while MATLAB or Octave for specialized calculations and color theory. OLSO may have a trial version, you can also try WinLens3D
Hey there. It is my first time on your channel and I am quite interested in learning more. Namely, I am interested in the detailed process of how you reverse engineered this lens, what software do you use, and the like. I am currently trying to design simple LED lenses, which do not require image fidelity and only usually attempt to narrow the light beam from a beam of 120 degrees to a one of 30 degrees or so. I was experimenting with the aspheric lens equation and got somewhat stuck. I was using open source software and mathematical solvers like sage. If you find the time to talk more about lens design and, for example, what equations, curvatures, and different types of glass there is, I would love to watch this content. The field leveller glass, for example, trades width for depth to achieve the outcome but how does it do that and do the surface follow an equation. Are lens surfaces all mathematical and how are they obtained (I mean, how are the formulas obtained). At a minimum, I would like to know what computational tools and/or software you use. Thanks!
I used ZEMAX, OSLO, Code V, LightTools. once in a while MATLAB or Octave for specialized calculations and color theory. OLSO may have a trial version, you can also try WinLens3D
this video was really great I have watched all your lens-related videos and really like them can you make one about UV lens protectors what effect do they have? thanks
also, why do reversed lenses work so well for macro?
short answer is the lens is already happy having the lens close to the film. so turning is has lens close to subject.
what do you want to know about uv filters?
Can you give the name of the optical design program? Would be fun to dabble in it
Zemax
I wish you had covered the Tessar lens.
just a triplet with an achromat back element.
But it’s a great lense.@@IMSAIGuy
Great video!
Has anyone explained how the focal lenght of a camera lens works? Like, where do you "fit" say 800mm of distance in a 800mm lens? Because most often, you cant, most lenses aren't even as long as their focal lenght.
(Unless ofc there is some glass trickery involved)
Excellent video, I would love to learn more about how to go about designing lenses. I am currently working on a project that uses Kodak disposable camera lenses and adapts them to other cameras.
get the book Modern Optical Engineering, by Warren Smith
get the software 'Zemax'
study for about 5-10 years
You could try a program could OSLO. It's an old version of a program like Zemax. It lets you model up to 10 surfaces for free. So if you're doing it unprofessionally that would be a good starting point. I use Zemax and it's great, but the license is expensive.
@@andrewdavies2358 my boss used OSLO and I used Zemax. we would both run optimizations to see if either has missed anything. Didn't know about the free 10 surfaces. that barely lets you design a triplet, but certainly would be useful for the astronomy crowd.
11:52 the fiber plane reminds me of a mineral called ulexite, also known as "TV rock". Would it be possible to make a fiber plane out of this mineral?
This is cool. I've been writing a renderer (for fun) and want to add realistic camera models. This kind of reverse engineering would be an awesome starting point. And you could try different parameters in the model, and see what the effect is. I'm not modelling wave effects so much, and no quantum effects, but it would be fun to play with.
the big boy lens design CAD packages allow you to render an image through a design. 2D only but it gives you some insight. 3D is possible for non-sequential ray tracing but will not have many of the aberrations included.
@@IMSAIGuy the fun part for me is the learning and design/coding. :-)
My code isn't intended for lens design. It's intended for rendering scenes (vfx). Being able to include a realistic lense helps match a simulated scene to a real one, and also adds aristic range
Where do I start if I want to measure a lens and figure out its parameters? I have a huge aerial photography lens, the first element is nearly 6 inches in diameter, and aside from focal length and aperture, I'm a bit in the dark. The front and rear halves unscrew from the shutter (no adjustable aperture, just open-close). Taking a giant lens spanner to the front, it's just two elements. The rear has 3 elemens, one single and two glued together. Since it's fixed focus as well, I'm pondering how to make it adjustable focus - the easy answer might be to mount the lens on a rack / bellows type movement, but I've been testing it out on the kitchen counter using a sheet of paper to gauge the focus and a ruler to see how much movement is required, and for the range of distances I'd like to use it at, I need over a meter of travel. The bellows would be bigger than the camera and require some support to keep out of the optical path. Just for fun, I calculated the hyperfocal distance and its around a mile.
that is a very complicated process. you need to measure the curvatures of each surface and the thicknesses and spaces, then you will need to know what glass type is used for each element (index and dispersion). these will all require special instruments. once you have these number they will not be accurate enough. you will need to put them in an optics design program to adjust them.
There are also physical limitations. DLRSs have the mirror s the lens is mounted typically at 45 mm from the sensor and the lens cannot go much deeper. Yet some lenses have focal lenghts as small as 10 mm. On telephoto the problem is the opposite. You want to keep the lens short, shorter than its focal length. Zoom lenses naturally add their complications.
The Minox spy camera needed the curved film especially for copying documents.
How eyes produce high resolution image with just cornea and one lens?
ruclips.net/video/wZ1dcruxYR0/видео.htmlsi=zmnkN9-E5hmBH8wP
Which the lens optical design software you used to reverse engineer the design.can u do a video on lens rehousing.there is not much material available about rehousing.or suggest books or online material you know about it.This was one great video.thx
lens cad program: Zemax
mechanical books: Opto-Mechanical Systems Design by Paul R. Yoder, Jr.
Fundamentals of Optomechanics (Optical Sciences and Applications of Light)
by Daniel Vukobratovich
@@IMSAIGuy Thanks you
could you recommend some books on how to design and calculate these lenses?
I would get the book Modern Optical Engineering by Warren Smith if you are interested in lens design.
CAD tools I've used: ZEMAX, OSLO, Code V, LightTools. once in a while MATLAB or Octave for specialized calculations and color theory. OLSO may have a trial version, you can also try WinLens3D
@@IMSAIGuythanks a lot!
Please send links to (free or demo with limited capabilities) optical CAD programs you recommend to start playing with.
Zemax. I don't know if they have a demo version
Current cheap plastic film cameras similar in build to disposable cameras also bend the film.
I wish this video was two hours
Interesting! I know virtually nothing about optics besides what I vaguely remember from undergrad physics class. I suppose nowadays one can use computers to optimize lens design in ways that weren't possible in pre-computer days?? Must be even harder to design good zoom lenses then?
Before computers they used ladies with adding machines. The job title was 'computer'. Each lady would be given one ray to trace. Computers make it much easier to design (and poorly design). I've seen lots of 'kids' that get an optics cad program and think they can design lenses. It doesn't go well. Yes, zoom design is a real art.
WA lenses has more complicated design mainly due to existenece of a mirror in SLRs which mimposes higher distance from lens to film/matrix. Extremly wide lens for large format or rangefinder camera is build of 6 to 8 lenses.
the SLR back focus requirements usually adds about or or two elements. of course the large format lens does not need as high of an MTF which helps. The Hasselblad HCD 24mm f/4.8 Lens still has 14 elements. and the Voigtlander Nokton 21mm rangefinder is 13 elements.
@@IMSAIGuy my experiences with LF lenses are that theirs sharpness is on par with SLR (not only medium format lenses), while design is much simpler (6 or 7 lens Gauss design - Schneider Super Angulon). 35mm lenses are more complicated, probably due to the vignette canceling formula (LF WA lenses use a central gray filter).
I'm curious: Is that a mathematically accurate statement to say that a perfect lens doesn't exist? If that's the case, is it also true if you replace surfaces with a smoothly changing refractive index gradient?
gradient index lenses have their own problems and are also not perfect.
I think using mirror lenses could get more corrective thanks to lack of dispersion effects.