Thread: Who is the father of all fast 50mm lenses? Planar vs Opic. Lens evolution.

Reply to Thread
Page 1 of 2 12 LastLast
Results 1 to 10 of 16
  1. #1 Who is the father of all fast 50mm lenses? Planar vs Opic. Lens evolution. 
    Senior Member Ilya O.'s Avatar
    Join Date
    Nov 2016
    Location
    Moscow
    Posts
    409
    Sometimes you can hear that any 50mm lens is Planar nowadays.
    Zeiss papers on Planar 1.4/50 promote it is as "the most successful lens design ever conceived. This lens type is the basis of almost all professional "workhorse" lenses on earth – and in space."
    Is that really so? Are all contemporary 50mils are really Planars? (NO)
    What is Planar as a brand and as lens design?
    Who is the father of all medium image field lenses?

    Let's dive a little into the history of lens design and all these Gauss, Double Gauss, Planar, Opic, Speed Panchro, Biotar, Xenon...and find who is the pioneer.
    Please participate and criticize, optical designers are WELCOME!

    Despite the popularity of Planar brand now, not many people actually know what was Planar lens, for what purpose it was created at 1896 at Zeiss Jena and why it was NOT popular.
    Was it really a breakthrough? Why every lens article traces Planar as a grandfather of all speed 50mms?

    According to the Patents' (DE 92,313 of 1896; US 583,336 of 1897) preface, Paul Rudolph's of Zeiss general goal was to improve Gauss telescope lens to make it usable for the photographic use.
    Gauss lens diagram (1817):



    Gauss lens is well corrected for chromatic aberration (read: can focus only two colors in one plane), lateral CA, spherical aberration and coma. Still, and this is important for this review, Gauss lens can not correct distortion and astigmatism (and so has high image curvature). It is very slow being f/14 at best. Gauss achromatic doublet was never mass-produced because of its steep curvatures and thin lenses, centration and mounting difficulties: too pricey and unpractical compared to lets say Fraunhofer doublet.

    So designers were playing with symmetrical lens design much at the middle of 1800-s since it corrects distortion and gives more speed (what was so needed for landscape and architecture shots).
    Also, late symmetrical lenses are relatively good corrected for CA, spherical aberration and coma. So there were Rapid Rectilinear of Dallmeyer and Aplanat of Steinheil, 1866, both f/8-f/6:



    The unsolved problem of Recilinear/Aplanat was astigmatism and severe filed curvature: wide open these lenses gave the flat image field of only ~35deg and at best one can get ~60deg when stopped down to hell...So that Aplanats were again not really photographic lenses cause they had very little useful image field and slow.

    Year 1888. Alvan G. Clark applied a patent (US 399,499) for Double Gauss Lens. Despite the hype everywhere in the papers now, the patent is not something optically new nor advanced in computing: Clark made an obvious move by simply placing 2 Gauss objectives of 1817 symmetrically in a nice tube. Here left and right parts are absolutely equal in lens' size and its curvatures:



    One can think from the draft that the spacing between the right lenses differs from the left, but it is not: the Patent says directly so. So Clark's 'invention' is in the fact that 2 Gauss telescope lenses (crown first mod) could become a photographic lens when placed in a tube! What he got was slow (~f/7.5-f/8) wide lens (covered 60deg image plane). Clark's Double Gauss lens shared the fate of Gauss lens, it was a poor seller, barely seen in Europe. Strangely enough, this Clark's Double Gauss lens is now considered as the basis of all modern fast Gauss lens designs, — I do not get it, guys.

    1890. Conquering the photographic market, Paul Rudolph of Zeiss was trying to overcome Recilinear/Aplanat unsolved aberrations with its Anastigmat (=no astigmatism) lenses (Patents DE 56,109; US 444,714), known as Protar since 1900. Max speed was f/4.5. Utilizing revolutionary 'new achromat' glass types for achromatisation and low spherical aberration, he went unsymmetrical route where left component has positive astigmatic difference, and the right achromatic doublet, triplet or even quadruplet has negative so that one cell corrects the other for astigmatism, and overall lens' performance achieves image flatness over quite big field of view:



    Mid 1890-s. When everybody was developing new unsymmetrical lenses utilising 'new achromat' glass types, and the ground braking Cooke Triplet f/3.7 of Harold Dennis Taylor was computed at 1893 (GB 22,607), Rudolph forgets his unsymmetrical Anastigmats and goes way back to the symmetry: he took Gauss lens of 1817 as a basis for new lens design! The symmetry allows to 'easy' correct coma, lateral CA and distortion.

    So Rudolph relates not to Double Gauss lens of Clark (1888), but to Gauss telescope lens of 1817 which with its 4 lens curvatures and 2 glass types for play, gives nice correction of chromatic & spherical aberration and, what is not widely known, of spherochromatism — Gauss lens corrected the spherical aberration for 2 different colors (cause, you know, spherical ab. is different for different wavelength). OKAY then, but to make it suitable for photo use, Rudolph must make it faster, distortionless, with anastigmatically flattened image, also chromatic & spherical aberration free. So, step one from the famous Planar Patent (DE 92,313 of 12 Nov. 1896; US 583,336 of 1897; GB 27,635 of 1896) was this lens:



    What Rudolph did here is make all lenses thicker, put the iris in front, use the outer curvatures to fight astigmatism and spherical aberration.
    So now the curvatures and thickness are fixed and pre-determined, but how on earth then to correct chromatism?! Besides, with the glass types available that time, Rudolph could not make this design achromatic...but then took the genius idea he patented a bit earlier (6 July 1896, US 576,896): to split a meniscus into the cemented doublet made of glass with (nearly)identical refractive index but different dispersion. The outer radii are equal to the original meniscus, but now Rudolph got an additional curvature (or in "buried surface") inside the cemented doublet (r4)! Varying it and dispersion powers, Rudolph corrected the chromatism without touching the previously made correction of the spherical aberration and astigmatism (since they both only affected by refraction, which is equal in both parts of this compound lens(L2&L3)!

    Step two. The lens above has barrel distortion since the iris is in front. And coma is not corrected. To eliminate it, Rudolph chose very simple way: put two described lenses symmetrically around the iris. That is how we all know Zeiss Jena Planar of Paul Rudolph was born, here is lens diagram:



    By the way, for some reason everywhere is written that the first Planar was f/4.5 — that is not true, the patent (DE 92,313 of 12 Nov. 1896) describes f/4 lens ('größte relative Öffnung 0.25'). The aperture of lenses in production was f/3.6 to f/5. Planar covered 52-72deg of image plane. Was it good? In a way, yes: when stopped down quite much (f/12 and more), it was the best lens on earth for reproduction, process work, for microscopy (you got flat field, unrivaled resolution and zero distortion). Can you imagine the Zeiss themselves did not recommended Planar for exterior shots?! They said better use Planar for portraiture work! Well, people were not since there were better and much faster lenses for that purpose...So Planar was not popular in photo/cine field. Wide open Planar was too glowy (coma is not corrected and can not be corrected in this symmetrical design, but Rudolph believed that), very low contrast, flare and mirror images due to the curvature's direction of the eight (8!) air-glass uncoated surfaces.

    Well then, everybody (designers and users) forgot Planar design till 1920 when Horace William Lee (Taylor, Taylor & Hobson Ltd.) patented Cooke Series 0 (zero) lens (Lee Opic lens) as fast as f/2. That is, not Biotar/Planar, the father of all later built speed cine and photo lenses of medium image field because of being asymmetrical. Here is Opic lens diagram as patented (GB 157,040):



    Looks symmetrical like the same old Planar, isn't it? No, it is not:-) You can not take the right lens' part, duplicate it and put it flipped behind like it was possible with Planar cells. Look at the curvatures (10 radii, r1 to r10) — they are all different! While Planar had only 5 unique radii in the front cell, the back was a symmetrical copy. The only symmetry left in Opic lens is the glass types, check the nD (refractive index) on the right and mirrored to the left.

    Yet, Opic is an anastigmat with flat image field and a successor of Planar, and Lee says that "the present invention is an improvement" over the Planar type. Lee Opic lens is 4 times faster and still has low spherical aberration, it is coma and CA corrected. Lee made it possible, first, when he left symmetry to have more radii for coma and other mono aberrations correction and to gain more speed; second, he chose other glass types: dense barium crown with high refractive index for all collective components and light flint glass for others (+made them thicker) to correct chromatic & spherical aberration.

    TTH Cooke Series 0 Anastigmat had unrivaled speed and performance that time. Despite the 35 mm (1.375'') f/2 lens for 35mm silent movie, TTH considered Ser.0 lineup (50, 76, 108, 140mm) as large format still lenses for some time:



    Now we know that was a big mistake!..
    [end of part 1]
    Last edited by Ilya O.; 07-24-2018 at 01:20 PM.
    Reply With Quote  
     

  2. #2  
    Senior Member
    Join Date
    Oct 2013
    Location
    Athens, Greece
    Posts
    216
    Fascinating and very informative read, thanks for this! (and all your other threads)
    Reply With Quote  
     

  3. #3  
    Senior Member Kwan Khan's Avatar
    Join Date
    Jan 2008
    Location
    NYC
    Posts
    3,761
    Thanks
    No, I do not RENT MY GEARS... - NYC (Times Square)

    EPIC-D with 9" touch, 4 RAVENs, RPP, Sigma 18-35 PL, Motion Mount, VF/RED-rock wireless FF, Pana 17", JVC 20", Seadicam with SEGWAY, Letus 4-Axis with L'Aigle Exoskeleton.

    Trash-Can 12-core Max'd-out) +MBP (Retina), RED Rocket (Sonnet), RAID system (Promise Pegasus II), Adobe Production (CC), FCS2, Resolve 12 with MC Color, 50" 4K Panel
    Reply With Quote  
     

  4. #4  
    Senior Member Ilya O.'s Avatar
    Join Date
    Nov 2016
    Location
    Moscow
    Posts
    409
    [part 2]
    Now we know that was a big mistake!

    Year 1925. German designers were overtaking and it is usually Albrecht Wilhelm Tronnier of Jos. Schneider & Co. Optische Werke with Xenon design who is named to follow and advance Lee Opic lens design. Well, I doubt that and I need help of a German native speaker. Please read the Tronnier Patent of 30 April of 1925 № 439,556, compare it to Lee's Opic Patent 157,040 and correct me if I'am wrong: Tronnier describes the same as Opic not-for-the-movies large image field lens (55-66deg), he registers one half of a lens, a triplet (100mm, f/4.5), which can be doubled (thus f~50mm, +twice more speed) around the iris and one can better choose "semi-symmetrical" for best correction; then he repeats Lee's invention concerning glass refractive characteristics but with slightly different numbers (0.02 for refractive index difference instead of 0.03 of Lee). Well, that was nothing new, here is Xenon lens diagram:



    What Schneider really did better than Taylor & Hobson is put new asymmetrical double-gauss lens into production for movie use really quick! Here is the optical scheme of "ultra-fast" Xenon f/1.8 for Cinematography which was on sale as early as 1926:



    On the 30th of September, 1927, Carl Zeiss Jena applied the Patent № 485,798 (in Britain: 1928 GB 297,823) for a f/1.4 (!) photographic lens which is known now as Biotar f/1.4, with the Willy Walter Merté as the lens designer behind. Funny enough, Zeiss describes its invention as a lens consisting out of 4 (four) menisci lenses, outer are collective, inner are dispersive — well, that we know since Clark's 1888 Double Gaus lens:-). Seems somebody tried not to be in a conflict with 1920 Lee Opic Patent:-) So this is original Biotar lens diagram, both of equal speed:



    Off course, you see here not 4 but 6 or 7 elements for all aberrations correction. What's new here comparing to Opic? Aside from the even higher refractive index glass usage, look at the radii and its correlations. Merté bended lenses curvatures higher than of Lee Opic, and put "distinctly biconvex" lenses and "distinctly meniscal" (=concave lens) in the Biotar lens to gather large aperture. Besides, this bending solves the ghosting problems of Planar lens by forwarding the reflections outside of the image field area (but still, Biotar is uncoated so that reflections only appear as diffused light/glow)

    Was Biotar a step forward Opic? Yes, it definitely was because it was solely made for 16&35mm movie cameras (Opic was not), Biotar's 35-42deg image field makes a perfect base for standard movie lens (Opic was ~50deg), and remember the main, a really shocking speed of f/1.4! Moreover, Biotar has very good spherical aberration correction for that aperture.

    From this point it seems Biotar of Zeiss was the speed king back then at 1927 and Merté was a pioneer, and Lee at Cooke (TTH) was doing nothing with its Opic till 1930 when he patented Speed Panchros lenses for particular and only movie use. At least that seems so from the short abstract about double gaussian lens evo in "A History of the Photographic Lens" by Rudolph Kingslake. But it is not, Cooke was not sleeping at 1927:-)

    It is not widely known, but Lee patented f/1.5 lens on 12th of October, 1927 (GB 298,769; US 1,779,257) — simultaneously with Merté's Biotar (Pat. applied on 30th of September)!!! And who knows, TTH (actually Kapella Limited that days) was applying many patents that time, and there could be more earlier ones to be discovered...maybe one day a Cooke representative read this thread and could kindly give us more info?
    So please enjoy Cooke's 1927th 'superspeed' f/1.5 pre-Panchro lens diagram:



    Notice the both internal dispersive components: each of them consisting out of 3 cemented elements! And both have large and thick biconvex, or "distinctly biconvex" as Merté names it, element. A nice example of independent design decisions yet resolved different. More, count the radii of Lee's speed lens, it is 12, all different, a broad horizon for aberrational manoeuvre:-)
    Last edited by Ilya O.; 06-25-2017 at 10:32 PM.
    Reply With Quote  
     

  5. #5  
    Junior Member
    Join Date
    Jun 2017
    Location
    Taipei
    Posts
    7
    Quote Originally Posted by Ilya O. View Post
    What Schneider really did better than Taylor & Hobson is put new asymmetrical double-gauss lens into production for movie use really quick! Here is the optical scheme of Cine Xenon f/1.8 which was on sale as early as 1926:
    The "Cine-Xenon" was a post war production, not in 1926. According to British Journal Almanac 1930, just 5 years later to patent DE439556, Schneider still use the name Xenon on cinematograph lens. Those early Xenon f/1.8 lenses looked identical to Lee's design. It could be a trick to avoid Lee's patent, that Schneider only applied "half lens" in that patent.
    Attached Images
    Reply With Quote  
     

  6. #6  
    Senior Member Ilya O.'s Avatar
    Join Date
    Nov 2016
    Location
    Moscow
    Posts
    409
    Thanks for the input, W. Su!

    You are right about Cine-Xenon as post-War brand name. Before WWII there was only Xenon as a "Cinematograph Lens".

    My error goes from the App061 lens diagram of Xenon as given in "A Lens Collector's Vade Mecum", it is entitled as "Schneider Cine Xenon f1.8 (1926). So, it is a kind of misunderstanding (the title means Xenon for Cine, not Cine-Xenon as a lens brand), and anachronism (there were no Cine-Xenons as lens name back at 1926).

    W. Su, could you please give more pictures or any info on early double-Gauss Xenons?
    Last edited by Ilya O.; 08-07-2017 at 03:48 AM.
    Reply With Quote  
     

  7. #7  
    Senior Member
    Join Date
    Dec 2006
    Posts
    4,735
    Very interesting thread Ilya O and W. Su
    DOP / Digital Lighting Camera Operator
    London, UK.

    Life moves pretty fast. If you don't stop and look around once in a while, you could miss it.
    Reply With Quote  
     

  8. #8  
    Junior Member
    Join Date
    Jun 2017
    Location
    Taipei
    Posts
    7
    The family tree of fast 50mm lenses should be:

    1920 TTH OPIC f/2 GBP157,040 → 1925 Schneider Xenon f/1.8 DEP439,556 & 1927~8 Zeiss Biotar f/1.4(DEP485,798 not in real product)



    1927 TTH Ultra Speed Panchro 50/1.4 GBP298,769 → 1935 Schneider Xenon 50/1.9 USP2,106,077



    1930 TTH Super Speed Panchro 58/1.3 GBP373,950 (USP2,019,985) → 1959 Zeiss Planar 55/1.4 DEP1,170,157 & 1964 Canon 50/0.95 JPP1964,010,178



    1964/66 Asahi Super Takumar 50/1.2 & 1.4 USP3,451,745

    →Modern 50/1.4 with 7 elements in 6 groups


    Although DEP485,798 was believed as Zeiss patent for Biotar, that patent never went into production. I have analyzed some first batch of Biotar 4cm/1.4 (sn. 808xxx, 1927) and first batch of Biotar 5cm/1.4 (sn. 916xxx, 1928), their optical design were very similar to Lee's Opic, not as asymmetrical as DEP485,798 described.

    British patent 298,769 has a product in real world, the TTH Ultra Speed Panchro 50/1.4. Like many British patents, it only has f/1.5 in patent document, but has f/1.4 in production. This should be the world's first 50mm f/1.4 lens, a new fast design and one-year earlier than Biotar 5cm production.

    The TTH Ultra-SP is very complicated, there is a modified lens design in patent GBP373,950, the TTH Super Speed Panchro 58/1.3. It has an additional glass over Opic design. When Super-SP licensed to Leica, the aperture was limited to f/1.5 only. The latter one was the famous Leica Xenon 5cm/1.5, each lens engraved with "Taylor-Hobson British Patent 373950 and U.S. Patent 2019985". From TTH Super-SP, the 7 elements in 5 groups formula became a standard of most fast lenses till 1960s, variations including Zeiss Planar 55/1.4 and very large aperture Canon-7 50/0.95.

    In 1935 a new Schneider f/1.9 Xenon came from the TTH Ultra-SP design. It removed a glass inside the front cemented group to an air separate, lower the lens speed to f/1.9 only (USP2,106,077). It has a simplified version 5cm f/2, replace the triple cemented lens by a doublet, become a 6 elements in 5 groups lens. However, Schneider only kept the patent of f/1.9, did not file patent for Xenon 5cm/2. Xenon's designer Tronnier took the f/2 optical scheme to Voigtlander after WWII, that was famous Voigtlander Ultron 50/2 (1950 USP2,627,204). Although Schneider is the first company to produce 6 elements with 5 groups lens, they lost in patent war.

    In 1964, Japanese Asahi Optical combined the TTH SSP type with 1935 Xenon type, that was Asahi Pentax Super Takumar 50/1.4. Its first version used to have 8 elements, then modified to 7 elements. The Super Takumar type (7 elements in 6 groups) soon became popular till nowadays, variations including the very fast Leica Noctilux 50/1.
    Last edited by W. Su; 06-28-2017 at 10:53 AM.
    Reply With Quote  
     

  9. #9  
    Senior Member Ilya O.'s Avatar
    Join Date
    Nov 2016
    Location
    Moscow
    Posts
    409
    Mr. Su, how do you know which patent or lens design went or not to production? What are your sources? And how do the actual production sample differs from the originating patent?

    Besides, it's interesting how you analyze lens diagram (you mentioned Biotar to Cooke)? What's the procedure? X-ray? I mean how do you compare lens curvatures in glued cells?
    Last edited by Ilya O.; 08-07-2017 at 03:51 AM.
    Reply With Quote  
     

  10. #10  
    Junior Member
    Join Date
    Jun 2017
    Location
    Taipei
    Posts
    7
    In our optical research work, we have collected many samples of each type of lenses, give us a chance to compare the differences between product and patent document. But how can we know the optical difference of each lens? That leads to your later question: Yes, we do use high energy rays and optical software to analyze them.

    For example, according to 1896 Deutsche patent 92133, the diagram of original Zeiss Planar should be:
    Attached Images
    Reply With Quote  
     

Posting Permissions
  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts