A camera lens (also known as photographic lens or photographic objective) is an optical lens or assembly of lenses used in conjunction with a camera body and mechanism to make images of objects either on photographic film or on other media capable of storing an image chemically or electronically.
While in principle a simple convex lens will suffice, in practice a compound lens made up of a number of optical lens elements is required to correct (as much as possible) the many optical aberrations that arise. Some aberrations will be present in any lens system. It is the job of the lens designer to balance these out and produce a design that is suitable for photographic use and possibly mass production.
There is no major difference in principle between a lens used for a still camera, a video camera, a telescope, a microscope, or other apparatus, but the detailed design and construction are different.
A lens may be permanently fixed to a camera, or it may be interchangeable with lenses of different focal lengths, apertures, and other properties.
Typical rectilinear lenses can be thought of as "improved" pinhole lenses. As shown, a pinhole lens uses a tiny aperture to block most rays of light, ideally selecting one ray to the object for each point on the image sensor. Pinhole lenses would be excellent except for a few serious limitations:
Such lenses can be thought of as an answer to the question "how can we modify a pinhole lens to admit more light and give higher resolution?" A first step is to put a simple convex lens at the pinhole with a focal length equal to the distance to the film plane (assuming the camera will take pictures of distant objects). This allows us to open up the pinhole significantly (below right) because the convex lens bends light in proportion to its angle of incidence on the lens. The geometry is almost the same as with a simple pinhole lens, but rather than being illuminated by single rays of light, each image point is illuminated by a focused "pencil" of light rays. Standing in front of the camera, you would see the small hole, the aperture. The virtual image of the aperture as seen from the world is known as the lens's entrance pupil; ideally, all rays of light leaving a point on the object that enter the entrance pupil will be focused to the same point on the image sensor/film (provided the object point is in the field of view). If one were inside the camera, one would see the lens acting as a projector. The virtual image of the aperture from inside the camera is the lens's exit pupil.
Practical photographic lenses include more lens elements. The additional elements allow lens designers to reduce various aberrations, but the principle of operation remains the same: pencils of rays are collected at the entrance pupil and focused down from the exit pupil onto the image plane.
A camera lens may be made from a number of elements: from one, as in the Box Brownie's meniscus lens, to over 20 in the more complex zooms. These elements may themselves comprise a group of lenses cemented together.
The front element is critical to the performance of the whole assembly. In all modern lenses the surface is coated to reduce abrasion, flare, and surface reflectance, and to adjust color balance. To minimize aberration, the curvature is usually set so that the angle of incidence and the angle of refraction are equal. In a prime lens this is easy, but in a zoom there is always a compromise.
The lens usually is focused by adjusting the distance from the lens assembly to the image plane, or by moving elements of the lens assembly. To improve performance, some lenses have a cam system that adjusts the distance between the groups as the lens is focused. Manufacturers call this different things. Nikon calls it CRC (close range correction), while Hasselblad and Mamiya call it FLE (floating lens element).[1]
Glass is the most common material used to construct lens elements, due to its good optical properties and resistance to scratching. Other materials are also used, such as quartz glass, fluorite,[2][3][4][5] plastics like acrylic (Plexiglass), and even germanium and meteoritic glass.[6] Plastics allow the manufacturing of strongly aspherical lens elements which are difficult or impossible to manufacture in glass, and which simplify or improve lens manufacturing and performance. Plastics are not used for the outermost elements of all but the cheapest lenses as they scratch easily. Molded plastic lenses have been used for the cheapest disposable cameras for many years, and have acquired a bad reputation: manufacturers of quality optics tend to use euphemisms such as "optical resin". However many modern, high performance (and high priced) lenses from popular manufacturers include molded or hybrid aspherical elements, so it is not true that all lenses with plastic elements are of low photographic quality.
The 1951 USAF resolution test chart is one way to measure the resolving power of a lens. The quality of the material, coatings, and build affect the resolution. Lens resolution is ultimately limited by diffraction, and very few photographic lenses approach this resolution. Ones that do are called "diffraction limited" and are usually extremely expensive.[7]
Today, most lenses are multi-coated in order to minimize lens flare and other unwanted effects. Some lenses have a UV coating to keep out the ultraviolet light that could taint color. Most modern optical cements for bonding glass elements also block UV light, negating the need for a UV filter. UV photographers must go to great lengths to find lenses with no cement or coatings.
A lens will most often have an aperture adjustment mechanism, usually an iris diaphragm, to regulate the amount of light that passes. In early camera models a rotating plate or slider with different sized holes was used. These Waterhouse stops may still be found on modern, specialized lenses. A shutter, to regulate the time during which light may pass, may be incorporated within the lens assembly (for better quality imagery), within the camera, or even, rarely, in front of the lens. Some cameras with leaf shutters in the lens omit the aperture, and the shutter does double duty.
The two fundamental parameters of an optical lens are the focal length and the maximum aperture. The lens' focal length determines the magnification of the image projected onto the image plane, and the aperture the light intensity of that image. For a given photographic system the focal length determines the angle of view, short focal lengths giving a wider field of view than longer focal length lenses. The wider the aperture, identified by a smaller f-number, allows using a faster shutter speed for the same exposure.[8]
The maximum usable aperture of a lens is specified as the focal ratio or f-number, defined as the lens' focal length divided by the effective aperture (or entrance pupil), a dimensionless number. The lower the f-number, the higher light intensity at the focal plane. Larger apertures (smaller f-numbers) provide a much shallower depth of field than smaller apertures, other conditions being equal. Practical lens assemblies may also contain mechanisms to deal with measuring light, secondary apertures for flare reduction,[9] and mechanisms to hold the aperture open until the instant of exposure to allow SLR cameras to focus with a brighter image with shallower depth of field, theoretically allowing better focus accuracy.
Focal lengths are usually specified in millimetres (mm), but older lenses might be marked in centimetres (cm) or inches. For a given film or sensor size, specified by the length of the diagonal, a lens may be classified as a:
An example of how lens choice affects angle of view. The photos above were taken by a 35 mm camera at a constant distance from the subject.
A side effect of using lenses of different focal lengths is the different distances from which a subject can be framed, resulting in a different perspective. Photographs can be taken of a person stretching out a hand with a wideangle, a normal lens, and a telephoto, which contain exactly the same image size by changing the distance from the subject. But the perspective will be different. With the wideangle, the hands will be exaggeratedly large relative to the head. As the focal length increases, the emphasis on the outstretched hand decreases. However, if pictures are taken from the same distance, and enlarged and cropped to contain the same view, the pictures will have identical perspective. A moderate long-focus (telephoto) lens is often recommended for portraiture because the perspective corresponding to the longer shooting distance is considered to look more flattering.
The complexity of a lens—the number of elements and their degree of asphericity—depends upon the angle of view and the maximum aperture, among other variables including intended price point. An extreme wideangle lens of large aperture must be of very complex construction to correct for optical aberrations, which are worse at the edge of the field and when the edge of a large lens is used for image-forming. A long-focus lens of small aperture can be of very simple construction to attain comparable image quality; a doublet (with two elements) will often suffice. Some older cameras were fitted with "convertible" lenses of normal focal length; the front element could be unscrewed, leaving a lens of twice the focal length and angle of view, and half the aperture. The simpler half-lens was of adequate quality for the narrow angle of view and small relative aperture. Obviously the bellows had to extend to twice the normal length.
Good-quality lenses with maximum aperture no greater than f/2.8 and fixed, normal, focal length need at least three (triplet) or four elements (the trade name "Tessar" derives from the Greek tessera, meaning "four"). The widest-range zooms often have fifteen or more. The reflection of light at each of the many interfaces between different optical media (air, glass, plastic) seriously degraded the contrast and color saturation of early lenses, zoom lenses in particular, especially where the lens was directly illuminated by a light source. The introduction many years ago of optical coatings, and advances in coating technology over the years, have resulted in major improvements, and modern high-quality zoom lenses give images of quite acceptable contrast, although zoom lenses with many elements will transmit less light than lenses made with fewer elements (all other factors such as aperture, focal length, and coatings being equal).[11]
Many Single-lens reflex cameras, and some rangefinder cameras have detachable lenses. A few other types do as well, notably the Mamiya TLR cameras. The lenses attach to the camera using a lens mount, which often also contains mechanical or electrical linkages between the lens and camera body. The lens mount is an important issue for compatibility between cameras and lenses; each major camera manufacturer typically has their own lens mount which is incompatible with others; notable exceptions are the Leica M39 lens mount for rangefinders, M42 lens mount for early SLRs, the later Pentax K mount, and the Four Thirds System mount for dSLRs, all of which are used by multiple camera brands. Most large-format cameras take interchangeable lenses as well, which are usually mounted in a lensboard or on the front standard.
A macro lens used in macro or "close-up" photography (not to be confused with the compositional term "Close up") is any lens that produces an image on the focal plane (i.e., film or a digital sensor) that is the same size or larger than the subject being imaged. This configuration is generally used to image close-up very small subjects. A macro lens may be of any focal length, the actual focus length being determined by its practical use, considering magnification, the required ratio, access to the subject, and illumination considerations. They can be special lens corrected optically for close up work or they can be any lens modified (with adapters or spacers) to bring the focal plane "forward" for very close photography. The depth-of-field is very narrow, limiting their usefulness. Lenses are usually stopped down to give a greater depth-of-field.[12][13]
Some lenses, called zoom lenses, have a focal length that varies as internal elements are moved, typically by rotating the barrel or pressing a button which activates an electric motor. Commonly, the lens may zoom from moderate wide-angle, through normal, to moderate telephoto; or from normal to extreme telephoto. The zoom range is limited by manufacturing constraints; the ideal of a lens of large maximum aperture which will zoom from extreme wideangle to extreme telephoto is not attainable. Zoom lenses are widely used for small-format cameras of all types: still and cine cameras with fixed or interchangeable lenses. Bulk and price limit their use for larger film sizes. Motorized zoom lenses may also have the focus, iris, and other functions motorized.
An ideal lens would image an object, point for point, with absolute accuracy in relative space. However, the laws of physics, the state of our knowledge of those laws, the limits of engineering, as well as the practical considerations of size, weight and cost, mean that no real lens can be ideal. The first century of the history of the photographic camera lens can be understood as a slow increase of optical knowledge; enough to bring optical aberrations of real lenses to an acceptable level. The second century of the history of the photographic camera lens can be regarded as technical applications of that knowledge; to slowly increase the variety and versatility of real lenses.
Note, the curvatures and spacing in the lens block diagrams are all approximate. In addition, they do not indicate the glass used. In other words, it is not possible to construct a usable lens solely from the diagrams. Note also, almost all the lens names given were trademarks; many are still properties of their respective owners and are used for identification purposes only.
The history of the photographic camera lens began with the Wollaston Meniscus. The single element concavo-convex Meniscus was invented in 1804 by William Hyde Wollaston (UK). It was first used for eyeglasses and was the first to be reasonably sharp over a wide field (about 50° at f/16) lens. Wollaston fitted it to an artist's aid camera obscura in 1812.[14]
Turned around so that the concave surface faced forward and with a front aperture stop, the Meniscus is called the first photographic lens because it was fitted to some of the camera obscuras adapted by Joseph Nicéphore Niépce (France) to his pioneering "heliography" experiments. The meniscus shape corrected the field curvature that limited the acceptably sharp field of the simple biconvex lens used on camera obscuras since Giambattista della Porta (modern Italy) in 1550.[15] Note, Niépce did not switch to a Meniscus until 1828;[16][17] he made the first permanent photograph on a bitumen photosensitized pewter plate in 1826 or 1827 with a biconvex lens.[18][19] Meniscus lenses were and are still used in simple focus-free box cameras, including innumerable Kodak Brownies.
Niépce and Louis-Jacques-Mandé Daguerre (France) shared the same optical supplier, Charles Chevalier (France), and Daguerre's daguerreotype experiments also began using camera obscuras with Meniscus lenses. However, the refractive index of glass increases from red to blue of the light spectrum (color dispersion). Blue is focused closer to the lens than red causing rainbow-like color fringing (chromatic aberration).[20][21][22] The lack any chromatic aberration control in a Meniscus meant it was impossible to focus accurately – the daguerreotype process was blue sensitive only, while the human eye focused primarily using yellow.[23]
Chevalier suggested a changeover to a Dollond Achromat Doublet (originally a telescope objective) in 1829. Although it is not inherently sharper than a Meniscus, an Achromatic Doublet cements a positive element of low refractive index and low dispersion crown (soda-lime) glass to a negative one of high refractive index and high dispersion flint (lead) glass to cancel out enough of their individual chromatic aberrations to bring blue and yellow to a common focus.
Modern photographic achromats (since about 1900) are normally designed to bring blue and red together – specifically 486 and 656 nanometer wavelengths.[24][25] Note, although John Dollond (UK) received the British Royal Society's Copley Medal in 1758 for the 1754 discovery (Isaac Newton [UK] had concluded in 1666 that the chromatic aberration of lenses was unsolvable),[26] the true inventor of the Achromat was Chester Moor Hall (UK) in 1729.[27][28]
On 22 June 1839, Daguerre contracted Alphonse Giroux (France) to manufacture his official daguerreotype apparatus, including the world’s first production photographic camera. The Giroux Le Daguerreotype camera used an almost 16 inch focal length (about 40 cm) f/16 Achromatic Doublet made by Chevalier to take 6½×8½ inch (about 16.5×21.5 cm) images.[29][30][31] An Achromat Doublet was also the specified lens in the official daguerreotype instructions issued by the French government 19 August 1839.[32][33] Chevalier would add a meniscus curve to the Achromat by the end of 1839 to combine field flattening and chromatic aberration control, and create the standard outdoors lens of the nineteenth century – the Achromat Landscape.[34][35]
The Achromat Landscape was hardly perfect. It was quite slow – its f/16 working aperture required twenty to thirty minute outdoor daguerreotype exposures – and the French Society for the Encouragement of National Industry offered an international prize in 1840 for a faster lens. Joseph Petzval (modern Hungary) was a mathematics professor without any optical physics experience but, with the aid of several human computers of the Austro-Hungarian army, took up the challenge of producing a lens fast enough for a daguerreotype portrait.
He came up with the Petzval Portrait (modern Austria) in 1840, a four element formula consisting of a front cemented achromat and a rear air-spaced achromat that, at f/3.6, was the first wide aperture, portrait lens. It was appropriate for one-to-two minute shaded outdoors daguerreotype exposures. With the faster colloidion (wet plate) process of 1851, it could take one-to-two minute indoor portraits. Due to national chauvinism, the Petzval did not win the prize, despite being far superior to all other entries.[36]
A 150mm focal length Petzval lens was fitted to a conical metal Voigtländer (modern Austria) camera taking circular daguerreotypes in 1841. The Voigtländer-Petzval was the first camera and lens specifically designed to take photographs, instead of being a modified artist's camera obscura.[37][38][39] The Petzval Portrait was the dominant portrait lens for nearly a century. It had what would now be considered severe field curvature and astigmatism. It was centrally sharp (about 20° field of view, 10° for critical applications), but quickly drifted out of focus to a soft outer field, producing a pleasant halo effect around the subject. The Petzval Portrait remains popular as a projection lens where the narrow angles involved means the field curvature does not matter.[40]
The Portrait was illegally copied by every lens maker and Petzval had a nasty falling out with Peter Voigtländer over unpayable royalties and died an embittered old man.[41] Although the Portrait was the first mathematically computed lens formula (not trial and error),[42] trial and error would continue to dominate photographic lens design for another half century, despite well established physical mathematics dating from 1856 (by Ludwig von Seidel [modern Germany], working for Hugo Adolph Steinheil [modern Germany]), to the retrospective detriment of lens advancement.[43]
The Achromat Landscape was also afflicted with rectilinear distortion – straight lines were imaged as curved. This was a pressing problem as architecture was an important photography subject early on – buildings do not move, making them popular to photograph with the early slow processes.[44] In addition, photographs of faraway places (especially in stereoscope form[45]) were a popular means to see the world from the comfort of one's home – the picture postcard is a mid-19th century invention.[46] The distortion got progressively worse as the field of view increased, which meant the Achromat Landscape could not be used as a wide angle lens.
The first successful wide angle (92° maximum field of view; 80° was more realistic) lens was the Harrison & Schnitzer Globe (USA) of 1862,[47] although with f/16 maximum aperture (f/30 was more realistic). Charles Harrison and Joseph Schnitzer's Globe had a symmetric four element formula – the name refers to the fact that if the two outer surfaces were continued and joined, they would form a sphere.[48][49]
Symmetry was discovered in the 1850s to automatically correct three (distortion, coma and transverse chromatic)[50] of the seven major lens aberrations (five monochromatic "Seidel sums": spherical, coma, astigmatism, field curvature and rectilinear distortion; plus two chromatic: axial [or longitudinal color] and transverse [or lateral color]) that prevent the formation of sharp images by simple lenses.[51][52][53] There are also decentration aberrations arising from manufacturing errors. A real lens will not produce images of expected quality if it is not constructed to or cannot stay in specification.[54] The more complex the design, the more sensitive it is to improperly polished or aligned elements.
There are additional optical phenomena that can degrade image quality but are not considered aberrations. For example, the oblique cos4θ light falloff, sometimes called natural vignetting,[55][56] and lateral magnification and perspective distortions seen in wide angle lenses are really geometric effects of projecting three dimensional objects down into two dimensional images, not physical defects.[57]
The Globe's symmetric formula directly influenced the design of the Dallmeyer Rapid-Rectilinear (UK) and Steinheil Aplanat (modern Germany). By coincidence, John Dallmeyer's Rapid-Rectilinear and Adolph Steinheil's Aplanat had virtually identical symmetric four element formulae, arrived at almost simultaneously in 1866, that corrected most optical aberrations, except for spherical and field curvature, to f/8. The breakthrough was to use glasses of maximum refractive index difference but equal dispersion in each achromat. The Rapid-Rectilinear and Aplanat were scalable over many focal lengths and fields of view for all contemporaneous film formats, and were the standard moderate-aperture, general purpose lenses for over half a century.[58][59]
The Landscape, the Portrait, the Globe and the Rapid-Rectilinear/Aplanat constituted the nineteenth century photographer's entire lens arsenal.[60]
It was known in the 1500s that an aperture stop would improve lens image quality.[61] It would be discovered that this was because a center stop that blocks off peripheral light rays limits the transverse aberrations (coma, astigmatism, field curvature, distortion, and lateral chromatic) unless the stop is so small that diffraction becomes dominant.[62] Even today, most lenses produce their best images at their middle apertures, at a compromise between transverse aberrations and diffraction.[63]
Therefore even the Meniscus had a permanent stop. However, the earliest lenses did not have adjustable stops, because their small working apertures and the lack of sensitivity of the daguerreotype process meant that exposure times were measured in many minutes. A photographer would not want to limit the light passing through the lens and further lengthen the exposure time. When the increased sensitivity wet colloidion process was invented in 1851, exposure times shortened dramatically and adjustable stops became practical.[64]
The earliest selectable stops were the Waterhouse stops of 1858, named for John Waterhouse. These were sets of accessory brass plates with sized holes mounted through a slot in the side of the lens.[65][66]
Around 1880, photographers realized that aperture size affected depth of field.[67] Aperture control gained much more significance and adjustable stops became a standard lens feature. The iris diaphragm made its appearance as an adjustable lens stop in the 1880s. It became the standard adjustable stop about 1900. The iris diaphragm had been common in early nineteenth century artists' aid camera obscuras and Niépce used one in at least one of his experimental cameras.[68] However, the specific type of iris used in modern lenses was invented in 1858 by Charles Harrison and Joseph Schnitzer.[69] Harrison and Schnitzer's iris diaphragm was capable of rapid open and close cycles, an absolute necessity for lenses with camera auto-aperture control.[70]
The modern lens aperture markings of f-numbers in geometric sequence of f/1, 1.4, 2, 2.8, 4, 5.6, 8, 11, 16, 22, 32, 45, 64, 90, etc. was standardized in 1949. Previously, this British system competed with the Continental (German) sequence of f/1.1, 1.6, 2.2, 3.2, 4.5, 6.3, 9, 12.5, 18, 25, 36, 50, 71, 100 ratios. In addition, the Uniform System (U.S., invented UK) sequence of 1, 2, 4, 8, 16, 32, 64, 128, etc. (where U.S. 1 = f/4, U.S. 2 = f/5.6, U.S. 4 = f/8, etc.), was favored by Eastman Kodak early in the twentieth century.[71][72][73]
A single-element camera lens is as long as its focal length; for example, 500 mm-focal-length lens requires 500 mm from the lens to the image plane. A telephoto lens is made physically shorter than its nominal focal length by pairing a front positive imaging cell with a rear magnifying negative cell. The powerful front group over-refracts the image, the rear restores the focal plane, thereby greatly shortening the back-focus length.[74] Originally, accessory negative cells were sold to attach to the rear of a regular lens. The Barlow lens, a negative achromat magnifier invented by Peter Barlow in 1833, is still sold to increase the eyepiece magnification of amateur telescopes.[75] The teleconverter is the modern photographic equivalent.[76][77]
In 1891, Thomas Dallmeyer and Adolph Miethe simultaneously attempted to patent new lens designs with nearly identical formulae – complete photographic telephoto lenses consisting of a front achromat doublet and rear achromat triplet. Primacy was never established and no patent was ever granted for the first telephoto lens.[78]
The front and rear cells of early telephotos were unmatched and the rear cell also magnified any aberrations, as well as the image, of the imaging cell. The cell spacing was also tunable, because that could be used to adjust the effective focal length, but that only worsened aberration problems. The first telephoto lens optically corrected and fixed as a system was the f/8 Busch Bis-Telar (Germany) of 1905.[79]
The photographic lens leapt forward in 1890 with the Zeiss Protar (Germany).[80] Paul Rudolph's Protar was the first successful anastigmat (highly corrected [for the era] for all aberrations, including properly for astigmatism) lens. It was scalable from f/4.5 portrait to f/18 super wide angle. The Protar was originally called the Anastigmat, but that descriptive term quickly became generic and the lens was given a fanciful name in 1900.[81]
The Protar is considered the first "modern" lens, because it had an asymmetric formula allowed by the new design freedom opened up by newly available barium oxide, crown optical glasses.[82] These glasses were invented by Ernst Abbe, a physicist, and Otto Schott, a chemist, (both Germany) in 1884, working for Carl Zeiss' Jena Glass Works. Schott glasses have higher refractive index than soda-lime crown glass without higher dispersion. The Protar's front achromat used older glass, but the rear achromat used high index glass.[83] Virtually all good quality photographic lenses since circa 1930 are anastigmat corrected. (The primary exceptions are deliberately "soft-focus" portrait lenses.)
Today's photographic lens state-of-the-art is apochromatic correction, which is, very roughly, twice as strict as anastigmatic.[84] However, such lenses require correcting for higher ordered aberrations than the original seven[85] with rare earth (lanthanum oxide) or fluorite (calcium fluoride) glasses of very high refractive index and/or very low dispersion of mid-twentieth century invention.[86][87][88] The first apochromatic lens for consumer cameras was the Leitz APO-Telyt-R 180mm f/3.4 (1975, West Germany) for Leicaflex series (1964, West Germany) 35mm SLRs.[89] Most professional telephoto lenses since the early 1980s are apochromatic.[90][91] Note, better-than-apochromat lenses are available for scientific/military/industrial work.[92]
The quintessential twentieth century photographic lens was the 1893 Taylor, Taylor & Hobson Cooke Triplet.[93] Dennis Taylor's (UK, not related to the Taylors of T, T & H) Cooke Triplet was a deceptively simple looking asymmetric three element anastigmat formula created by reexamining lens design from first principles to take maximum advantage of the advances in new Schott optical glasses. The elements were all of such strong power that they were highly sensitive to misalignment and required tight manufacturing tolerances for the era.[94]
The Cooke Triplet became the standard "economy" lens of the twentieth century. For example, the Argus Cintar 50mm f/3.5 for the Argus C3 (1937, USA), probably the best-selling rangefinder camera of all time, used a Cooke triplet.[95]
The Triplet was adequate for contact prints from medium format roll film cameras and small enlargements from 35mm "miniature" format cameras, but not for big ones. The films of the first half of the twentieth century did not have much resolving power either, so that was not necessarily a problem.
Paul Rudolph developed the Tessar from dissatisfaction with the performance of his earlier Protar,[96] although it also resembles the Cooke triplet. The Tessar was originally an f/6.3 lens. It was refined to f/2.8 by 1930, although f/3.5 was the realistic limit for best image quality.[97]
The Tessar was the standard high-quality, moderate-aperture, normal-perspective lens of the twentieth century. The Kodak Anastigmat Special 100mm f/3.5 on the Kodak Super Six-20 (1938, USA), the first autoexposure still camera, was a Tessar[98] the D. Zuiko 2.8 cm f/3.5 on the Olympus Pen (1959, Japan), the original Pen half frame camera;[99] the Schneider S-Xenar 40mm f/3.5 on the late version of the Rollei 35 (1974, West Germany/Singapore);[100][101] and the AF Nikkor D 45mm f/2.8P Special Edition for the Nikon FM3A (2001, Japan), the last manual focus 35mm SLR released by a major maker.[102] It was fitting that the Zeiss Stiftung's last camera, the Zeiss Ikon S 312, had a Zeiss Tessar 40mm f/2.8 (1972, West Germany).[103]
It is often incorrectly stated that the Leitz Elmar 50mm f/3.5 fixed to the Leica A (1925, Germany), Leitz's first camera, was a Tessar.[104]. However, at the time the Leica was introduced the 50mm f/3.5 Kino Tessar had only been designed to cover the cine format of 18x24mm, which was insufficient for the new 24x36mm format of the Leica, and Leitz had to develop a new lens to provide adequate full frame coverage. It was only when Zeiss Ikon were designing the Contax in response to the success of the Leica that a 50mm Tessar which could cover the 24x36mm format was designed. The Elmar was based on a modified Cooke Triplet with a different computation to the Tessar and with the stop in the first air space.[105]
With anastigmat image quality achieved, attention next turned to increasing aperture size to allow photography in lower light or with faster shutter speeds. The first common very wide aperture lens suitable for candid available light photography was the Ernemann Ernostar (Germany) of 1923.[106] Ludwig Bertele's formula was originally a 10 cm f/2 lens, but he improved it to 10.5 cm and 85mm f/1.8 in 1924.[107] The Ernostar was also a Cooke Triplet derivative; it has an extra front positive element or group.[108]
Mounted on the Ernemann Ermanox (1923, Germany) camera and in the hands of Erich Salomon, the Ernostar pioneered modern photojournalism. French Premier Aristide Briand once said: "There are just three things necessary for a[n international] conference: a few Foreign Secretaries, a table and Salomon."[109] Note, American photojournalists favored flash use into the 1950s (see Arthur Fellig [Weegee]).
Bertele continued Ernostar development under the more famous Sonnar name after Ernemann was absorbed by Zeiss in 1926. He reached f/1.5 in 1932 with the Zeiss Sonnar 50mm f/1.5[110][111] for the Contax I 35mm rangefinder camera (1932, Germany).[112]
The Sonnar was (and is) also popular as a telephoto lens design – the Sonnar is always at least slightly telephoto because of its powerful front positive elements. The Zeiss Olympia Sonnar 180mm f/2.8 for the Contax II (both 1936, Germany) is a classic, if not mythic, example.[113]
In 1817 Carl Friedrich Gauss improved the Fraunhofer telescope objective by adding a meniscus lens to its single convex and concave lens design. Alvan Clark further refined the design in 1888 by taking two of these lenses and placing them back to back. The lens was named in honour of Gauss. The current design can be traced back to 1895, when Paul Rudolph of Carl Zeiss Jena used cemented doublets as the central lenses to correct for chromatic aberration.
Later the design was developed with additional glasses to give high-performance lenses of wide aperture. The main development was due to Taylor Hobson in the 1920s, resulting in the f/2.0 Opic and later the Speed Panchro designs, which were licensed to various other manufacturers. The design forms the basis for many camera lenses in use today, especially the wide-aperture standard lenses used on 35 mm and other small-format cameras. It can offer good results up to f/1.4 with a wide field of view, and has sometimes been made at f/1.0.
The design is presently used in inexpensive-but-high-quality fast lenses such as the Canon EF 50mm f/1.8 and Nikon 50 mm f/1.8D AF Nikkor. It is also used as the basis for faster designs, with elements added, such as a seventh element as in both Canon[114] and Nikon's 50 mm f/1.4 offerings[115] or an aspherical seventh element in Canon's 50 mm f/1.2.[116] The design appears in other applications where a simple fast normal lens is required (~53° diagonal) such as in projectors.
Surface reflection was a major limiting factor in nineteenth century lens design. With a four to eight percent (or more) reflective light loss at every glass-air interface dimming the light transmission plus the reflected light scattering everywhere producing flare, a lens would not be of practical use with more than six or eight losses. This, in turn, limited the number of elements a designer could use to control aberrations.[117]
Some lenses were marked by T-stops (transmission stops) instead of f-stops to indicate the light losses.[118] T-stops were "true" or effective aperture stops and were common for motion picture lenses,[119] so that a cinematographer could ensure that consistent exposures were made by all the different lenses used to make a movie. This was less important for still cameras and only one still lens line was ever marked in T-stops: for the Bell & Howell Foton 35mm rangefinder camera. Bell & Howell was normally a cinematographic equipment maker. The Foton's standard lens was the Taylor, Taylor & Hobson Cooke Amotal Anastigmat 2 inch f/2 (T/2.2) (1948; camera USA; lens UK, a Double Gauss).[120] The quarter stop difference between f/2 and T/2.2 is a 16% loss.
It was noticed by Dennis Taylor in 1896 that some lenses with glass tarnished by age counterintuitively produced brighter images. Investigation revealed that the oxidation layer suppressed surface reflections by destructive interference.[121][122] Lenses with glass elements artificially "single-coated" by vacuum deposition of a very thin layer (approximately 130-140 nanometers[123]) of magnesium or calcium fluoride to suppress surface reflections[124] were invented by Alexander Smakula working for Zeiss in 1935[125][126] and first sold in 1939.[127] Antireflection coating could cut reflection by two-thirds.[128]
In 1941, the Kodak Ektra (USA) 35mm RF was introduced with the first complete antireflection coated lens line for a consumer camera: the Kodak Ektar 35mm f/3.3, 50mm f/3.5, 50mm f/1.9, 90mm f/3.5, 135mm f/3.8 and 153mm f/4.5.[129] World War II interrupted all consumer camera production and coated lenses did not appear in large numbers until the late 1940s. They became standard for high quality cameras by the early 1950s.
The availability of antireflection coating permitted the Double Gauss to rise to dominance over the Sonnar. The Sonnar had more popularity before World War II because, before antireflection coating, the Sonnar's three cell with six air-glass surfaces versus the Double Gauss's four and eight made it less vulnerable to flare.[130] Its telephoto effect also made the lens shorter, an important factor for the Leica and Contax 35mm RFs designed to be compact.
As maximum aperture continued to increase, the Double Gauss's greater symmetry promised easier aberration correction. This was especially important for SLRs because, without the parallax error of RFs, they also began offering much closer focusing distances (typically a half meter instead a whole meter).[131] The Double Gauss became the preferred normal lens design in the 1950s with the availability of antireflection coating and new generation extra high refractive index rare earth optical glasses.[132]
Coating lenses with up to a dozen or more different layers of chemicals to suppress reflections across the visual spectrum (instead of at only one compromise wavelength) were a logical progression. Asahi Optical's SMC Takumar lenses (1971, Japan) were the first all multicoated (Super-Multi-Coated) lenses for consumer cameras (M42 screw mount Asahi Pentax SLRs).[133] Modern highly corrected zoom lenses with fifteen, twenty or more elements would not be possible without multicoating.[134][135] The transmission efficiency of a modern multicoated lens surface is about 99.7% or better.[136]
Antireflection coating does not relieve the need for a lens hood (a conical tube slipped, clipped, screwed or bayoneted onto the front of a lens to block non-image forming rays from entering the lens) because flare can also result from strong stray light reflecting off of other inadequately blacked internal lens and camera components.[137][138][139]
Regular wide angle lenses (meaning lenses with focal length much shorter than the format diagonal and producing a wide field of view) need to be mounted close to the film. However, SLR cameras require that lenses be mounted far enough in front of the film to provide space for the movement of the mirror (the "mirror box"); about 40 mm for a 35mm SLR compared to less than 10 mm in non-SLR 35mm cameras. This prompted the development of wide field of view lenses with more complex retrofocus optical designs. These use very large negative front elements to force back-focus distances long enough to ensure clearance.[140][141]
In 1950, the Angénieux Retrofocus Type R1 35mm f/2.5 (France) was the first retrofocus wide angle lens for 35mm SLRs (Exaktas).[142] Except for the front element, Pierre Angénieux' R1 was a five element Tessar. Note, "retrofocus" was an Angénieux trademark before losing exclusive status. The original generic term was "inverted" or "reversed telephoto." A telephoto lens has a front positive cell and rear negative cell;[143] retrofocus lenses have the negative cell in front and positive cell to the rear.[144] The first inverted telephoto imaging lens was the Taylor, Taylor & Hobson 35mm f/2 (1931, UK) developed to provide back-focus space for the beamsplitter prism used by the full-color via three negatives Technicolor motion picture camera.[145] Other early members of the Angénieux Retrofocus line included the 28mm f/3.5 Type R11 of 1953 and the 24mm f/3.5 Type R51 of 1957.[146]
Retrofocus lenses are extremely asymmetric with their large front elements and therefore very difficult to correct for distortion by traditional means. On the upside, the large negative element also limits the oblique cos4θ light falloff of regular wide-angle lenses.[147][148][149]
Retrofocus design also influenced non-retrofocus lenses. For example, Ludwig Bertele's Zeiss Biogon 21mm f/4.5,[150] released in 1954 for the Contax IIA (1950, West Germany) 35mm RF, and its evolution, the Zeiss Hologon 15mm f/8[151] of 1969, fixed to the Zeiss Ikon Hologon Ultrawide (West Germany), were roughly symmetrical designs. However, each half can visualized as retrofocus. The Biogon and Hologon designs take advantage of the large negative elements to limit the light falloff of regular wide angle lenses.[152][153] With a 110° field of view, the Hologon would otherwise have had a 3¼ stop corner light falloff, which is wider than the exposure latitude of contemporaneous films. Nonetheless, the Hologon had a standard accessory radially graduated 2 stop neutral density filter to ensure completely even exposure. The distance from the Hologon's rear element to the film was only 4.5 mm.[154]
Many normal perspective lenses for today's digital SLRs are retrofocus, because their smaller-than-35mm-film-frame image sensors require much shorter focal lengths to maintain equivalent fields of view, but the continued use of 35mm SLR lens mounts require long back-focus distances.
A fisheye lens is a special type of ultra-wide angle retrofocus lens with little or no attempt to correct for rectilinear distortion. Most fisheyes produce a circular image with a 180° field of view. The term fisheye comes from the supposition that a fish looking up at the sky would see in the same way.[155]
The first fisheye lens was the Beck Hill Sky (or Cloud; UK) lens of 1923. Robin Hill intended it to be pointed straight up to take 360° azimuth barrel distorted hemispheric sky images for scientific cloud cover studies.[156] It used a bulging negative meniscus to compress the 180° field to 60° before passing the light through a stop to a moderate wide angle lens.[157] The Sky was 21mm f/8 producing 63mm diameter images.[158] Pairs were used at 500 meter spacing producing stereoscopes for the British Meteorological Office.[159]
Note, it is impossible to have 180° rectilinear coverage because of light falloff. 120° (12mm focal length for the 35mm film format) is about the practical limit for retrofocus designs; 90° (21mm focal length) for non-retrofocus lenses.[160]
Strictly speaking, macrophotography is technical photography with actual image size ranging from near life-size (1:1 image-to-object ratio) to about ten or twenty times life-size (10 or 20:1 ratio, at which photomicrography begins). "Macro" lenses were originally regular formula lenses optimized for close object distances, mounted on a long extension tube or bellows accessory to provide the necessary close focusing, but preventing focusing on distant objects.[161]
However, the Kilfitt Makro-Kilar 4 cm f/3.5 (West Germany/Liechtenstein) of 1955 for Exakta 35mm SLRs changed the everyday meaning of macro lens.[162] It was the first lens to provide continuous close focusing. Version D of Heinz Kilfitt's (West Germany) Makro-Kilar focused from infinity to 1:1 ratio (life-size) at two inches; version E, to 1:2 ratio (half life-size) at four inches.[163] The Makro-Kilar was a Tessar mounted in an extra long draw triple helical. SLR cameras were best for macro lenses because SLRs do not suffer from viewfinder parallax error at very close focus distances.[164]
Designing close-up lenses is not really that hard – an image size that is close to object size increases symmetry. The Goerz Apo-Artar (Germany/USA) photoengraving process lens was apochromatic in 1904,[165] although ultra-tight quality control helped.[166] It is getting a sharp image continuously from infinity to close-up that is hard – before the Makro-Kilar, lenses generally did not continuously focus to closer than 1:10 ratio. Most SLR lens lines continue to include moderate aperture macro lenses optimized for high magnification.[167] However, their focal lengths tend to be longer than the Makro-Kilar to allow more working distance.[168]
"Macro zoom" lenses began appearing in the 1970s, but traditionalists object to calling most of them macro because they stray too far from the technical definition – they usually do not focus closer than 1:4 ratio with relatively poor image quality.[169][170]
A supplementary lens is an accessory lens clipped, screwed or bayoneted to the front of a main lens that alters the lens' effective focal length. If it is a positive (converging) only supplement, it will shorten the focal length and reset the infinity focus of the lens to the focal length of the supplementary lens. These so-called close-up lenses are often uncorrected single element menisci, but are a cheap way to provide close focusing for an otherwise limited focus range lens.[171][172]
An afocal attachment is a more sophisticated supplementary lens. It is a so-called Galilean telescope accessory mounted to the front of a lens that alters the lens' effective focal length without moving the focal plane. There are two types: the telephoto and the wide angle. The telephoto type is a front positive plus rear negative cell combination that increases the image size; the wide angle has a front negative and rear positive arrangement to reduce the image size. Both have cell separation equal to cell focal length difference to maintain the focal plane.[173][174]
Since afocal attachments are not an integral part of the main lens' formula, they degrade image quality and are not appropriate for critical applications.[175] However, they have been available for amateur motion picture, video and still cameras since the 1950s.[176] Before the zoom lens, afocal attachments were a way to provide a cheap sort of interchangeable lens system to an otherwise fixed lens camera. In the zoom lens era, they are a cheap way to extend the reach of a zoom.
Some afocal attachments, such as the Zeiss Tele-Mutar 1.5× and Wide-Angle-Mutar 0.7× (1963, West Germany) for various fixed lens Franke and Heidecke Rolleiflex brand 120 roll film twin-lens reflex cameras, were of higher quality and price, but still not equal to true interchangeable lenses in image quality. The very bulky Mutars could change a Rolleiflex 3.5E/C's Heidosmat 75mm f/2.8 and Zeiss Planar 75mm f/3.5 (1956, West Germany) viewing and imaging lenses into 115mm and 52mm equivalents.[177][178] Afocal attachments are still available for digital point-and-shoot cameras.[179][180]
The Kodak Retina IIIc and IIc (USA/West Germany) collapsable lens 35mm rangefinder cameras of 1954 took the supplementary lens idea to the extreme with their interchangeable lens "components." This system allowed swapping the front cell component of their standard Schneider Retina-Xenon C 50mm f/2 lenses (a Double Gauss) for Schneider Retina-Longar-Xenon 80mm f/4 long-focus and Schneider Retina-Curtar-Xenon 35mm f/5.6 wide-angle components.[181][182] Component lens design is tightly constrained by the need to reuse the rear cell and the lenses are extremely bulky, range limited and complex compared with fully interchangeable lenses,[183] but the Retina's interlens Synchro-Compur leaf shutter restricted lens options.
The zoom lens is a natural consequence of the telephoto lens, the original lens to manipulate focal length. Varying the spacing between a telephoto's front positive and rear negative cells changes the lens' magnification. However, this will upset focus and aberration optimization, and introduce pincushion distortion. A real zoom lens needs a compensating cell to push the focal plane back to the appropriate place and took decades of development to become practical. The earliest zooms came out between 1929 and 1932 for professional motion picture cameras and were called "Traveling," "Vario" and "Varo" lenses.[184]
The first zoom lens for still cameras was the Voigtländer-Zoomar 36-82mm f/2.8 (USA/West Germany) of 1959,[185] for Voigtländer Bessamatic series (1959, West Germany) 35mm leaf shutter SLRs.[186] It was designed by Zoomar in the United States and manufactured by Kilfitt in West Germany for Voigtländer.[187] The Zoomar 36-82 was very large and heavy for the focal length[188] – 95mm filter size.[189]
Frank Back (Germany/USA) was the early champion of zoom lenses and his Zoomars would hurl far into the future the lance of zoom lens development and popularity, starting with his original Zoomar 17-53mm f/2.9 (1946, USA)[190] for 16mm motion picture cameras.[191] The image quality of early zoom lenses could be very poor – the Zoomar's has been described as "pretty rotten."[192]
Japanese photographic lens production dates from 1931 with the Konishiroku (Konica) Hexar 10.5 cm f/4.5[193] for the Konishiroku Tropical Lily small plate camera. However, the Japanese advanced quickly and were able to manufacture very high quality lenses by 1950[194] – LIFE magazine photographer David Douglas Duncan's "discovery" of Nikkor lenses is an oft-told tale.[195][196][197]
In 1954, the Japan Camera Industry Association (JCIA) began promoting the development of a high quality photographic industry to increase exports as part of Japan's post-World War II economic recovery. To that end, the Japan Machine Design Center (JMDC) and Japan Camera Inspection Institute (JCII) banned the slavish copying of designs and the export of low quality photographic equipment, enforced by a testing program before issuance of shipping permits.[198][199]
By the end of the 1950s, the Japanese were seriously challenging the Germans. For example, the Nippon Kogaku Nikkor-P Auto 10.5 cm f/2.5 of 1959, for the Nikon F 35mm SLR (1959), is reputed to be one of the best portrait lenses ever made, with superb sharpness and bokeh. It originated as the Nikkor-P 10.5 cm f/2.5 (1954) for the Nikon S series 35mm RF, was optically upgraded in 1971 and available until 2006.[200]
In 1963, the Tokyo Kogaku RE Auto-Topcor 5.8 cm f/1.4 came out along with the Topcon RE Super/Super D (1963) 35mm SLR. The Topcor is reputed to be one of the best normal lenses ever made.[201] The Nikkor and the Topcor were sure signs of the Japanese optical industry eclipsing the Germans'. Topcon in particular was highly avant-garde in producing two ultra-fast lenses by 1960 - the R-Topcor 300 F2.8 (1958) and the R-Topcor 135 F2 1960). The former was not eclipsed until 1976. Germany had been the optical leader for a century, but the Germans turned very conservative after World War II; failing to achieve unity of purpose, innovate or respond to market conditions.[202][203] Japanese camera production surpassed West German output in 1962.[204]
Early Japanese lenses were not novel designs: the Hexar was a Tessar; the Nikkor was a Sonnar; the Topcor was a Double Gauss. They began breaking new ground around 1960: the Nippon Kogaku Auto-Nikkor 8.5–25 cm f/4-4.5 (1959), for the Nikon F, was the first telephoto zoom lens for 35mm still cameras (and second zoom after the Zoomar),[205] the Canon 50mm f/0.95 (1961), for the Canon 7 35mm RF, with its superwide aperture, was the first Japanese lens a photographer might lust after,[206][207] and the Nippon Kogaku Zoom-Nikkor Auto 43-86mm f/3.5 (1963), originally fixed on the Nikkorex Zoom 35mm SLR, later released for the Nikon F, was the first popular zoom lens, despite mediocre image quality.[208][209]
German lenses disappear from this history at this point. After ailing throughout the 1960s, such famous German nameplates as Kilfitt, Leitz, Meyer, Schneider, Steinheil, Voigtländer and Zeiss went bankrupt, were sold off, contracted production to East Asia or became boutique brands in the 1970s.[210][211] Names for design types also disappear at this point. Apparently the Japanese are not fans of lens names, they use only brand names and feature codes for their lens lines.[212]
The JDMC/JCII testing program, having fulfilled its goals, ended in 1989 and its gold "PASSED" sticker passed into history.[213] The JCIA/JCII morphed into the Camera & Imaging Products Association (CIPA) in 2002.[214]
Catadioptric photographic lenses (or "CAT" for short) combine many historical inventions such as the Catadioptric Mangin mirror (1874), Schmidt camera (1931), and the Maksutov telescope (1941) along with Laurent Cassegrain's Cassegrain telescope (1672). The Cassegrain system folds the light path and the convex secondary acts as a telephoto element, making the focal length even longer than the folded system and extending the light cone to a focal point well behind the primary mirror so it can reach the film plane of the attached camera. The Catadioptric system, where a spherical reflector is combined with a lens with the opposite spherical aberration, corrects the common optical errors of a reflector such as the Cassegrain system, making it suitable for devices that need a large aberration free focal plane (cameras).
The first general purpose photographic catadioptric lens was Dmitri Maksutov 1944 MTO (Maksutov Tele-Optic) 500mm f/8 Maksutov–Cassegrain configuration, adapted from his 1941 Maksutov telescope.[215] Designs followed using other optical configurations including Schmidt configuration and solid catadioptric designs (made from a single glass cylinder with a maksutov or aspheric form polished into the front face and the back spherical surface silvered to make the "mirror"). In 1979 Tamron was able to produce a very compact light weight catadioptric by using rear surface silvered mirrors, a "Mangin mirror" configuration that saved on mass by having the aberration corrected by the light passing through the mirror itself.[216]
The catadioptric camera lens' heyday was the 1960s and 70s, before apochromatic refractive telephoto lenses. CATs of 500mm focal length were common; some were as short as 250mm, such as the Minolta RF Rokkor-X 250mm f/5.6 (Japan) of 1979 (a Mangin mirror CAT roughly the size of a 50mm f/1.4 lens).[217] The CAT is the only reasonable solution for 1000+ mm lenses.
Dedicated photographic mirror lenses fell out of favor in the 1980s for various reasons. However, commercial reflector astronomical Maksutov–Cassegrain and Schmidt–Cassegrain telescopes with 14 to 20 inch (or even larger) diameter primary mirrors are available. With an accessory camera adapter, they are 4000mm f/11 to f/8 equivalent.[218][219]
Most early zoom lenses produced mediocre, or even poor, images. They were adequate for low resolution requirements such television and amateur movie cameras, but usually not still photography. For example, Nippon Kogaku always apologetically acknowledged that Takashi Higuchi's Zoom-Nikkor Auto 43-86mm f/3.5, the first popular zoom lens, did not meet its normal image quality standards.[220] However, efforts to improve them were ongoing.
In 1974, the Ponder & Best (Opcon/Kino) Vivitar Series 1 70-210mm f/3.5 Macro Focusing Zoom (USA/Japan) was widely hailed as the first professional-level quality very close focusing "macro" zoom lens for 35mm SLRs. Ellis Betensky's (USA) Opcon Associates perfected the Series 1's fifteen element/ten group/four cell formula by calculations on the latest digital computers.[221] Freed from the drudgery of hand computation in the 1960s, designs of such variety and quality only dreamt of by earlier generations of optical engineers became possible.[222][223] Modern computer created zoom designs may be so complex that they have no resemblance to any of the classical human created designs.
The optical zooming action of the Series 1 was different from most earlier zooms such as the Zoomar. The Zoomar was an "optically compensated" zoom. Its zooming cell and focal plane compensating cell were fixed together and moved together with a stationary cell in between.[224] The Series 1 was a "mechanically compensated" zoom. Its zooming cell was mechanically cammed with a focal plane compensating cell and moved at different rates.[225] The tradeoff for greater optical design freedom was this increase in mechanically complexity.
The external controls of the Series 1 were also mechanically more complex than the Zoomar. Most early zooms had separate twist control rings to vary the focus and focal length – a "two touch" zoom. The Series 1 used a single control ring: twist to focus, push-pull to zoom – a "one touch" zoom. For a short time, about 1980-1985, one-touch zooms were the dominant type, because of their ease of handling. However, the arrival of interchangeable lens autofocus cameras in 1985 with the Minolta Maxxum 7000 (Japan; called Alpha 7000 in Japan, 7000 AF in Europe) necessarily forced the decoupling of focusing and zooming controls and two touch zooms made an instant comeback.
In 1977, zoom lenses had advanced far enough that the Fuji Fujinon-Z 43-75mm f/3.5-4.5 (Japan) became the first zoom lens to be sold as the primary lens for an interchangeable lens camera, the Fujica AZ-1 (1977, Japan) 35mm SLR, instead of a prime.[226]
Small quick framing "supernormal" zooms of around 35-70mm focal length became popular 50mm substitutes in Japan by 1980.[227] However, they never gained much of a foothold in the United States,[228] although 70-210mm telephoto zooms were very popular as second lenses. The first auto-everything 35mm point-and-shoot camera with built-in zoom lens, the camera type that dominated the 1990s, was the Asahi Optical Pentax IQZoom (1987, Japan) with Pentax Zoom 35-70mm f/3.5-6.7 Tele-Macro.[229]
The next landmark zoom was the Sigma 21-35mm f/3.5-4 (Japan) of 1981. It was the first super-wide angle zoom lens for still cameras (most 35mm SLRs). Previously, combining the complexities of rectilinear super-wide angle lenses, retrofocus lenses and zoom lenses seemed impossible. The Sigma's all-moving eleven element/seven group/three cell formula was a triumph of computer-aided design and multicoating.[230]
Along with optical complexity, the mechanical complexity of the Sigma, with three cells moving at differing rates, required the latest in manufacturing technology. Super-wide angle zoom lenses are even more complicated for most of today's digital SLRs, because the usually smaller-than-35mm-film-frame image sensors require much shorter focal lengths to maintain equivalent fields of view, but the continued use of 35mm SLR lens mounts require the same large back-focus distances.
Japanese zoom interchangeable lens production surpassed that of prime lenses in 1982,[231] and to say that zooms are ubiquitous today, while primes are not, is stating the obvious.
The increasingly complex internal movements of zoom lenses also inspired improved prime lens designs. Traditionally, prime lenses for rigid cameras were focused closer by physically shifting the entire lens toward the object in a helical or rack and pinion mount. (Cameras with bellows expanded the bellows to shift the lens forward.) However, element spacing for best aberration correction may be different for near versus far objects.
Therefore, some prime lenses of this era began using "floating elements" – zoom-like differential cell movement in nested helicals for better close-up performance.[232] For example, retrofocus wide angle lenses tend to have excessive spherical aberration[233] and astigmatism at close focusing distances and so the Nippon Kogaku Nikkor-N Auto 24mm f/2.8 (Japan) of 1967 for Nikon 35mm SLRs had a Close Range Correction system with a rear three element cell that moved separately from the main lens to maintain good wide aperture image quality to a close focus distance of 30 cm/1 ft.[234]
Other prime lenses began using "internal focusing," such as Kiyoshi Hayashi's Nippon Kogaku Nikkor 200mm f/2 ED IF (Japan) of 1977. Focusing by moving only a few internal elements, instead of the entire lens, ensured the lens' weight balance would not be upset during focusing.[235][236]
Internal focusing was originally popular in heavyweight, wide-aperture telephoto lenses for professional press, sports and wildlife photographers, because it made their handling easier. IF gained all-around significance in the autofocus era, because moving a few internal elements instead of the entire lens for focusing conserved limited battery power and eased the strain on the focusing motor.[237]
Note, floating elements and internal focusing produces a zooming effect and the effective focal length of an FE or IF lens at closest focusing distance can be one-third shorter than the marked focal length.[238]
Bokeh is the subjective quality of the out-of-focus or blurry part of the image. Traditionally, time consuming hand computation limited lens designers to correcting aberrations for the in-focus image only, with little consideration given to the out-of-focus image. Therefore, approaching and outside the specified circle of confusion or depth-of-field, aberrations built up in the out-of-focus image differently in different lens design families. Differences in the out-of-focus image can influence the perception of overall image quality.
There is no precise definition of bokeh and no objective tests for it – as with all aesthetic judgments. However, symmetrical optical formulae such as the Rapid-Rectilinear/Aplanat and the Double Gauss are usually considered pleasing, while asymmetric retrofocus wide angle and telephoto lenses are often thought harsh.[239] The unique "donut" bokeh produced by mirror lenses because of the optical pathway obstruction of the secondary mirror is especially polarizing.[240][241]
In the 1970s, as increasing powerful computers proliferated, the Japanese optical houses began to spare computing cycles to study the out-of-focus image.[242] An early result of these explorations was the Minolta Varisoft Rokkor-X 85mm f/2.8 (Japan) of 1978 for Minolta 35mm SLRs. It used floating elements to allow the photographer to deliberately under-correct the spherical aberration of the lens system and render unsharp specular highlights as smoothly fuzzy blobs without affecting focus or other aberrations.[243] In effect, the Varisoft, and later variable soft focus portrait lenses, attempt to recreate the qualities the Petzval Portrait had accidentally. Note, the Varisoft, except for the floating elements, is a Tessar.
Bokeh is now a normal lens design parameter for very high quality lenses. However, bokeh is virtually irrelevant for the tens of millions of very small sensor digital point-and-shoot cameras sold every year. Their very short focal length and small aperture lenses have enormous depth-of-field – almost nothing is out of focus. Since wide aperture lenses are rare today, most contemporary photographers confuse bokeh with shallow depth-of-field, having never seen either. Many are even unaware of their existence.
Despite the grousing of traditionalists that lenses were better in the past, lenses have improved over time. On average, lenses are sharper today than they were in the past.[244]
The easiest way to prove this is to remember that camera image format sizes have been steadily shrinking over the last two centuries, while standard print sizes have stayed about the same. It is therefore obvious that today's lenses must have higher resolving power than lenses of past eras to maintain an equal level of print quality with the required higher level of enlargement. For example: the human eye can resolve about five lines per millimeter at distance of one foot (about 30 cm). Therefore, a lens must produce a minimum resolution of forty lines per millimeter on a 24×36 mm 35mm film negative if it to provide a linear enlargement of eight times to an 8×10 inch (about 20×25 cm) print and still appear sharp when viewed at one foot.[245] A lens for an APS-sized (about 16×24 mm) digital SLR sensor needs a minimum resolution of fifty-two lines per millimeter to be enlarged thirteen times to a sharp 8×10 inch print.
Another way to understand how lenses have improved is to know the level of analysis that optical engineers devote to their lens formulae. In the nineteenth century, opticians dug to the level of the Seidel aberrations, called mathematically the third order aberrations, to reach basic anastigmatic correction. Opticians needed to calculate for the fifth order aberrations by the mid-twentieth century to produce a high quality lens.[246] Today's lenses require seventh order aberration solutions.[247]
Note, the best photographic lenses from forty or fifty years ago were already of very high image quality (twice the minimum resolution mentioned above) and it may not be possible to conclusively demonstrate the superiority of the best of today's lens without comparing 20×30 inch (about 50×75 cm) enlargements of exactly the same scene side by side.[248][249]
Typical lens elements have spherically curved surfaces. However, this causes off-axis light to be focused closer to the lens than axial rays (spherical aberration); especially severe in wide angle or aperture lenses. This can be prevented by using elements with convoluted aspheric curves. Although this was theoretically proven by René Descartes in 1637,[250] the grinding and polishing of aspheric glass surfaces was extremely difficult and expensive.[251][252]
The first camera lens with an inexpensive mass-produced molded glass aspheric element was the unnamed 12.5mm f/2.8 lens built into the Kodak Disc 4000, 6000 and 8000 (USA) cameras in 1982. It was said to be capable of resolving 250 lines per millimeter. The four element lens was a Triplet with an added rear field-flattener. The Kodak Disc cameras contained very sophisticated engineering. They also had a lithium battery, microchip electronics, programmed autoexposure and motorized film wind for US$68 to US$143 list. It was the Disc film format that was unable to record 250 lpm.[253]
Kodak began using mass-produced plastic aspheres in viewfinder optics in 1957, and the Kodak Ektramax (USA) Pocket Instamatic 110 cartridge film camera had a built-in Kodak Ektar 25mm f/1.9 lens (also a four element Triplet) with a molded plastic aspheric element in 1978 for US$87.50 list.[254] Plastic is easy to mold into complex shapes that can include an integral mounting flange.[255] However, glass is superior to plastic for lens making in many respects – its refractive index, temperature stability, mechanical strength and variety is higher.[256]
The new freedom allowed by inexpensive precision molded plastic or glass aspheric elements is one of the greatest influences on lens design in the last quarter century, producing a breathtaking variety of lenses.
The hunger for one lens able to do everything, or at least as much as possible, is probably the other great influence on lens design in the last quarter century. The Kino Precision Kiron 28-210mm f/4-5.6 (Japan) of 1985 was the first very large ratio focal length "superzoom" lens for still cameras (most 35mm SLRs). The fourteen element/eleven group Kiron was first 35mm SLR zoom lens to extend from standard wide angle to long telephoto, able to replace 28, 35, 50, 85, 105, 135 and 200mm prime lenses, albeit restricted to a small variable maximum aperture to keep size, weight and cost within reason (129×75 mm, 840 g, 72mm filter, US$359 list).[257][258][259]
Early 35mm SLR zooms focal length ratios rarely exceeded 3 to 1, because of unacceptable image quality issues. However, zoom versatility, despite increasing optical complexity and stricter manufacturing tolerances, continued to increase. Despite their many image quality compromises, convenient superzooms (sometimes with ratios over 10 to 1 and four or five independently moving cells) became common on amateur level 35mm SLRs by the late 1990s. They remain a standard lens on today's amateur digital SLRs,[260] with the Tamron AF18-270mm f/3.5-6.3 Di II VC LD Aspherical (IF) MACRO attaining 15× in 2008.[261] Superzooms also sell by the millions on digital point-and-shoots.[262]
The desire for an all-in-one lens is hardly a new phenomenon. "Convertible" lenses, still used by large format film photographers (insofar as large format photography is used), consisting of two cells that could be used individually or screwed together, giving three-lenses-in-one,[263] date back to at least the Zeiss Convertible Protar (Germany) of 1894.[264]
Convenience of a different sort was the major feature of the Tokina SZ-X 70-210mm f/4-5.6 SD (Japan) of 1985. It was the first ultra-compact zoom (85×66 mm, 445 g, 52mm filter); half the size of most earlier 70-210 zooms[265] (the third generation Vivitar Series 1 70-210mm f/2.8-4 [1984, USA/Japan] was 139×70 mm, 860 g, 62mm filter).[266] Like the Kiron 28-210mm, the twelve element/eight group/three cell Tokina had a small variable maximum aperture, but added low dispersion glass and a new bidirectional nonlinear zooming action, to bring size and weight down to an absolute minimum.[267]
Small aperture 35mm format lenses were made practical by the availability of snapshot quality, high sensitivity ISO 400 color films in the 1980s (and ISO 800 in the 1990s), as well as cameras with built-in flash units. During the 1990s, point-and-shoot cameras with compact small aperture zooms were the dominant camera type. Compact variable aperture zoom (some superzoom, some not) lenses remain a standard lens on today's digital point-and-shoot cameras.
At about this time the image quality of zooms equalled that of primes.[268]
Note, many of today's superzooms are not "parfocal"; that is, not true zooms. They are "varifocal" – the focus point shifts with the focal length – but are easier to design and manufacture. The focus shift usually goes unnoticed as they are mounted on autofocus cameras that will automatically refocus.[269]
Since autofocus is primarily an electromechanical feature of the camera, not an optical one of the lens, it did not greatly influence lens design. The only changes wrought by AF were mechanical adaptations: the popularity of "internal focusing", the switch back to "two touch" zooming and the inclusion of AF motors or driveshafts, gearing and electronic control microchips inside the lens shell.[270]
However, for the record: the first autofocus lens for a still camera was the Konishiroku Konica Hexanon 38mm f/2.8[271] built into the Konica C35 AF (1977, Japan) 35mm point-and-shoot; the first autofocus lens for an SLR camera was the unnamed 116mm f/8[272] built into the Polaroid SX-70 Sonar (1978, USA) instant film SLR; the first interchangeable autofocus SLR lens was the Ricoh AF Rikenon 50mm f/2 (1980, Japan, for any Pentax K mount 35mm SLR),[273] which had a self-contained passive electronic rangefinder AF system in a bulky top-mounted box; the first dedicated autofocus lens mount was the five electrical contact pin Pentax K-F mount on the Asahi Optical Pentax ME F (1981, Japan) 35mm SLR camera with a TTL contrast detection AF system for its unique SMC Pentax AF 35mm-70mm f/2.8 Zoom Lens;[274] the first built-in TTL autofocus SLR lens was the Opcon/Komine/Honeywell Vivitar Series 1 200mm f/3.5 (1984, USA/Japan, for most 35mm SLRs),[275] which had a self-contained TTL passive phase detection AF system in a underslung box and the first complete autofocus lens line was the twelve Minolta AF A mount lenses (24mm f/2.8, 28mm f/2.8, 50mm f/1.4, 50mm f/1.7, 50mm f/2.8 Macro, 135mm f/2.8, 300mm f/2.8 APO, 28-85mm f/3.5-4.5, 28-135mm f/4-4.5, 35-70mm f/4, 35-105mm f/3.5-4.5 and 70-210mm f/4)[276] introduced with the Minolta Maxxum 7000 (1985, Japan) 35mm SLR and its TTL passive phase detection AF system.
Even with a high optical quality lens, it is still easy to produce deficient images. Exposure error was solved by electronic autoexposure in the 1970s and focusing errors were alleviated by autofocus in the 1980s.
In 1994, the unnamed 38-105mm f/4-7.8 lens built into the Nikon Zoom-Touch 105 VR (Japan) 35mm point-and-shoot camera was the first consumer lens with built-in image stabilization.[277] Its Vibration Reduction system could detect and counteract handheld camera/lens unsteadiness, allowing sharp photographs of static subjects at shutter speeds much slower than normally possible without a tripod. Although image stabilization is an electromechanical breakthrough, not optical, it was the biggest new feature of the 1990s.
The Canon EF 75-300mm f/4-5.6 IS USM (Japan)[278] of 1995 was the first interchangeable lens with built-in image stabilization (called Image Stabilizer; for Canon EOS 35mm SLRs). Image stabilized lenses were initially very expensive and used mostly by professional photographers.[279] Stabilization surged into the amateur digital SLR market in 2006.[280][281][282][283][284] However, the Konica Minolta Maxxum 7D (Japan) digital SLR introduced the first camera body-based stabilization system in 2004[285] and there is now a great engineering and marketing battle over whether the system should be lens-based (counter-shift lens elements) or camera-based (counter-shift image sensor).[286][287]
With computer-aided design, aspherics, multicoating, very high refraction/low dispersion glass and unlimited budget, it is now possible to control the monochromatic aberrations to almost any arbitrary limit – subject to the absolute diffraction limit demanded by the laws of physics. However, chromatic aberrations remain resistant to these solutions in many practical applications.
In 2001, the Canon EF 400mm f/4 DO IS USM (Japan) was first diffractive optics lens for consumer cameras (for Canon EOS 35mm SLRs).[288] Normally photographic cameras use refractive lenses (with the occasional reflective mirror) as their image forming optical system. The 400 DO lens had a multilayer diffractive element containing concentric circular diffraction gratings to take advantage of diffraction's opposite color dispersion (compared to refraction) to correct chromatic and spherical aberrations with less low dispersion glass, fewer aspheric surfaces and less bulk.[289][290][291]
As of 2010, there have been only two expensive professional level diffractive optics lenses for consumer cameras,[292] but if the technology proves useful, prices will drop and its popularity will rise.
At first glance, digital photography would seem not to affect lenses, since it is a camera technology for the capture and storage, but not the creation, of images. However, electronic image processing provides an opportunity to improve lens images far beyond a simple contrast boosting Unsharp Mask.
In 2004, the Kodak (Sigma) DSC Pro SLR/c (USA/Japan) digital SLR was loaded with optical performance profiles on 110 lenses so that the on-board computer could correct the lateral chromatic aberration of those lenses, on-the-fly as part of the capture process.[293] Also in 2004, DO Labs DoX Optics Pro (France) computer software modules were introduced, loaded with information on specific cameras and lenses, that could correct distortion, vignetting, blur and lateral chromatic aberration of images in post-production.[294]
Lenses have already appeared whose image quality would have been marginal or unacceptable in the film era, but are acceptable in the digital era because the cameras for which they are intended automatically correct their defects. For example, onboard automatic software image correction is a standard feature of 2008's Micro Four Thirds digital format. Images from the 2009 Panasonic 14-140mm f/4-5.8 G VARIO ASPH. MEGA O.I.S. and the 2010 Olympus M. Zuiko Digital 14-150mm f/4-5.6 ED lenses (both Japan) have their severe barrel distortion at the wide angle settings automatically reduced by a Panasonic LUMIX DMC-GH1 and Olympus Pen E-P2, respectively. The Panasonic 14-140mm lens also has its chromatic aberration corrected. (Olympus has not yet implemented chromatic aberration correction.)[295][296]
Some notable photographic optical lens designs are:
Some lens manufacturers (2009):