How the World Became an Image

The camera obscura (Latin for "dark room") is a natural optical phenomenon. Light passing through a small opening travels in straight lines and forms an upside-down image on the opposite wall. A lens makes the image brighter and sharper.

The principle was described in the Chinese text associated with Mozi (5th–4th century BCE) and in the problems attributed to Aristotle, who noticed crescent-shaped images of a partially eclipsed sun cast through the gaps between leaves. The eleventh-century scholar Ibn al-Haytham (Alhazen), working in Cairo, investigated it systematically and used it to explain how light travels and how vision works.

Astronomers adopted the dark room as a safe way to watch the sun. In 1545 the Flemish mathematician Gemma Frisius published the first known illustration of one, showing the solar eclipse he had observed at Louvain on 24 January 1544.

From there the instrument grew. Gerolamo Cardano described fitting a glass disc, probably a lens, in 1550; Giambattista della Porta popularised the darkened room as a spectacle and drawing aid in Magia Naturalis (1558); Johannes Kepler gave it its Latin name and built a portable tent version. By the eighteenth century, box-sized camera obscuras with mirrors and ground-glass screens were common studio tools. The device could show an image; it could not keep one.

The camera lucida

Patented by William Hyde Wollaston in 1806, the camera lucida ("light room") used a prism to make a subject appear superimposed on the paper beneath it. It was a drawing aid, not a camera: the artist traced what they saw. In October 1833, sketching on the shores of Lake Como, William Henry Fox Talbot found the results with one dispiriting. He later recalled thinking

"how charming it would be if it were possible to cause these natural images to imprint themselves durably, and remain fixed upon the paper!"Talbot, The Pencil of Nature, 1844

The frustration helped turn an old optical problem into a chemical one.

Some silver compounds are photosensitive: light changes their chemical structure and produces dark metallic silver. In 1725–27 the German professor Johann Heinrich Schulze showed that a mixture of chalk and silver salts darkened where sunlight struck it. He cut letters out of paper, wrapped them around a bottle of the mixture and watched words appear. He proved the cause was light, not heat.

In 1777 the Swedish chemist Carl Wilhelm Scheele found that darkened silver chloride resisted ammonia, which dissolved the unaltered compound. It pointed to a way of stabilising an image by removing the sensitive material that light had not touched, though no one used it for decades.

Around 1800 Thomas Wedgwood, son of the potter, working with Humphry Davy, made shadow pictures of leaves, insect wings and painted glass on paper and white leather coated with silver nitrate. Davy published their results in 1802. The images could only be looked at briefly by candlelight; daylight darkened the whole sheet. They had made photographs, but could not fix them. (A leaf photogram once attributed to Wedgwood surfaced at auction in 2008; most scholars now think it is by Talbot.)

Joseph Nicéphore Niépce (1765–1833) was a gentleman inventor from Chalon-sur-Saône in Burgundy. With his brother Claude he had already built the Pyréolophore, one of the first internal combustion engines. In 1813 he caught the French craze for lithography, the new method of printing from drawings made in greasy ink on limestone. Niépce could not draw; his son Isidore made the designs. He began looking for a way to let light do the drawing instead.

His first camera experiments, in 1816, used paper coated with silver chloride. He made small negatives of the view from his workroom, which he called rétines, and sent descriptions to Claude. But he could not fix them, and he disliked that the tones were reversed.

He then turned to a material printers already knew: bitumen of Judea, a natural asphalt used as an acid-resisting ground for etching. Bitumen hardens and becomes insoluble where light strikes it. Niépce coated plates of glass, stone, zinc, copper and pewter, laid an engraving made translucent with oil over them, and left them in the sun. Where the black lines of the print had shielded the bitumen, it stayed soft and could be washed away with oil of lavender and white petroleum, exposing bare metal. Acid then bit those lines into the plate, which could be inked and printed. He called the process heliography, sun-writing.

A first heliographic copy of a portrait of Pope Pius VII, made in 1822, was destroyed when he tried to print from it. The oldest that survive date from 1825: a print of a seventeenth-century Dutch etching of a boy leading a horse, and one of a woman at a spinning wheel. In 1826 he copied Isaac Briot's engraved portrait of Cardinal Georges d'Amboise; the plates were etched more deeply by the Paris engraver Augustin François Lemaître, and prints were pulled from them in 1827. They are the ancestors of photogravure, photolithography and every photomechanical printing method that followed.

Niépce also put bitumen plates in a camera obscura, aiming at the courtyard below his upstairs workroom at Le Gras, his estate at Saint-Loup-de-Varennes. One pewter plate made in 1826 or 1827 survives. It shows a pigeon-house on the left, a pear tree, the sloping roof of a barn and another wing of the house on the right. The bitumen is so slow that the exposure took at least eight hours and may have lasted days; sunlight appears to strike both sides of the courtyard because the sun crossed the sky while the plate was exposed.

In 1827 Niépce travelled to Kew to visit his ailing brother and tried to present heliography to the Royal Society. Because he would not reveal the full method, the Society declined. He left the plate with the botanical artist Francis Bauer. It passed through private hands, vanished after 1905, and was tracked down in 1952 by the historian Helmut Gernsheim, who commissioned the enhanced reproduction that is usually printed. Since 1963 it has been at the Harry Ransom Center in Austin, Texas.

The original is hard to read: a mirror-like metal plate on which the image appears only at a raking angle. It is also extraordinary: a metal surface physically altered by light reflected from the world itself.

Louis-Jacques-Mandé Daguerre (1787–1851) trained under the panorama painter Pierre Prévost and made his name designing stage effects for the Ambigu-Comique and the Opéra, where his moonlight and sunrise effects drew reviews of their own. In 1822, with the architectural painter Charles-Marie Bouton, he opened the Diorama in a purpose-built hall on the rue Sanson, near the Château-d'Eau in Paris (the site of today's Place de la République).

The audience, roughly 350 people, sat in a darkened, circular auditorium that could be turned on a pivot to face one of two or three stages in turn. Each stage was seen down a long, dark tunnel that hid its edges, so there was nothing to judge its size against. At the far end hung an enormous painting, around 22 by 14 metres, on semi-transparent linen or calico.

The trick was in the lighting. Skylights and windows above and behind the picture were fitted with blinds, shutters and coloured screens worked by stagehands. Light from the front showed the image painted on the front of the canvas. As it dimmed and light was let in from behind, a second image, painted on the reverse, shone through: day turned to dusk, empty church naves filled with worshippers at midnight mass, snow fell, an Alpine valley was buried by a landslide. A show lasted about fifteen minutes per picture.

London, fire and Bry-sur-Marne

A second Diorama opened in Regent's Park, London, in September 1823, in a building designed by Augustus Charles Pugin and James Morgan behind John Nash's Park Square East terrace. It showed Daguerre and Bouton's pictures, among them Canterbury Cathedral, the Valley of Sarnen and Holyrood Chapel by moonlight. The façade still stands. Imitations followed in Berlin, Edinburgh, Liverpool, Manchester and Dublin.

Daguerre lived and worked in the Paris Diorama building, and it was there that he carried on the experiments that became the daguerreotype. On 8 March 1839, two months after his invention was announced, the Diorama burned down, taking his paintings with it.

The only Daguerre diorama to survive is a smaller one he painted in 1842 for the parish church at Bry-sur-Marne, where he had retired: a trompe-l'œil Gothic choir behind the altar, restored in the 2010s and still lit by daylight.

Why the Diorama matters to photography

The Diorama was a machine for controlling light falling on a picture: dimming it, colouring it, changing its direction to reveal one image and hide another. Daguerre used a camera obscura to make preparatory studies for his huge canvases, and he wanted an automatic way of recording what it showed.

His famous 1839 manual even put the two inventions side by side: Historique et description des procédés du daguerréotype et du diorama. To contemporaries, the daguerreotype looked like the work of a man who had spent his career manufacturing realism, which is exactly what it was.

The daguerreotype used a highly polished, silver-plated copper plate. Iodine vapour made the surface light-sensitive. After exposure, an invisible latent image was brought out over heated mercury vapour, then fixed (at first with hot salt water, soon with Herschel's hypo) and washed. Daguerre's key discovery, around 1835, was that development could reveal an image after a short exposure, instead of waiting for light to do all the work.

The result was a unique object: depending on the angle, it looks positive or negative, like a mirror with a memory. It was also laterally reversed, as a mirror is.

Daguerre's earliest reliably dated plate, from 1837, is a still life of plaster casts in his studio. A portrait inscribed "M. Huet 1837", if genuine, would be the oldest photograph of a person. The best known early plate, of the Boulevard du Temple, was made in spring 1838 from a window high above the street. The exposure was several minutes, so moving traffic left no trace. The exception is a man standing still to have his boots polished, the first people known to appear in a photograph.

William Henry Fox Talbot (1800–1877) began experimenting in 1834, soon after his Como frustration. He soaked writing paper in weak salt solution, brushed it with silver nitrate, and found that strong salt solution (later potassium iodide) slowed further darkening. Laying leaves, lace and feathers on the paper in sunlight gave crisp white silhouettes on a lilac-brown ground. He called these photogenic drawings.

He also loaded sensitive paper into tiny wooden cameras that his wife Constance called "mousetraps", left around the grounds of Lacock Abbey in Wiltshire. In August 1835 one of them recorded the oriel window of the abbey's south gallery: a negative about an inch across, the oldest camera negative to survive. Talbot noted that the 200 or so panes could be counted with a magnifying glass.

Then he put the work aside. News of Daguerre's announcement in January 1839 sent him rushing: on 25 January he showed photogenic drawings at the Royal Institution, and on 31 January read Some Account of the Art of Photogenic Drawing to the Royal Society. Neither party yet knew how different the two processes were.

The calotype and the latent image

In September 1840 Talbot found what Daguerre had found: a brief exposure leaves an invisible latent image that can be developed chemically, here with gallic acid and silver nitrate. Exposures fell from many minutes to seconds. He called the improved process the calotype (from Greek kalos, beautiful) and patented it in February 1841. Waxed paper negatives were contact-printed onto salted paper to make positives, as many as were needed.

Calotypes were softer than daguerreotypes; the paper fibres blurred fine detail. But they could be multiplied, pasted into books and mailed. Talbot set up a printing works at Reading with his former valet Nicolaas Henneman, and between 1844 and 1846 issued The Pencil of Nature in six parts with 24 tipped-in photographs: buildings, books, a haystack, a ladder, an open stable door, shelves of china. It was the first commercially published book illustrated with photographs, and an argument for everything the medium could do, from recording and copying to making pictures for their own sake.

Scotland, outside the patent

Talbot's patents were strictly enforced in England and Wales, which slowed the calotype there. They did not extend to Scotland. In Edinburgh, Talbot's friend Sir David Brewster encouraged its use, and from 1843 the painter David Octavius Hill and the young chemist Robert Adamson made some 3,000 calotypes at Rock House on Calton Hill. They began with reference portraits for Hill's vast painting of the founding of the Free Church of Scotland, with 474 figures, and went on to the fisherfolk of Newhaven, among the first sustained documentary series in photography.

Hércules Florence, Brazil, 1833

In the village of São Carlos (now Campinas), the French-born painter Hércules Florence wanted a way to print labels and diplomas in a country with almost no presses. In 1833 he made contact copies on silver-salt paper, fixed them with ammonia, and in his notes called the process photographie, years before the word was used in Europe. His work stayed unknown until the historian Boris Kossoy examined his manuscripts in the 1970s.

Hippolyte Bayard, Paris, 1839

Hippolyte Bayard, a clerk at the Ministry of Finance, invented a direct positive paper process in early 1839 and on 24 June showed some thirty prints at a charity sale in Paris, arguably the first public exhibition of photographs. Arago, Daguerre's champion, persuaded him to delay publishing. Bayard got a small grant; Daguerre got the glory. In October 1840 Bayard staged a protest: a photograph of himself slumped as a drowned man, with a bitter note on the back explaining that the inventor had killed himself after being ignored by the government. It is often called the first staged photograph, and the first photographic hoax.

Herschel's blue prints and Anna Atkins

Herschel kept experimenting with every light-sensitive substance he could find. In 1842 he published the cyanotype, which uses iron salts rather than silver and produces a durable Prussian blue. Architects later used it for their "blueprints". His friend Anna Atkins, a botanist and illustrator, used it to make contact images of seaweeds. Beginning in October 1843 she issued Photographs of British Algae: Cyanotype Impressions in parts to a small circle of friends. It is generally recognised as the first book illustrated with photographs, preceding Talbot's, and Atkins as the first woman to make photographs.

Firsts, 1838–1840

First people in a photograph: the Boulevard du Temple, 1838. First photographic exhibition: Bayard, June 1839. First surviving photograph on glass: Herschel, September 1839. Earliest American portrait: Cornelius, autumn 1839. First portrait studio: Alexander Wolcott and John Johnson, New York, 1840. First staged photograph: Bayard, 1840. First photographically illustrated book: Atkins, 1843.

"First" is slippery in photographic history: claims depend on what survives, what counts as a photograph, and who published. The dates above are the ones most historians accept.

Glass, albumen and the portrait boom

In 1850 Louis-Désiré Blanquart-Evrard introduced albumen paper, coated with egg white, which gave sharper, glossier prints and dominated for forty years. In 1851 Frederick Scott Archer published the wet collodion process: glass negatives with the sharpness of a daguerreotype and the reproducibility of a calotype, but the plate had to be coated, exposed and developed before the plate dried. Photographers hauled darkroom tents across battlefields and mountains. Collodion also gave cheap one-off formats: the ambrotype on glass (1854) and the tintype on iron (1856).

André-Adolphe-Eugène Disdéri's carte de visite (1854) put eight small portraits on one plate and launched a collecting craze. Stereoscopic views, popularised by Sir David Brewster's lens stereoscope, brought three-dimensional images of distant places into parlours by the million.

Colour and motion

In May 1861 James Clerk Maxwell, with the photographer Thomas Sutton, demonstrated three-colour photography at the Royal Institution: a tartan ribbon photographed through red, green and blue filters and projected through the same filters, one over another. In 1878 Eadweard Muybridge used a row of cameras triggered by threads to show that a galloping horse lifts all four hooves off the ground, and in doing so opened the way to cinema.

Photography becomes an industry

Richard Leach Maddox's gelatin dry plate (1871), made commercially from the late 1870s, could be bought ready-made and developed later; exposures fell to fractions of a second. Halftone screens, first used in a newspaper in 1880, let photographs be printed alongside type. George Eastman's roll film and the Kodak camera (1888) completed the shift: the buyer took a hundred pictures and posted the whole camera back to Rochester for processing. "You press the button, we do the rest." The one-dollar Brownie followed in 1900.

Small cameras, colour film, mass culture

Gabriel Lippmann's interference colour process won a Nobel Prize in 1908; the Lumière brothers' Autochrome, marketed from 1907, made colour practical for enthusiasts, using dyed potato-starch grains as filters. The Leica (1925) popularised 35mm film; Kodachrome (1935) made colour transparencies routine; Edwin Land's Polaroid (1948) delivered a print in a minute. Photojournalism, family snapshots, fashion and advertising each developed their own languages. Photography stopped being a record of exceptional things and became part of everyday life.

The digital turn

In 1957 Russell Kirsch and colleagues at the US National Bureau of Standards built a drum scanner for the SEAC computer and digitised a 5 cm snapshot of his three-month-old son, Walden, as a grid of 176 × 176 pixels. The charge-coupled device (CCD), invented at Bell Labs in 1969, became the electronic sensor; in 1975 Steven Sasson at Kodak built a self-contained digital camera that took 23 seconds to record a 0.01-megapixel image to cassette tape. Bryce Bayer's colour filter mosaic (1976), the JPEG standard (1992) and the first camera phones (2000) followed.

By the 2010s more photographs were made each year than in the whole of the film era. Today computational photography merges many exposures into one, and generative systems produce photographic-looking images without any light from the scene at all. That unsettles the old assumption that a photograph is a trace of something that stood before a lens.