Daguerreotype: What It Is and How It Worked
A direct positive on polished silver, no negative and no copies: how it was sensitized with iodine, developed with mercury and why it took minutes.
A daguerreotype is a direct positive formed on a silver-plated copper sheet polished to a mirror finish: one plate, one image, no negative and no copies. The plate acted as film, iodine vapor sensitized it, a box of heated mercury developed it and a salt solution fixed it. With that process, presented in France in 1839, photography became a trade.
What is a daguerreotype?
The image forms on the same plate that was inside the camera. There is no intermediate negative, and since the plate is opaque it cannot be contact-printed either: in its day, duplicating a daguerreotype meant photographing it again with another camera (a worse copy) or turning it into a lithograph or engraving.
Two oddities show up when you hold one in your hand:
It comes out laterally reversed, as a mirror image, unless the camera had a mirror or a prism fitted. Daguerre's own Boulevard du Temple is reversed and you can tell from the signs; that's why text that reads the right way round can be proof that a plate is a copy of an original.
It is seen by scattering. The Canadian Conservation Institute explains it this way: the shadows are bare silver and the highlights are a microscopic texture that scatters light. You have to find the angle, light it with a single source at about 45°, and as you rotate the plate the same image goes from positive to negative.
Sizes went by fractions of a whole plate, which measured 6½ × 8½ inches (about 16.5 × 21.6 cm); the sixth plate, the usual one in portrait cases, about 7 × 8.3 cm. The process fits between Niépce's heliography and the calotype within the history of photography in 12 milestones.
How does it differ from the calotype and from a digital sensor?
| Daguerreotype (1839) | Calotype (1841) | Digital sensor | |
|---|---|---|---|
| Support | Silver-plated copper, mirror-polished | Paper coated with silver iodide | Silicon, and a file on a card |
| Positive or negative | Direct positive, no negative | Translucent negative on paper | Neither: data |
| Copies possible | None directly; rephotograph the plate or turn it into an engraving | Contact copies on salted paper, with no practical limit | Unlimited identical copies |
| Exposure time | From 3-15 min (Library of Congress) to 20-30 min (Britannica); under a minute by 1842 | A minute or two in full sun | From 1/8000 s to hours |
| Permanence | Many are still in good condition almost two hundred years later, if nobody opened the seal | Uneven; many faded early | Depends on the backup, not on the medium |
The copies row decided the contest. Talbot discovered on September 23, 1840 that an exposure of seconds left a latent image that could be developed with gallic acid, and he patented the calotype in February 1841: his negative gave copies, even though the paper fiber showed through in each one and the image came out softer than a silver plate.
The permanence row is the one that has changed in nature. A daguerreotype's lives in the object and is lost with it; a file's lives in the backup policy, which is what the 3-2-1 rule organizes. The remaining differences between a chemical medium and a sensor are in analog vs digital photography.
How was a plate made, step by step?
Step: 1. Polishing; In the 1840 workshop: Silver burnished to a mirror shortly before use; Modern equivalent: Clean sensor
Step: 2. Sensitizing; In the 1840 workshop: Iodine vapor until the plate turns yellow-pink; Modern equivalent: Loading film, choosing ISO
Step: 3. Accelerator; In the 1840 workshop: Bromine or chlorine vapor, with a final iodine refuming; Modern equivalent: Raising the ISO
Step: 4. Exposure; In the 1840 workshop: The lens is uncapped by hand; there is no shutter; Modern equivalent: Shutter speed
Step: 5. Development; In the 1840 workshop: Hot mercury vapor, 2 to 5 min, face down over the bath; Modern equivalent: Chemical development
Step: 6. Fixing; In the 1840 workshop: Hot table salt, then sodium thiosulfate; Modern equivalent: Fixer
Step: 7. Gilding and sealing; In the 1840 workshop: Gold chloride, glass, mat and gummed paper; Modern equivalent: Varnish and mounting
Color was the indicator: the yellow-pink of silver iodide signaled that the plate was ready. The latent image stayed invisible until the mercury condensed where light had arrived and formed a silver-mercury amalgam that scatters light. Those are the highlights.
Gilding was not cosmetic. Hippolyte Fizeau introduced it around 1840-1841 by pouring gold chloride over the plate and briefly heating it: it physically reinforces the particles and raises contrast. Without gilding, the image is "as delicate as the dust on a butterfly's wing" and a finger wipes it away.
An old portrait in its case: the headrest was imposed by the clock
How long a 1839 exposure lasted is the most disputed fact of the process, and the sources do not overlap: the Library of Congress gives three to fifteen minutes and calls it "almost impracticable for portraiture"; Britannica gives twenty to thirty. The controlled re-trials by daguerreotypist and researcher Mike Robinson measure 12 minutes with a plate sensitized with iodine only. Between a few minutes and half an hour, depending on whom you ask.
Two things brought it down to seconds in a couple of years:
The lens. The Chevalier landscape lens worked at around f/14; the Petzval portrait lens, calculated in Vienna in 1840 and in general use from 1841, at around f/3.6. The math is (14 / 3.6)² = 15.1: about fifteen times more light, a little under 4 stops. It was paid for in weak edges, which don't matter for a face and do for a landscape.
The accelerators. Where iodine alone asked for 12 minutes, iodine chloride asked for two and a half minutes, and bromine came out about 60 times faster than iodine alone. John Frederick Goddard published the bromine finding on December 12, 1840.
Studio figures confirm it. Voigtländer's August 1841 sheet, with iodine-only plates and the model outdoors, called for three and a half minutes under a winter overcast sky, a minute and a half to two minutes in the shade on a sunny day and 40-45 seconds in full sun. Antoine Claudet, already using accelerators in London, wrote in September 1841 that in June 10-20 seconds were enough for him and in September he needed 60-90.
Forty seconds of stillness is still a lot. Headrests came from painted portraiture and were standard studio equipment, but at those times they stopped being an accessory: they held the skull so the face wouldn't come out smeared. Studios were glazed on rooftops like greenhouses, and sometimes filtered in blue so the sitter could stand the light on the face, because the plate was almost only sensitive to blue and removing the rest of the spectrum barely lengthened the shot. A Swedish amateur nearly cost his sitter an eye by keeping them looking almost straight at the sun for five minutes.
What ISO did a daguerreotype plate have?
There is no standardized measurement. Robinson's thesis, the most rigorous modern work, gives no ISO or ASA figure and expresses everything in relative proportions; the "0.001-0.005 ISO" that circulates online has no method or measurement behind it, and the sensitometric study that would hold the real figure, done at RIT in 1971, is not online. ISO doesn't fit either: it is defined for gelatin emulsions, and a plate that is color-blind and sensitive only to blue and ultraviolet doesn't see what a light meter measures.
What does fit is a reconstruction with the sunny 16 rule, checkable with a calculator:
In full sun, ISO 100 and f/16 call for 1/100 s.
The 1839 process worked at f/14, practically the same, and called for between 3 and 15 minutes according to the Library of Congress.
Going from 1/100 s to 12 minutes (720 s) is multiplying by 72,000, about 16 stops.
Sixteen stops below ISO 100 puts the plate on the order of a thousandth of ISO 1: roughly between ISO 0.006 at 3 minutes and ISO 0.001 at 15.
With bromine-accelerated plates, about 60 times faster, the reconstruction rises to around ISO 0.1, still about ten stops below ISO 100. It is a reconstruction, not a measurement: people who make plates today sum it up by saying there is no easy way to determine exposure and prior experience rules. To see what sixteen stops look like on controls you can actually touch, the exposure triangle simulator walks through them one by one.
Can you make a daguerreotype today?
Without mercury, yes. With mercury, you shouldn't. Development consisted of suspending the plate face down over a heated mercury bath, and the temperature the sources give ranges from about 60 °C to about 90 °C depending on the manual and the bath: there is no single defensible number. Mercury's toxicity was known in the nineteenth century and precautions were still rarely taken; there are documented reports of photographers with mad hatter syndrome, the same mercury poisoning that affected hatters.
Mercury wasn't even the worst thing in the workshop. Elemental bromine is a corrosive liquid that at room temperature already gives off a red vapor capable of causing severe burns and pulmonary edema; mercury is a cumulative neurotoxin, but it is controlled with ventilation and containment. Today's daguerreotypists have abandoned mercury vapor and work with the less toxic Becquerel process, in exchange for plates about 10 times slower.
This is not cyanotype. A cyanotype is a Saturday plan with a brush and sunshine; a classic daguerreotype is iodine, bromine and hot mercury, and it is not a process to reproduce at home.
What do I do if I have a daguerreotype at home?
Don't open the case, don't take out the plate, don't touch the surface, use clean lint-free gloves and consult a conservator. All the guides agree, and the reason they give is always the same: abrasion, tarnish and fragility of the glass. No conservation institution presents the plate as a toxic risk to whoever handles it, and none declares it harmless; it is known that mercury remains chemically present in the image particles, and there the verifiable ends.
Tarnish is what kills the image, and the mechanism explains the severity: the image is scattered light, so a layer of silver oxides, chlorides or sulfides doesn't darken it, it erases it. It appears when the case is opened or the seal is broken, starts light brown at the edge and progresses to an opaque blue-black stain. Cleaning it with chemical solutions was common practice and today it shows up as veiling, green fluorescence and pitting that reveals the copper. On displaying it, the most recent and cautious guide sets a maximum of three or four months and advises against intense light.
Frequently asked questions
How long did a daguerreotype exposure last?
It depends on the year, and for 1839 nobody gives the same figure: the Library of Congress says three to fifteen minutes and Britannica twenty to thirty, ranges that don't even overlap. With the Petzval lens and the bromine or chlorine accelerators it came down to 40-45 seconds in full sun in 1841, under a minute by 1842 and a few seconds in the studio in the 1850s.
Why does a daguerreotype look like a mirror?
Because it is one: the plate is silver-plated copper polished to a mirror finish, and the image is formed by microscopic particles that scatter light on that surface. The shadows are bare silver reflecting whatever is in front of it, so you have to light it with a single source at about 45° and rotate it until you find the angle at which it reads as a positive.
Can you make copies of a daguerreotype?
Not by contact: the plate is opaque and there is no negative, so each daguerreotype is a unique piece. In its day it could only be duplicated by rephotographing it with another camera, with a loss of quality, or by turning it into a lithograph or engraving.
Who invented the daguerreotype?
Louis-Jacques-Mandé Daguerre, a set painter and owner of the Paris Diorama, building on the work of Joseph Nicéphore Niépce, with whom he partnered in 1829 and who died in 1833. François Arago showed the results to the Académie des Sciences on January 7, 1839 without revealing the method, and the full details were published on August 19 of that year in exchange for lifetime pensions of 6,000 francs a year for Daguerre and 4,000 for Niépce's son.
What the daguerreotype makes clear about your settings
Exposure has always been the same calculation. Those operators had no light meter or shutter, but they went from f/14 to f/3.6 to gain almost four stops and added bromine to gain another six, the same arithmetic you do when you open the aperture or raise the ISO. For them a stop cost minutes in a chair with a headrest; for you it costs a click.
And honest figures carry their margin. When the range runs from three to thirty minutes depending on the source, the useful thing is to write down the range, not to pick the number that looks best.
The 44 cards from Enfogram come with the aperture, shutter speed and ISO steps printed by scene, in A4 and A5, so you do the math looking at a card and not at the screen.