Every old photograph format, and how to tell which one is in your hand.

A photograph is a physical process before it is a picture.

Every photograph up to the 1990s was made by a specific chemical process on a specific support — metal, glass, or paper — and each process was commercially dominant for a known span of years. That makes the object itself a date stamp. A cased mirror-like plate cannot be later than about 1860. A card mount the size of a postcard with a photographer's name printed along the bottom is almost certainly 1866 to 1914. A glossy black-and-white print with a white border is twentieth century, probably after 1910.

People misread this constantly. They date a photograph from the clothes, which works only if the clothes are the sitter's own and the print is contemporary with the sitting, and from the faces, which works not at all. Format, mount and surface are contemporaneous with manufacture. Clothing can be a hand-me-down; negatives were reprinted for decades.

The one caveat worth stating plainly: format dates the object, not the sitting. A photograph can be a later print from an earlier negative — a 1900s gelatin silver print from an 1870s glass plate is common. Format tells you the earliest possible date with confidence, and the likely window with reasonable confidence.

If you would rather answer the question with four observations than with this page, the date estimator does the interval arithmetic for you.

The nine formats, with what to look for and what they prove.

Two more you will meet and should not handle casually. Glass negatives (1850s–1930s) are fragile, and a broken plate can still be printed — do not throw the pieces away. Nitrate negatives (roughly 1889–1951) are chemically unstable and flammable; they are the one item in a family collection that should be stored apart from everything else, digitised, and then managed according to local disposal rules rather than kept in a cupboard.

Three supports, three entirely different sets of risks.

Grouping by support rather than by name is the fastest way to remember what to do.

Metal

Daguerreotypes and tintypes. The daguerreotype is a silvered copper plate so soft that a fingertip leaves a permanent mark, sealed under glass in a case with a gilt mat; the case is part of the object and part of its value. Never open one, never clean one, never try to rehouse one. A tintype is a lacquered iron sheet with a collodion emulsion, robust by comparison and easily bent — bending cracks the image layer off.

Glass

Ambrotypes and glass negatives. Both fail by fracture, and both are damaged by water trapped between the plate and its cover glass. An ambrotype is a negative pretending to be a positive, and its backing is what makes it work; if the backing is flaking, the image goes pale in patches. Keep glass vertical in storage, never stacked flat, and never in a damp room.

Paper

Albumen, gelatin silver, cyanotype, postcards, and every colour print. Paper is the most forgiving support and the most chemically complicated. Albumen's egg-white binder yellows and cracks, and its silver image sits on the surface where a thumb will remove it. Gelatin silver is comparatively stable but silverings, and its gelatin layer swells with humidity and sticks to anything pressed against it. Colour prints made between roughly 1960 and 1985 are the most at-risk material in a typical family collection — their dyes fade in daylight and in the average cupboard, and there is no treatment that brings them back.

The practical rule that follows: store by support, not by subject. Metal and glass want stable, dry, dark, undisturbed. Paper wants cool, dry, sleeved and flat. Colour wants the dark, and wants it sooner than anything else.

The four words that cause most of the confusion.

Sepia

Not a format and not always a process. It is a brown tone — either an intentional sulfide or selenium toning applied to a black-and-white print, or the natural yellowing of an albumen print, or simply the colour a scan takes on. A brown photograph is not automatically old.

Tintype

Contains no tin. It is a collodion image on a japanned iron plate. The name is a nineteenth-century trade word, and it appears in the same period as "ferrotype".

CDV and cabinet card

These name the mount, not the print. Both usually hold an albumen print. That is why the size of the card is the dating evidence and the print technique is the condition problem.

Reprint and copy

A later print from an original negative, or a photograph of a photograph. Both are common in family collections, and both defeat naive dating: a 1950s gelatin silver print of an 1870s sitter is not evidence of a 1950s photograph. Look for the mount, the paper and the corner wear of the print itself.

One more useful distinction: albumen versus gelatin is the biggest single tell between a nineteenth-century and a twentieth-century paper print. Albumen prints are warm, thin, card-mounted and crackle. Gelatin prints are neutral, on a stiffer paper base, and usually have a white border.

What the format changes about value, scanning and restoration.

Format is not a curiosity. It decides three practical things.

Value. A cased daguerreotype of an identifiable person can be worth hundreds; the same image as a later gelatin silver print is worth a few dollars. Rarity of process is most of the difference, and identification is most of the rest — which is why what actually sells is a question about names, not about age.

Scanning. A daguerreotype is a mirror, so a flatbed scan of one is a photograph of the scanner. Cased plates, glass, and anything with a curved tintype surface need a camera, not a scanner, and need the reflections managed. Everything paper and flat scans well. The scanning arithmetic covers the rest.

Restoration. A cracked albumen print and a scratched gelatin print fail differently, and the honest answer differs: one is repaired on the object by a conservator, the other is reconstructed in a file. If the damage is what brought you here, the costs of each route are set out without a sales voice.

If the object is unstable — mould, water, flaking emulsion, a case that has come apart — stop and read the first 48 hours before doing anything else. Handling decisions made in the first day are the ones that cannot be undone.