Analogue craft

Original or dupe? How to tell duplicate transparencies apart

A woman with short hair rests her arms across a raised knee, wearing necklaces and bangles.
Seated portrait in black and white.Photograph by Robert Harper. Robert Harper Photograph Library, Archive.London. From the archive, shown for illustration.

For decades the picture business ran on copies of copies. A dupe can look perfect on a lightbox and still be a generation away from the photograph. Here is how to tell, and why to find out before scanning.

Every transparency in a box is either the film that was in the camera or a photograph of that film, and the catalogue should say which.

An original transparency is the piece of film that was in the camera. A duplicate transparency, universally shortened to dupe, is a copy of it made later on a second piece of reversal film. On a lightbox the two can look alike. Under a loupe, and above all under a scanner, they are different objects: one is a photograph, the other a photograph of a photograph.

Most advice on the subject is written for collectors worried about being sold a copy. Ours is written for the person about to digitise a filing cabinet. Picture libraries ran on dupes for decades, so any archive that passed through that trade will contain both, mixed together and rarely labelled. Sorting them before scanning saves money, protects quality and prevents the catalogue from describing one picture five times. The signs are learnable, and there are eight of them.

What is the difference between an original and a duplicate transparency?

Reversal film produces a positive image on a transparent base, so the film that comes out of the processing tank is the finished picture. There is no negative to print from. That is the format's glory and its weakness: every original is unique.

Making a second copy means photographing or contact-printing the first one onto fresh film. The simplest method is a slide copier, an optical attachment that fits on a camera's lens mount and lets one slide be photographed onto another frame. Professional laboratories used dedicated duplicating equipment and special film stock, and produced dupes by the hundred.

A third category causes most of the confusion. Photographers who knew a picture would be needed in several places simply exposed several frames at the time, often at slightly different exposures. Agencies and studios called these in-camera duplicates, or similars. They are not dupes. Each one is a first-generation original, and each is very slightly different from its neighbours: a changed expression, a shifted highlight, a third of a stop.

Camera original In-camera duplicate or similar Laboratory dupe
Made At the shoot At the shoot, as extra frames Afterwards, from an existing transparency
Generation First First Second or later
Film Camera stock Same camera stock, same roll or batch Duplicating stock
Identical to the others? Not applicable No: small differences in pose, light or exposure Yes in content; differs in contrast and sharpness
Best use now Scan as the master Scan the selected frame; record the rest Scan only if no original survives

Why did picture agencies run on dupes?

Because the alternative was posting irreplaceable film around the world. A library that wanted to show one photograph to a magazine in London, a publisher in Munich and a partner agency in Tokyo in the same week had one original and three customers. Duplicates solved the arithmetic. They also solved the insurance problem: a lost dupe could be remade, and a lost original could not.

The trade was large enough for manufacturers to design film for it. A late example, Kodak's Ektachrome Duplicating Film EDUPE, is described in its data sheet as "a low-contrast color reversal duplicating film designed for making high-quality duplicates from originals on Kodak Ektachrome or Kodachrome Films". It was processed in the same E-6 chemistry as camera Ektachrome, a process first made available in the mid-1970s. And it was sold in sizes that tell their own story about how the business worked: 35 mm in rolls of 100, 400 and 1,000 feet, 46 mm, 61.5 mm and 70 mm rolls, 120, and sheets of 4 x 5 and 8 x 10 inches. Nobody buys a thousand feet of film to copy a holiday slide.

The larger sizes had a purpose. A 35mm picture copied up onto 70 mm or sheet film was easier for an art director to judge on a lightbox and looked more substantial beside medium-format originals. An enlarged dupe was a sales tool.

Why does a duplicate look different from the original?

Every copy loses something. The general term is generation loss: the fall in quality between a copy and the thing it was copied from, which compounds with each further copy.

With transparencies the first casualty is tone. Camera reversal film is built with high contrast, because it has to turn a real scene into a brilliant image for projection. Copy a transparency onto the same film and that contrast is applied a second time: shadows close up, highlights lose their gradation and colours turn harsher. Duplicating film is made low in contrast precisely to cancel this, and Kodak's data sheet lists "faithful reproduction of original tone scale" as the aim, noting that contact printing gives "slightly higher contrast" than optical copying.

A well-made dupe on proper stock is therefore close. It is never identical. A little shadow detail goes, fine texture softens because the image has passed through a second optical system or a second layer of emulsion, and the colour balance depends on how carefully the laboratory filtered its light source.

How can you tell a dupe from an original? Eight signs

No single test is conclusive, and laboratories varied. Work through the list and let the evidence accumulate.

  1. Edge markings. Look at the rebate, the clear or black strip outside the image, with a loupe. Camera film normally carries the maker's name, a film code and frame numbers printed along the edge. Duplicating stock supplied in long bulk rolls often carries different markings or none of the usual frame numbering. A film name that is a duplicating stock settles the matter.
  2. Format against camera. Ask whether a camera could have produced this piece of film. A picture with 35mm proportions sitting on a 4 x 5 inch sheet was not shot that way. Film 70 mm wide with perforations down both edges was a standard duplicating size, although a few cameras also took 70 mm magazines, so check it against the other signs.
  3. Frame edges. An original's image area is shaped by the camera's film gate and its edges are crisp. A dupe sometimes shows two borders: the mask used in copying, and inside it a slightly soft image of the original's frame edge, occasionally with the original's own edge numbers visible as part of the picture.
  4. Contrast. Compare with known originals from the same shoot. Blocked shadows, chalky highlights and a harder look overall suggest a copy, especially one made on camera film instead of duplicating stock.
  5. Sharpness and grain. A dupe carries the grain of the original, photographed, plus its own. At high magnification detail is softer than it should be for the format.
  6. Dust that will not move. Dust or a hair that sat on the original during copying is recorded in the dupe as a sharp dark shape. It is part of the image. It cannot be brushed away, there is no mark on the film surface when you tilt it to the light, and infrared dust removal ignores it, for the reason given in our guide to dust removal in film scans: the scanner's infrared channel sees physical debris, not dye.
  7. Surface relief. Tilt the emulsion side to a light. Kodachrome originals show the image in faint relief on the surface. A duplicate of a Kodachrome made on E-6 stock does not.
  8. Mounts and labels. A mount stamped or written "dupe" or "duplicate" is what it says. Agency stickers, reference numbers and uniform laboratory mounts across a whole set point the same way. A processing laboratory's dated mount, by contrast, usually belongs to an original.

The most reliable method is comparison. Put the candidate beside a frame you know to be original from the same job and the difference is usually obvious within seconds.

Why does the distinction matter before scanning?

For four practical reasons.

Quality. The original holds more detail and a longer tonal range. If both exist, the original is the master and the scan should be made from it. Scanning guidance asks for the full frame, border included; the US National Archives' guidelines recommend capturing "the entire negative including the full border" where the border carries information, which is exactly where the evidence of generation sits.

Cost. A box of forty transparencies may contain eight pictures. Scanning all forty at preservation quality spends five times the budget for no gain. Identify the originals, scan those, and record the rest as items held. The scan resolution calculator will show what each avoided scan saves in storage.

Counting. An honest inventory distinguishes unique photographs from total frames. If dupes are counted as pictures the archive appears larger than it is, and a licensee may be offered the same image under several numbers. After scanning, perceptual hashing will catch what the eye missed; our guide to finding duplicate photos in an archive with hashes explains the method, and the free image hash generator lets you try it on a pair of files.

Dating and provenance. A dupe was made after the original, sometimes years after, and its mount and stock date the copying, not the photograph. Where a dupe bears an agency's label, that is evidence of where the picture was distributed. It is not evidence of who owns the rights in it.

None of this makes dupes worthless. When the original was lost, sold or returned to a client, the dupe is the picture. Some were made with corrections, a tighter crop or a cleaner colour balance, and record how the image was actually circulated. Keep them, house them properly and describe them accurately. Film duplication at scale is now a historical practice: Kodak had discontinued all its Ektachrome products by late 2013, and although camera Ektachrome returned in 2018, the working infrastructure of the dupe trade did not.

What does the Robert Harper archive's inventory show?

The archive's published figures draw the same line this article does. The inventory counts 95,487 unique photographs and 177,143 frames including duplicates and variants. On the analogue side that is 52,684 unique photographs against 134,340 frames, an average of about two and a half frames per picture. Individual folders show the pattern: one advertising folder is listed as 355 frames from 74 shots.

What the inventory does not yet say is which of the additional frames are in-camera variants, each an original, and which are laboratory copies. The counts come from the archive's own inventory and have not been independently reconciled. Making that distinction, folder by folder, is part of the cataloguing that lies ahead, and it will determine what is scanned first. The principal format is 6 x 7 colour transparency film, described on the analogue collections page.

Picture libraries and agencies with experience of sorting originals from dupes are exactly the partners this work needs; the page for picture agencies explains how to get in touch.

Frequently asked questions

What is a duplicate transparency?

A duplicate transparency, or dupe, is a copy of a slide or transparency made on another piece of colour reversal film. It is produced either by contact printing or by photographing the original through a lens. It is a second-generation image: a photograph of a photograph, not the film that was in the camera when the picture was taken.

How can you tell a dupe slide from an original?

Check several things together: the film stock named in the edge markings, whether the film format matches the camera that would have been used, the look of the frame edges, raised contrast and blocked shadows, slight softness, dust or scratches that are part of the image, and any label or stamp on the mount. No single sign is conclusive.

Why did picture agencies use duplicate transparencies?

Because an original exists once. An agency wanted to send the same picture to several clients, and to partner agencies abroad, at the same time, without risking a unique piece of film in the post. Duplicates could be made in quantity, mounted and distributed, and replaced if a client lost or damaged one.

Is a duplicate transparency worth scanning?

Yes, if the original cannot be found; a good dupe may be the only surviving version of a picture. Where both exist, scan the original, which has more detail and a longer tonal range, and record the dupes without scanning each one. Either way, note the generation in the catalogue so users know what they are looking at.

Are in-camera duplicates the same as dupes?

No. Photographers often exposed several frames of the same set-up, sometimes bracketing the exposure, so that more than one original could be sent out. Those frames are all camera originals, each slightly different. A dupe is made later, from one of them, on duplicating film. The distinction affects quality, dating and how the items are counted.

Sources

Published by Archive.London. Corrections are welcome under our editorial policy.

Help digitise the archive

52,684 analogue originals, mostly medium-format transparencies, are waiting to be scanned and catalogued. We are speaking to scanning, cataloguing and tagging partners.

See the digitisation briefWrite to the archive