A clip test sacrificed a frame or two so the rest of the roll could be developed to within a third of a stop. The instruction survives in biro on the bag, and it explains more than most captions.
The scribble on a film bag is the only place where the photographer, the lab and the light of that particular day are all recorded together.
A clip test is a few inches of exposed film, cut from the end of a roll in the dark and developed ahead of the rest. The photographer looked at it on a lightbox, decided the pictures were a touch dark or a touch light, and told the lab to develop the remainder for slightly longer or slightly less. That is the whole idea, and manuals describe it in a paragraph. Our angle is what happened next. The instruction was written on the bag, the bag went into a box, and decades later those few characters are often the only surviving explanation of why frames from a single shoot differ in density, contrast and colour. For anyone scanning, captioning or building a dataset from old film, the bag is evidence and should be treated as such.
What is a clip test?
It is a trial run using real film from the real job. There were three common versions.
- Roll film. The lab cut a short length from the head or tail of the roll, enough for a frame or two, and processed it at the normal time. The rest of the roll, called the balance, was held back.
- Sheet film. One sheet from a set of identical exposures was processed first. This was the cleanest method, since nothing had to be cut.
- A test roll. On a large job, one whole roll was processed ahead of the others.
The photographer or an assistant then viewed the clip and gave the instruction for the balance: normal, or a push or pull expressed as a fraction of a stop. Because the cut was made blind, it generally went through the middle of a frame, so photographers who expected to clip would expose a few expendable frames at one end, or simply accept the loss.
Usage varied from lab to lab, and so did the name. Some called it a snip test. We describe the general practice here and not any one laboratory's rules.
Why did professionals bother?
Because of the film they used. Colour transparency film was the preferred material for publication until about 1995, valued for its contrast and resolution, and it has fewer stops of exposure latitude than colour negative film. With a negative, a modest error is corrected when the print or scan is made. With a transparency, the piece of film from the camera is the final image. Half a stop too dense and the shadows block; half a stop too light and the highlights are empty, and no later stage can put them back.
Studio photographers had already narrowed the odds before the film reached the lab. They metered, they proofed on instant film, as described in Polaroid proofing before digital, and they bracketed, exposing the same set-up at several settings a fraction of a stop apart. The clip test was the last adjustment, made after the event, and it corrected for things no meter could see: variation between emulsion batches, the true output of a flash pack that day, or the lab's own chemistry. It also meant that an entire roll could be shifted towards the best exposure, where bracketing alone gave only one best frame in three.
What do push and pull processing do?
Pushing means developing for longer than standard, which builds density in film that was underexposed. Pulling means developing for less, which holds back film that was overexposed. In both cases the adjustment is made in the first critical development step.
| Standard | To push | To pull | |
|---|---|---|---|
| E-6 (colour transparency) | First developer about 6 minutes at roughly 38 degrees C | Lengthen the first developer; one kit maker suggests about 25 per cent per stop | Shorten the first developer |
| C-41 (colour negative) | Developer 3 minutes 15 seconds at 37.8 degrees C | Kodak recommends 30 seconds more for one stop | Less common |
| Black and white | Depends on film and developer | Extend development | Reduce development |
The side effects are well documented. Push processing increases contrast and grain and reduces resolution, and in colour film it can saturate and distort colours. Pull processing lowers contrast and softens the image. With transparency film a push also tends to weaken the deepest blacks. This is why professional labs offered adjustments in small steps, commonly thirds or halves of a stop, and why a clip test was normally run before anything more than a mild correction.
Pushing was also used on purpose, with no error to correct. Rating a film at double its speed and pushing it one stop bought a faster shutter speed in poor light, at a price in grain that some photographers positively wanted. The black-and-white version of that bargain is discussed in Tri-X, HP5 and FP4.
How do you read the notes on an old film bag?
Notation was never standardised, so read each bag in the context of its own job and its own lab. The following are typical forms and their usual meanings.
| Marking | Usual meaning |
|---|---|
| N, Norm, Normal | Processed at the standard time |
| +1/3, +1/2, +1, Push 1 | Pushed by that fraction of a stop |
| -1/3, -1/2, Pull 1/2 | Pulled by that fraction of a stop |
| Clip, C/T, Snip | A clip test was requested or made from this film |
| Hold, Held | The balance was kept back awaiting the result of a clip |
| Bal, Balance, Run | The remainder of the roll, and the instruction given for it |
| Roll 3 of 8 | Position in the job; rolls were not always processed alike |
| E-6, C-41, B/W | The process, which identifies the type of film |
| A job or docket number | The lab's reference, sometimes matched on a delivery note |
Two cautions. First, a plus sign on a bag is a processing instruction, and should not be confused with exposure brackets marked on a transparency mount or sleeve, which record what was done in the camera. Second, absence of a note does not prove normal processing. Bags were reused, replaced and separated from their contents. Treat the note as testimony: valuable, and to be weighed.
Why do frames from one shoot look different?
Knowing the lab routine turns an apparent muddle into a sequence. In a sleeve of transparencies from one studio session you may find:
- Brackets. The same composition three or five times, stepping from lighter to darker. These differ in exposure and nothing else.
- The clip. One or two frames, often cut or incomplete, sometimes filed separately or missing altogether. They were processed normally whatever happened to the rest.
- The balance. The remaining frames, pushed or pulled, with slightly different contrast and colour from the clip.
- Different rolls. Later rolls processed to a different instruction once the first results were seen.
- Age. Dye fading that varies with storage, described in why colour slides fade.
Only the last is damage. The first four are the photographer and the lab doing their jobs, and the frame eventually chosen for publication was the product of all of them.
What should a scanning operator do with this?
Record it, and resist tidying it up.
- Capture the bag. Photograph or transcribe every envelope, sleeve header and docket before the film is separated from it. A snapshot of the bag linked to the frame numbers costs seconds.
- Put the note in the metadata. A simple field such as "processing: E-6, pushed 1/2 stop, clip tested" travels with the file. Our guide to captioning from contact sheets and job bags covers where such notes belong.
- Scan to the film. A preservation master should record the density and colour actually on the film, using fixed, profiled settings. Automatic exposure on a scanner will quietly equalise a bracket and erase the evidence.
- Keep variants linked. Brackets and clips are near-duplicates. Group them under one shot identifier so that nobody later counts five frames as five pictures. Techniques for this are in finding duplicate photographs with hashes, and two frames can be checked side by side with our image compare tool.
- Flag pushed film. Pushed transparencies have more grain and contrast, which affects sharpening, dust removal and any automated quality check.
The same points apply to anyone preparing images as data. A set in which the same subject appears five times at slightly different exposures is a different thing from five independent photographs, and it should be described as such.
What the Robert Harper inventory shows
This is where the subject stops being historical for us. The inventory of the Robert Harper Photograph Library is counted in two ways throughout: shots and frames. For the analogue material it records 134,340 frames against 52,684 unique photographs. A folder of cocktails, for example, is listed as 250 shots across 630 frames. The difference between the two figures is duplicates and variants, and the working methods described above are among the ordinary reasons a studio archive contains them. We have not yet analysed how much of the difference is bracketing, how much is clip tests and how much is duplicate transparencies, and we will not guess. The counts come from the inventory and have not been independently reconciled.
More than half of the archive is still on film and has yet to be scanned, most of it 6 x 7 colour transparency film in the analogue holdings. The folder-level figures are published in the archive inventory. A digitisation partner who knows what to do with a biro note reading "bal +1/3" is exactly the kind we are looking for, and the page for digitisation partners explains how to get in touch.
Frequently asked questions
What is a clip test in film processing?
It is a trial development. The lab cuts a short piece from one end of an exposed roll, or takes one sheet from a batch, and processes it normally. The photographer judges the result on a lightbox and tells the lab how to process the remainder: normally, or pushed or pulled by a stated fraction of a stop.
What does push one stop mean?
It means the film is developed for longer than standard, as if it had been rated at twice its box speed. An ISO 100 film pushed one stop is treated as ISO 200. The image gains density and contrast, with more visible grain and some loss of shadow detail. Pull one stop is the opposite: shorter development for film that was overexposed.
Why do slides from the same shoot look different?
Several reasons, often together. The photographer may have bracketed, exposing the same picture at slightly different settings. Part of the film may have been a clip test processed differently from the rest. Different rolls may have been pushed or pulled by different amounts. And transparencies stored in different conditions can fade at different rates.
Does a clip test destroy part of the film?
Usually, yes. Cutting a strip from a roll in darkness will slice through a frame at an unknown point, and the clipped frames are processed at a time that may not be ideal. Photographers allowed for this by exposing a few expendable frames at the start or end of a roll, or an extra sheet.
Should scans be corrected so pushed and normal frames match?
Not in the master file. A preservation scan should record the film as it is, including its density and colour, with the processing note captured in the metadata. Matching frames to each other is an editorial decision that belongs in a derivative copy, where it can be reversed and where the original difference remains on record.
Sources
Published by Archive.London. Corrections are welcome under our editorial policy.

