From photo to photograph

Enlarge, select a film stock, and develop the grain. A single photograph or a whole roll follows the same process. Each successfully developed image uses one $1 image credit, including whichever of these steps you choose. Failed images keep their credit.

A grain filter lays one frozen sheet of noise over the whole picture, the same specks in the shadows as in the highlights, the same on every image you run through it. Real film does not work that way, and neither does this.

Film is a bed of microscopic silver-halide crystals suspended in gelatin. Each crystal is all-or-nothing: catch enough light and development turns the entire grain into an opaque speck of silver, miss the threshold and it washes away in the fixer. A photograph is that random field of developed specks. Grain is not laid on top of the image; the image is made of grain. That is why a real photograph reads as something that happened rather than something that was rendered.

Because the specks are counted, not painted, the texture follows the light. Where more photons landed, more crystals developed, so the statistics of the grain shift smoothly from the shadows through the mid-tones into the highlights. Color film stacks three of these emulsions, one per dye layer, and the blue-sensitive layer is coarser than the red, so color grain is not gray noise, it has its own structure.

Photofest builds the picture the same way. It works in optical-density space, the quantity the chemistry actually responds to, and models the crystals as a random field whose size and count are set stock by stock from each film's published RMS granularity (the number the manufacturer measured on real scans). Then it develops that field per dye layer. Nothing is overlaid and nothing is invented; the grain is computed from the same physics the film obeyed, which is why it sits differently in every tone of every image instead of the same everywhere.

Every "film look" tool asks you to trust its taste. This one you can check. Kodak and Fuji published a measured number for how grainy each film is (RMS granularity, read off real developed film with a microdensitometer). Photofest's engine is calibrated so each stock reproduces that exact number; the film look here is a measurement, not a vibe.

And there is a built-in lie detector. Work the calibration backwards through Selwyn's law and it tells you how big the grains would have to be. The answers come out at the actual size of silver crystals and dye clouds in real emulsions, a few square microns. If this were noise dressed up as film, that arithmetic would expose it. Here is the whole table:

Every stock lays grain down differently, and the difference is the whole point. These four are the same neutral mid-tone patch developed on four black-and-white films, shown at actual pixels. Nothing is drawn on: the texture is the engine building each emulsion from its published granularity, fine on a slow film and coarse on a fast one.

These are the emulsions the master printer can load, and what makes it reach for each one. You do not have to take its choice: every stock here is available manually, in the single-print lab or for a whole roll at the drop-off.

One image credit covers the selected enlargement, film-stock selection, and grain development. Buy exactly how many you need. No subscription. Failed images keep their credit, and retrying the same development does not use another one.

Buy image credits and view your billing history