Why Photochemical
There are two ways to make an image look like film. You can copy the finished look, or you can rebuild the process that produced it. Copying the look means measuring a finished print and baking that one result in: a single transform that holds how a particular stock, developed a particular way, rendered one set of choices. Rebuilding the process means running your image through the stages a shot used to run through, the negative, the lab, the print stock, and the light that passed through all of them, so those choices stay open as you work. Film/Emulsion™ takes the second route, so the final look is a consequence of the stages rather than a copy of one result.
Two ways to make a color image
A screen and a strip of film build color in opposite directions. Kodak describes the screen as additive: it starts from black and adds red, green, and blue light until the eye reads the color you want. Film is subtractive. It starts from white light and takes color away with dyes, each dye removing part of the spectrum and passing the rest. Kodak notes that the earliest color photographs were in fact made additively and that the approach works, but it lost on efficiency. An additive dye passes only a narrow slice of the spectrum and blocks the rest, so a viewable image needs a great deal of light. Subtractive dyes each remove only a narrow band and let most of the light through, so three dyes at varying strengths cover a wide range of color at ordinary viewing levels.
Kodak frames the whole of color film as three fixed dyes whose densities vary. Not different dyes for different colors, the same three, laid down more or less heavily. A film image is a three-channel record of density, and every color in it is a combination of how much cyan, magenta, and yellow sit at that spot. A film model built on three channels of density follows the way film itself makes color, not just the look of a finished frame.
The image is a record of density
When a lab or a scanner reads a developed negative, Kodak describes it measuring how much red, green, and blue light gets through at each point, one transmittance for each channel. Density, the measure film is characterized in, is the log of the reciprocal of that transmission, and rises as the film blocks more light. Either way the file already carries what the stock did to the light, its color, its contrast, the way its highlights roll off, before any grade begins. It records the negative’s response, not the scene.
This is why the order of operations is not cosmetic. The negative records the scene; the print reads the negative; halation and grain sit in the physical film and travel with it. Change something early and everything after it sees the change, because each later stage works from what the stage before it produced.
The chain, run twice
Kodak describes the release chain as a series of printings: the camera negative is printed to an interpositive, that to a duplicate negative, and release prints come off the duplicate. Colorists describe each of those duplications as softening the image and coarsening its grain and color a little. The core of it they put more plainly still: a negative and a print are the same process run twice. Light exposes an emulsion, development builds dye where the light fell, and the result restricts the light that passes through it next. The negative inverts the scene; the print inverts it back. Two stages, one mechanism.
Film/Emulsion™ keeps that structure. The Overview lays out the full path, but the shape of it is the argument: a Negative stage that records the scene onto a stock you choose, a Print stage that reads that negative onto a print stock, then halation and grain on the finished image. The lab folds into the two film stages, the way a real lab does, because a lab has no image of its own; it changes how the negative develops and how the print is struck.
Film talks back
The reason to model the process rather than the look is that the process is coupled in ways a single look table cannot hold. Colorists who grade on film describe its controls as non-orthogonal: push one and others move, sometimes the way you wanted and sometimes not. Overexpose the negative and bring it back to a normal brightness, and Kodak describes the color returning more saturated and the grain finer than a normal exposure would have given. That is not two effects stacked together; it is one physical system responding as a whole.
A single baked result freezes one set of choices: this stock, this development, this print. Change your mind about the exposure and you are outside what that table was built to show. The stages keep the choices open, because the coupling is modeled rather than pre-computed: the negative stage records a brighter or darker scene, and the print stage responds to the denser or thinner negative that results. When you move Exposure on the Camera tab and watch the grain and the highlight rolloff shift with it, that coupling is doing the work.
Honesty about “accurate”
Reproducing the process does not mean chasing accuracy, and it is honest not to pretend otherwise. Kodak is blunt that no color-photographic process is accurate in an absolute sense. Careful measurement of even a fine photograph shows real differences from the original subject, and because the viewer’s own perception adds to that, Kodak treats perfect reproduction as a goal that does not mean anything. The stocks were tuned to be pleasing, not correct. Reproducing the process inherits that honestly: the aim is the character film actually had, including the parts that are, strictly speaking, wrong.
What photochemical emulation cannot do
Modeling the chain does not recover information the chain never had. Kodak notes that a blown-out highlight cannot be brought back after the fact; if your input has already clipped, no stage here restores it. Balance and expose the scene before the plugin, which is what the Resolve Quick Start asks you to do.
It is also one stage in a longer pipeline, not the whole look. Some colorists argue that a large part of what we call the film look lives in the display-rendering math applied to every shot, not in the medium alone. Film/Emulsion™ does not include a display transform; it hands the printed image back in the space it received and lets your output transform finish the job. The Overview covers that boundary. The plugin gives you the film; where it sits in your grade is still yours.
The account of additive versus subtractive color, the three-dye density model, the scanning of a negative as transmittance, and the release-printing chain all draw on Kodak’s Color Films, Motion Picture Color Theory, Exploring the Color Image, and Essential Reference Guide. The description of the negative and print as one process run twice, of film’s controls as coupled, and of the look residing partly in display rendering comes from colorists’ accounts of working photochemically rather than from a manufacturer.
Next: The Negative follows the scene into the camera stock and explains what a negative actually records.