As with anything I do, I like to take the most scientific approach possible. So… prepare for a deep dive. There are usually two ways film grain can be emulated: with a grain plug-in, or through manual compositing.

What most plug-ins out there do

Normally, a grain plug-in creates random noise that is statistically inspired by film, and composites that onto the image. Then you can tweak the amount, or compositing strength, of the grain in different luminance regions like shadows, midtones, and highlights. And that’s pretty much it. The problem is that this approach neglects an important characteristic of film: the grain appearance is density-dependent.

Taking DaVinci Resolve grain as an example, the shadow, midtone, and highlight controls modify how strongly the grain is composited onto each luminance range, but they don’t change the underlying noise that is being composited. So yes, you can make the grain stronger or weaker in different parts of the image. But the grain itself is still the same grain.

A better approach, used by very fancy and expensive plug-ins like LiveGrain, is to shoot grey charts and create grain plates at different densities, or exposures, and then composite those plates into their appropriate luminance ranges. This already brings us one step closer, because it takes into account how film behaves at different exposure levels. But still…Film grain is more complex than that. In fact, this approach still doesn’t solve a fundamental problem: