Process of storing image information on celluloid—analog medium with continuous tonal range and distinctive grain. Nostalgic now, but technically superior on scan.
You're shooting 35mm film, and suddenly you realize: it's a different language than digital. Film recording works with continuous chemical processes — silver halide crystals react to light, and the result is a physical, non-hierarchical density of information that scanners have to "retrieve" again today. It's not a sensor that samples discretely. It's analog memory.
On set, this means: you expose, the grain works for you — not against you like with early digital cameras. At set light levels of T2 or T2.8, 500 ASA material captures what your naked eye can barely distinguish. At the same time: you need a light ratio that the emulsion can handle — expose too long, and the shadows clip; too short, and the highlights are blown out. This tension between rendering and contrast is the fundamental melody of all film recording. Digital sensors are more brutal: they show you immediately when something isn't right. Film forgives — but only during the scan do both see what's really there.
The practical side: you're tied to the film length of a reel — 400 or 1000 feet, depending on the camera. This forces a rhythm in shooting that digital cameras don't have. Long takes are expensive. This shapes the staging, the editing frequency, even the work with actors. And the sound quality? Optical film recording with Dolby or DTS stores on the film edge — perfectly synchronized, no drift, no problems with unsynchronization like with separate digital audio recorders of the past.
Today, film recording is a conscious choice. You choose it for aesthetics (the grain, the color modulation), for archival security (a well-stored 35mm negative lasts 100+ years), or for hybrid workflows — shoot on film, scan in 4K, edit digitally, back-conform to the original for archive. This isn't going backward, it's redundancy with depth.