Craft over hypeNo paid reviewsSpoilers always flagged
After the TrailerWhat the two-minute cut left out

Archive

Film restoration is interpretation, not just cleaning

Every restoration makes choices about what the film looked like. Those choices are contested and rarely explained.

Close-up of a curled filmstrip against a white background, emphasizing vintage cinema.
Photograph by Pietro Jeng via Pexels
Editorial note. Independent reporting and analysis. Nothing here is sponsored or paid for. How we work.

The points below about film restoration are ordered by how much difference they make, not by how often they get repeated.

What matters most

  • Original release appearance is frequently unknowable and has to be inferred.
  • Digital tools allow changes that go beyond repair into revision.
  • Nitrate and colour stocks decay in ways that destroy information permanently.

The reference does not exist

Restorers rarely have a pristine original print, because prints were projected until worn and negatives were used until damaged. What a film looked like in 1948 has to be reconstructed from surviving elements, documentation and knowledge of the stocks and processes used.

That reconstruction involves judgement, and different teams reach different conclusions from the same materials. This is why two restorations of the same title can look meaningfully different and both be defensible.

Decay destroys specific information

Nitrate film is chemically unstable and flammable, and a great deal of early cinema is simply gone. Colour stocks fade unevenly by dye layer, which is why so much surviving material has drifted toward magenta. Where a layer has faded past recovery, the original colour relationship cannot be recovered, only estimated.

The acetate stock that replaced nitrate has its own decay, which releases acetic acid and accelerates once it starts, so storage temperature and humidity do more for survival than any amount of laboratory work afterwards.

Digital tools invite overreach

Grain reduction, sharpening and automated damage removal can be applied aggressively, and frequently are. The results look clean and remove texture that was part of the original photographic image.

The distinction between restoring and modernising is exactly where most criticism of commercial restorations sits. The pressure behind it is commercial, since a smooth image sells better on a large modern display, and a distributor is rarely weighing fidelity against profit as an abstract question.

Frame rate and aspect ratio are decisions

Silent films were shot at varying rates and projected variably, so any single playback speed is an interpretation. Aspect ratio was inconsistent for decades, and cropping decisions in restoration can change composition substantially. These are the changes most visible to an ordinary viewer and least often explained on the packaging.

In practice, sound raises the identical question, because a mono optical track can be presented as it was recorded or spread across modern channels, and the second option is common and almost never labelled as a choice.

Documentation is what makes it accountable

Good restorations publish what elements were used, what was done and what was inferred. Archives increasingly treat this as standard practice, and commercial releases often do not. Where no documentation exists, a restoration is a claim rather than a record.

The stronger version is preserving the raw scan alongside the graded result, so a later team can start from the same data instead of returning to physical elements that will have degraded further by then.

Digital preservation is not a solved problem either

A restoration finished today exists as files that have to be copied onto new storage repeatedly and checked for corruption, which is a recurring cost rather than a completed task. Film in cold storage survives decades of neglect; a digital master survives neglect for exactly as long as the institution holding it keeps paying for it.

Archives address this by keeping both, and by writing digital restorations back out onto film as a long-term copy, which is expensive and frequently unfunded. The formats and codecs chosen are a dependency in themselves, so an archive is committing not only to storing what it wrote but to still being able to read it.

Everything above, in order of what to do first

  1. The reference does not exist. Restorers rarely have a pristine original print, because prints were projected until worn and negatives were used until damaged.
  2. Decay destroys specific information. Nitrate film is chemically unstable and flammable, and a great deal of early cinema is simply gone.
  3. Digital tools invite overreach. Grain reduction, sharpening and automated damage removal can be applied aggressively, and frequently are.
  4. Frame rate and aspect ratio are decisions. Silent films were shot at varying rates and projected variably, so any single playback speed is an interpretation.
  5. Documentation is what makes it accountable. Good restorations publish what elements were used, what was done and what was inferred.
  6. Digital preservation is not a solved problem either. A restoration finished today exists as files that have to be copied onto new storage repeatedly and checked for corruption, which is a recurring cost rather than a completed task.

The takeaway

Ask what elements were used and what was inferred. A restoration without that is an opinion.

Watch the transitions. That is where the argument of a film usually is.

Questions readers ask

Is a 4K restoration always better?

Not necessarily. Resolution is one variable; the condition of the source element and the restraint of the grading matter more to how faithful it is.

Why do old films look so clean now?

Often because grain has been digitally reduced. Grain is part of the photographic image, and removing it is a stylistic choice rather than a repair.

Archiverestorationarchivepreservationcolour
Sloane Mercer
Editor, After the Trailer

Sloane edits After the Trailer and is more interested in the second act than the opening weekend.

Also by Sloane Mercer