The film stock that burns is the film stock that survived

There is a short answer about nitrate film stock and a useful one, and they are not the same. What follows is the useful one.

Two different chemistries, two different failures

Film manufactured before the middle of the last century generally used a cellulose nitrate base, which is chemically related to materials used as propellants. It burns readily, sustains its own combustion without external oxygen and cannot be reliably extinguished once it is properly alight. The replacement, cellulose acetate, was marketed as safety film and removed the fire risk almost entirely.

Acetate then revealed its own decay pathway, in which the base breaks down, releases acetic acid and causes the film to shrink and buckle. Neither material is stable across a century, and the archival problem is a choice between two different kinds of loss.

Nitrate decay is progressive and self-accelerating

Decomposition begins with a faint smell and mild stickiness, moves through softening and blistering, and ends with the emulsion detaching entirely. The products of decay catalyse further decay, so a deteriorating reel accelerates and can affect neighbouring material stored with it.

In the edit, because of this, archives isolate suspect reels and inspect holdings on a schedule rather than waiting for a problem to announce itself. The optical quality of surviving nitrate prints is frequently remarkable, which is why archives take the trouble rather than simply copying and discarding. Projection of nitrate is permitted only in specially equipped venues under strict conditions, and very few such venues remain anywhere.

Acetate decay is quieter and just as final

The characteristic sign is a vinegar smell, produced as acid is released from the decomposing base, which gives the phenomenon its informal name. Once it starts, the process is self-sustaining in the same way nitrate decay is, and it spreads through a store if material is not separated.

By the second act, shrinkage is the practical consequence, since a shrunken film cannot be transported through a scanner or a printer designed for its original dimensions. Specialised equipment exists for handling shrunken material and it is slow, expensive and available in relatively few places. A collection can therefore be physically present and effectively unreadable, which is a distinction the public rarely hears about.

Polyester changed the calculation again

Later polyester base is considerably more stable than either predecessor and is now standard for preservation copies. It is also very strong, which introduces a practical hazard because a jam in a projector will damage the machine rather than break the film. Stability of the base does not protect the image, since colour dyes fade independently of what they are coated onto.

This is why black and white separation copies were historically made from colour originals, storing the colour information as three stable records.

Preservation therefore addresses base and image as separate problems with separate solutions.

Storage does most of the work

Low temperature and controlled humidity slow every chemical decay pathway substantially, and this is the single most effective intervention available. Cold storage is expensive to build and expensive to run continuously, and it fails badly if allowed to cycle rather than remaining stable.

What the cut is doing: material moved from cold storage must be acclimatised gradually to prevent condensation, which adds time to any access request. Underfunded archives are frequently forced to choose between storing more material poorly and less material well. That choice, made repeatedly over decades, has shaped what survives far more than any decision about cultural value.

Copying is not the same as preserving

Transferring content to a new carrier preserves the information and discards whatever the original carrier itself recorded about how the film was made. Archives therefore retain originals wherever possible even after digitisation, because a better scanner will exist in twenty years.

Digital storage introduces its own maintenance requirement, since files must be migrated, checked and duplicated indefinitely rather than simply shelved. The running cost of digital preservation is continuous where a can on a cold shelf costs almost nothing per year. Serious archives run both, and the argument that one replaces the other has been settled repeatedly in favour of keeping the object.

The takeaway

Cold, dry and stable storage buys decades. Nothing available buys reversal.

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