Metadata and provenance

Metadata is descriptive information that an application stores alongside your production and media. It can include titles, descriptions, keywords, rights information, camera notes, language-specific text, or fields defined by another production application.

The same item can legitimately have several values. A clip might have several keywords or titles in different languages. PostProject keeps those values separate and does not silently pick one.

Some metadata belongs to the logical asset, while other metadata describes one particular representation such as a camera original or proxy. PostProject keeps that distinction so an application does not accidentally present representation-specific facts as facts about every copy.

The CLI can optionally inspect imported files with an installed ffprobe and store container, stream, codec, dimensions, rates, channel layout, bit depth, pixel format, duration, timecode, and other embedded tags as typed metadata:

postproject media add production.pproj camera.mov --inspect

Inspection is optional. If ffprobe is missing or rejects the file, import still succeeds and the command reports unavailable or failed rather than inventing technical values.

Provenance answers a different question: how was a result produced? An activity connects input representations, an operation, and its output representations. It can also record the responsible tool, version, agent, timing, and processing parameters. Descriptive metadata and production provenance can complement each other, but neither is proof that content is authentic. Cryptographic trust systems such as C2PA remain a separate layer.

All current metadata remains inside the local production file. PostProject does not upload it, contact vocabulary services, or look up identifier registries.