Types of Content Metadata

There are many ways of organizing Content Metadata. If your Content Metadata is focused on scripted narrative, your organizational structure will almost certainly revolve around Scene and Shot.

With more free form Content Metadata, it can be useful to organize using some sort of grouping structure. I am a strong advocate that we need to keyword:

  • Location,
  • Person,
  • Activity, and
  • Content.

These are the categories that we will use for organization and story pre-editing. If we have B-roll having it organized by Location, and having Locations organized together makes sense. If we’re organizing a Panel discussion, or want to be be able to find a person talking about a subject (content). If you’re looking to string out all the plays by a particular player from a sports game or practice, then having Person-based Content Metadata becomes very valuable.

There are situations where tracking an activity makes sense, such as the plays in a sporting match. And, of course, being able to find all the subjects discussed in many interviews, all organized by Keyword, or in a  pre-edit string-out, is fundamental to documentary editing.

Many NLEs have ways of generating compound searches, for example Smart Collections in Final Cut Pro X, and Search Bins in Premiere Pro CC. These live searches help find and filter to exactly what you want to find and start editing with.

So, based on the way Content Metadata gets used, these four categories make sense.

These are very broad categories. For example, Location could refer to:

  • The fictional location from a script,
  • A broad description of the general area, such as a town or famous location,
  • A very specific location: “Marlon’s Apartment”, “Bob Muravez’ Garage”, etc, or even as specific as “Foredeck” and “Aft deck” if you were on a boat.

People can be animals – literally, if you are working on a project that revolves around any group of animals.

Ultimately this metadata is what forms the foundation our goal of editor-ready, pre-edit string-outs.

Continue to Where do we get Content Metadata…