What is metadata?
Metadata is data about data. Too easy, right? Metadata is data that describes a data object. In our case, the data objects are video clips.
Why is it important?
Metadata helps people and systems process, administer, and discover the content you create. Good metadata is critical to ensuring your clips are licensed correctly and easily found by customers.
Help us help you
Think about metadata as a fundamental part of your content creation process. You wait for the best light, you hike to the perfect vantage point, you work with the right talent — do you describe your content with the same care and attention? If you don’t, you’re doing your work a disservice.
Your descriptive metadata should reflect the quality of your content. Better footage warrants better metadata. Above all else, our customers want accuracy in the product. Concise, accurate descriptions and straightforward keyword sets are crucial. Every customer should be able to find the right clip, and every clip should be able to be found by its customer.
You can help ensure this happens by providing the best metadata possible. Here’s how:
Provide a short, clearly written description of your clip. This description will be both visible and searchable on Dissolve, so strive for accuracy in capitalization, spelling, and grammar. Use natural, everyday language.
For a good description:
- Focus on what the clip is of, rather than what the clip is about. You may include the latter, but you should always include the former.
- Strong descriptions cover the following:
- The subject(s): who or what is covered in the clip
- The action: what takes place in the clip
- The location: where the action takes place
- If your clip uses a special technique for the shot (e.g., timelapse), include that in the description: Timelapse of a yellow flower blooming in a garden.
- Qualitative descriptions of the subject are acceptable if they are relevant: Happy couple gazes at each other over a candle-lit dinner table.
- If ethnicity is relevant, include it in the keywords only. It should not be referenced in the description.
Keywords — either single terms (e.g., flower) or multiple words (e.g., basketball court) — must be separated with commas (e.g., sand, beach towel, beach umbrella, sunscreen).
The strongest keywords are specific, cover the entirety of the clip, and do not introduce irrelevant concepts. There is no limit to the number of keywords allowed for each clip, but excessive or irrelevant keywords will harm the retrieval of your clips in search.
For good keywords
- Include a specific reference to what the clip is of (e.g., hydrangea) and to what the clip is about (e.g., botany).
- Include the subjects, the action, the location, and techniques used to produce the clip.
- When keywording genres, think about what customers may be searching for (e.g., travel, nature, performing arts), but be sure they are relevant.
- Strive to be exhaustive, but only with respect to the primary subject of a clip and what the clip is about. Don’t worry about describing incidental subjects, such as a rock in the background.
- Use your knowledge of the shoot to include specific terms that may not be obvious to a customer. For example, if a clip shows a particular dance movement, include the name of that move.
Specific keyword fields
In the upload site’s metadata editor, you will be prompted with two specific keyword fields:
- Geographic location depicted — If you wish, and if it is relevant, you may provide the geographic place name of the location in your clip (e.g., San Francisco, California, United States).
- Shot types — Provide terms that describe the type of shot, technique used, or camera movements from the list of available shot types.
- People Count — The number of people depicted in the media. Must be 0 or more.
- People Ages — A list of age ranges of the people depicted in the clip or photo.
More on keyword types
- It might be helpful to specify the number of people in a clip (e.g., group of four). If the number of people is not obvious, crowd or large group will provide enough context for retrieval.
- Include gender, ethnicity, and age group (e.g., infants, teenagers, older adults) if possible.
- Include roles and/or relationships (e.g., firefighters, mother and daughter) if they are relevant. Keyword only what is depicted.
- If possible, describe attributes of the people depicted, such as hair or eye color (e.g., brown eyes), specific articles of clothing (e.g., sundress), or a general style (e.g., casual, formal).
- If relevant, use qualitative terms to describe the location depicted (e.g., crowded street).
- If it is an important part of the subject, include a reference to the time of day depicted in the clip (e.g., dawn, dusk, evening, night, day, sunrise, noon).
- If a clip is related to a particular time period (e.g., 1950s public service announcements), note that period, preferably by decade (e.g., 1940s, 1970s). If you are not certain of the time period, avoid the temptation to guess.
- If there is a predominant mood to the clip, it may be keyworded, but be careful not to pigeonhole a scene with irrelevant or tenuous mood descriptions.
- Avoid the temptation to describe a variety of potential moods. Clips with numerous mood keywords, especially conflicting moods, will not perform well in search and may be edited.
- If the clip was shot with a specific style or theme in mind, or if the clip depicts scenes that are easily identifiable as belonging to a specific genre (that is, a readily identifiable style, form, subject matter, etc.), include that information.
- Savvy users may be searching for clips with a specific cinematographic technique. If the clip depicts a tracking shot, a tilt shot, timelapse, shallow focus, rack focus, etc., include the relevant terms.