What does good metadata look like? As we saw in an earlier article, it is extensive, precise, unique, allows inference, and is portable. Healthy metadata also helps audition tracks, and organize them.
Of course, that’s the conceptual approach. What are some hard and fast rules of composing metadata?
Not sure how to write the best metadata for your sound fx collections? Want to spend as little time as possible and receive superior results sound pros will love?
Today’s post shares a list of rules or “commandments” you must follow to create the best sound fx metadata for your field recordings.
Adopting Metadata Rules
Well, it’s true. I’m using the term with a bit of tongue-in-cheek mock gravity. Just the same, is it reasonable to have absolute rules for writing sound fx metadata?
Why? Well, sound clip metadata is used for one prevaling purpose: to use text to help sound pros find and select sound fx and field recordings quickly.
Since we know that, it is completely sensible to design fixed rules to make this task easier, more efficient, and more precise.
What benefit can these “commandments” give you? How can you use them? Think of this list as a proofreader’s checklist for auditing your sound library metadata.
The 7 Sound FX Metadata Commandments
1. You shall write unique descriptions
No two metadata Description fields must be the same. Duplicate metadata is inscrutable. It forces sound pros to audition every repeated clip, wasting their time. It also makes your collection seem as if it stacked with filler audio instead of precise, unique, and valuable clips.
|Car pass 01||Car passing by quickly with tire sizzle|
|Car pass 02||Car passing by medium fast while kicking up gravel|
Bonus tip: can’t write a unique name for every sound? That’s a sign that the clips are too similar and are essentially copies. Duplicate sounds shortchange your fans of value and contribute to library bloat. Keep one and delete the rest.
2. You shall use strong, active terms
Powerful words are compelling. They create vivid imagery in a listener’s mind, tempting them to audition and select that sound. Avoid poor metadata with passive sentences and weak words. Instead, compose Description field metadata with active sentences with strong terms. Examples:
|A cup being dropped/A mug falling||Porcelain cup dropping and shattering with a burst and fragments scattering and skipping away|
|Dog makes some noise with other sounds||Bulldog snarling viciously with rippling growls and occasional collar rattles.|
3. You shall use words everyone knows
Avoid abbreviations or esoteric trade name terms in all metadata fields. Casual listeners will be mystified by shorthand terms used by ace mixers and sound supervisors. Help them by simplifying language that everyone will understand, not just sound pros.
Using common words ensures people will type familiar keywords that will locate your sounds. Some examples:
|EXT||Outside or Exterior|
|AMB||Ambience or Atmosphere|
|CU||Close (not "close up") or omit entirely since a nearby distance is implied|
|Dist||Describe the distance or position (e.g., 5 meters)|
|POV||Describe the perspective (e.g., behind door)|
|Foley||Describe the Foley performance instead|
|Library stock or release numbers (e.g., AIR-017)||Omit entirely as they have minimal practical use for your fans|
Bonus: familiar keywords help with Internet SEO.
Tip: skip Soundminer's ShortID field for this purpose. Its format doesn't contribute to useable, understandable metadata for others.
4. You must write a minimum length
Use at least 75 characters of text or 12 words in the Description field. The minimum length ensures a sound is described fully and is packed with juicy keywords.
|Mossberg shotgun blast 03||Mossberg 12-guage shotgun racking quickly and firing, shot echoes amongst the mountains with long decay|
Note: don’t stack words just to create length. Each term must create value.
5. You shall complete as many fields as possible
Have you written an amazing metadata Desription? Don’t rest there. If they are available in your metadata software, fill out the Category, Sub-Category, Microphone, URL, and more. Used together with the metadata Description, these fields allow for inference and context, and give an impression of the sound larger than the text alone.
6. You shall be more consistent than correct
This idea evolves from the concept that in terms of providing a predicitable, reliable data set, “it’s better to be consistent than correct.” Consistent metadata teaches fans how to find your library sounds, even if it uses language unique to your collection alone. In short, ensure terms and formatting are the same throughout your collection.
|Colour then using color later||Pick one term and stick with it|
|POV alternating with perspective|
|Writing conversational descriptions (e.g., Cat meowing strongly) and then scientific, linear ones (e.g., Cat, Meow, Strong)||Pick one method and stick with it|
|Placing some monster voices in the "Creature" category, and others in "Sound Design"|
7. Always check your work
Ensure words are spelled correctly with proper punctuation. If necessary, export your metadata and spell-check it in Excel or Numbers. Consult a punctuation guide to ensure commas, colons, semi-colons, and others are all being used, and used correctly.
The best way to check your work? Read everything you've written aloud. That will catch obvious errors and ensure the metadata is readable and flows properly.
The Benefit of Metadata Rules
Does every sound fx library’s metadata need to follow these seven rules?
No, of course not. Adopting even one of these metadata commandments will strengthen your sound fx. A collection will improve with each additional tip the sound fx includes. You’ll notice that elite metadata includes all these features.
The best result?
You’ll find yourself using these powerful sound clips again and again.
Tweet Follow @paulvirostek
To stay in touch, receive free updates by email newsletter or RSS feed. | Follow on SoundCloud