Every style pack in packages/videojs/styles/, with each semantic class of
examples/fixtures/sample.overtone.json in styled and plain mode, at 16:9 (horizontal profile) and
9:16 (vertical profile); the overview adds a 1:1 frame (a profile cut with overtone recut --profiles 1:1).
Frames are rendered with the player's own renderCue() and applyProfileToLayer() at the
moment the styled word is spoken, so motion like the social pack's "pop" is visible. A red outline marks a line
that wraps; the line-fit check renders every cue at 16:9, 1:1 and 9:16 widths.
Make your own: docs/contributing/style-packs.md.