Expressive Captions

Captions that go beyond literal word-for-word transcription to convey the prosodic, emotional, or speaker-identity information that traditional captions strip out. Expressive captions may modulate font weight, size, colour, position, or animation to signal loudness, pitch, excitement, or emotional tone; some systems add haptic or visual icons. The goal is to give d/DHH viewers access to extra-speech information (ESI) that hearing audiences absorb automatically from vocal delivery. Research has shown that viewers prefer valence-encoding typography (clear emotional categories) over arousal-encoding (continuous intensity), and that readability must be balanced against expressiveness. Variable fonts (e.g. Recursive) are commonly used because their weight and width axes can be modulated per-word in real time.