Can text be made to sound more than just its words? (2022)
20 days ago
- #Typography
- #Human-Computer Interaction
- #Accessibility
- Captions typically represent words the same way regardless of vocal nuances like bawling, whispering, or yelping.
- The paper proposes embedding visual representations of paralinguistic qualities (loudness, pitch, duration) into captions using typography (font-weight, baseline shift, letter-spacing).
- An evaluation showed participants could match speech-modulated typography to original audio with 65% accuracy, with no significant difference between animated or static text.
- Participants' mental models of speech-modulated typography varied widely, indicating diverse interpretations of the visual cues.