Tiny demo demonstrating a speech transcription interface where the the font size of the resulting transcription is proportional to the loudness of the speaker's voice. Think of it as an initial attempt at exploring "augmented transcriptions" that retain more data about the input source than merely their content.