Your guitar talks. Named for the raccoon dog, the canid that Japanese folklore says can take on any shape and voice. Tanuki tracks the note you play, and every note you pick starts a syllable in a synthesized voice at that pitch. There are five characters, all synthesis with no recordings: a bright singer, a laid-back "wow", a gravelly grunt, a tiny robot and a sour shrieker. Click anything on the drawing.
Top view. One page, no paging: every control does one thing. DIP 1 is the top switch on the right side.
Pick a note and the voice says the next syllable at that pitch. Hammer-ons, pull-offs and slides count as new notes too. The voice gets louder and quieter with your playing. Stop for a second and a half and the phrase starts over from its first syllable. It's monophonic: play single notes, not chords.
The voice is built from scratch every time. A buzzing source at your pitch passes through four resonances that move like a mouth (the formants), with bursts of breath and hiss for the consonants. TALKBOX swaps the buzz for your guitar, so the guitar itself is shaped into words.
VOICE (knob 1) picks one of five characters. PHRASE (knob 2) picks what it says: one held vowel, a cycle of syllables, or its catchphrase across your notes.
Toggles change how much of each character you get. The DIPs change what the voice is made of.