TouchDesigner
Video Pipeline
Camera In
Video capture
TOP
Pixelation
Effect chain
TOP
to image transform
Image In
Still image source
TOP
Image Preview
Pre-diffusion frame
TOP
prompt + image
StreamDiffusion Node
Real-time img2img generation
Prompt from transcript
TOP
generated image
Audio Pipeline
Audio Device In
Mic capture 44100Hz
CHOP
Audio File Out
Record toggle by script
CHOP
Analysis CHOP
RMS Power — voice trigger
CHOP
Audio Filter
Bandpass 4–8kHz
Clap detection
CHOP
rms value
Script Pipeline
CHOP Execute
RMS trigger
onValueChange()
DAT
CHOP Execute
Button override
onOffToOn()
DAT
call API
Python Script
urllib.request → OpenAI Whisper
multipart/form-data WAV upload
DAT
transcribed text
Text + Compositing
Transcript Output
Latest speech text
Resets each take
DAT
Text Container
Renders text overlay
TOP
Current Question
Clap-navigated
loops at 6
DAT
Image Transform
Scale / position
Color adjust
TOP
composite
Movie File Out
Snapshot on silence
take_TIMESTAMP.jpg
TOP