Skip to main content
← All Projects

DKN Technologies · Hardware + Software

DKN Visuals — Music-Reactive Lighting System

IoTLED LightingAudio DSPESP32Machine LearningDesktop App
DKN Visuals — Music-Reactive Lighting System project interface

Overview

A complete hardware + software lighting platform: a React + PixiJS desktop studio with a top-down room canvas and live per-LED preview, a FastAPI backend running a real-time music-analysis engine and 60fps render loop, and native C++ firmware for ESP8266/ESP32 nodes. The engine phase-locks a beat grid to the music and fires light changes early by the measured pipeline latency, so LEDs land exactly on the audible beat.

Challenge

Turning any room into a synchronised, music-reactive light canvas demands sub-frame timing across audio capture, analysis, network transport, and dozens of independent LED nodes — without perceptible lag or drift.

Solution

Engineered a near-zero-latency pipeline (WASAPI loopback capture, 60fps render loop, UDP pixel transport, adaptive look-ahead beat scheduler) paired with production-grade clock-synced node firmware.

Outcome

A room-scale lighting rig where an ML DJ auto-mode reads BPM, genre, and energy live and drives 25 club effects across the strip like a real fixture rig — landing changes on the beat with <2ms inter-node skew.

Key Highlights

  • Near-zero-latency path — native WASAPI loopback capture, 60fps render loop, UDP pixel transport, and an adaptive look-ahead beat scheduler
  • ML DJ auto-mode detects BPM, genre, and energy live, switching between 25 built-in club effects on downbeats across the room strip-by-strip
  • Production-grade node firmware: <2ms inter-node skew, streaming SHA-256-verified OTA updates, current limiting, watchdog, and a captive-portal setup wizard

Tech Stack

PythonFastAPIReactPixiJSTypeScriptC++PlatformIOESP8266/ESP32
DKN Visuals — Music-Reactive Lighting System | Dylan Kyle Netto