audio-dsp（音频 DSP）

Audio DSP Skill

Expert knowledge for implementing audio digital signal processing in web applications.

When to Use

• Implementing audio effects (reverb, delay, chorus)
• Real-time audio processing with Web Audio API
• FFT analysis and visualization
• Convolution-based effects
• Audio worklet development

Web Audio API Patterns

Basic Audio Context

const audioContext = new AudioContext();
const analyser = audioContext.createAnalyser();
analyser.fftSize = 2048;

Audio Worklet (Real-time Processing)

// Register processor
await audioContext.audioWorklet.addModule('processor.js');
const node = new AudioWorkletNode(audioContext, 'my-processor');

Common Algorithms

Convolution Reverb

const convolver = audioContext.createConvolver();
const response = await fetch('/impulse-response.wav');
convolver.buffer = await audioContext.decodeAudioData(await response.arrayBuffer());

FFT Analysis

const dataArray = new Float32Array(analyser.frequencyBinCount);
analyser.getFloatFrequencyData(dataArray);

Knowledge Base Resources

Query for DSP algorithms:

SELECT * FROM algorithm WHERE tags CONTAINS 'audio' OR tags CONTAINS 'dsp';
SELECT * FROM paper WHERE abstract @@ 'convolution reverb impulse';

Performance Guidelines

1. Use AudioWorklet for real-time processing
2. Avoid main thread blocking during audio
3. Buffer sizes: 128-512 for low latency
4. Sample rate: Match device (usually 44100 or 48000)
5. Use SharedArrayBuffer for worklet communication

React Integration

// Custom hook pattern
const useAudioAnalyser = (audioContext: AudioContext) => {
  const analyserRef = useRef<AnalyserNode>();

  useEffect(() => {
    analyserRef.current = audioContext.createAnalyser();
    return () => analyserRef.current?.disconnect();
  }, [audioContext]);

  return analyserRef;
};

Related KB Queries

-- Find reverb implementations
SELECT * FROM knowledge WHERE content @@ 'reverb implementation';

-- Get validated algorithms
SELECT * FROM algorithm WHERE success_rate > 0.8;