API Reference Overview

Complete reference for the public APIs exported by webgpu-fft.

These API pages are maintained source documentation inside docs/. Built site output under docs/.vitepress/dist is generated and is not the documentation source of truth.

Package Exports

The library provides three entry points:

Entry	Description
webgpu-fft	Core FFT APIs (engines, CPU fallback, types, errors)
webgpu-fft/utils	Utility functions (complex, bit-reversal, windows)
webgpu-fft/validation	Validation helpers

Core APIs (webgpu-fft)

API	Description
FFTEngine	Main GPU-accelerated FFT engine with complex and real-input APIs
cpuFFT	CPU complex FFT fallback plus real-input helpers
SpectrumAnalyzer	CPU-based real-time audio frequency analysis
ImageFilter	CPU-based frequency-domain image filtering

GPU Detection

import { isWebGPUAvailable, hasWebGPUSupport } from 'webgpu-fft';

// Async full check (verifies adapter can be obtained)
if (await isWebGPUAvailable()) {
  const engine = await createFFTEngine();
}

// Sync basic check (only checks navigator.gpu existence)
if (hasWebGPUSupport()) {
  console.log('WebGPU API detected');
}

Utilities (webgpu-fft/utils)

import {
  // Complex operations
  complexAdd, complexMul, complexMagnitude, complexConj,
  // Bit-reversal
  bitReverse, log2, isPowerOf2, bitReversalPermutation,
  // Window functions
  hannWindow, hammingWindow, blackmanWindow, applyWindow
} from 'webgpu-fft/utils';

API	Description
Window Functions	Hann, Hamming, Blackman, Flat Top, and Rectangular windows
Complex Numbers	Complex arithmetic helpers
Bit Reversal	Bit-reversal permutation utilities

Validation (webgpu-fft/validation)

import {
  validateFFT, validateFFT2D,
  validateRealFFTInput, validateRealIFFTInput
} from 'webgpu-fft/validation';

Types

All TypeScript types are exported from the main entry:

import type {
  FFTEngineConfig,
  SpectrumAnalyzerConfig,
  FilterType,
  FilterShape,
  ImageFilterConfig,
  Complex,
} from 'webgpu-fft';

Error Handling

The library uses custom error classes for clear error messages:

import { FFTError, FFTErrorCode } from 'webgpu-fft';

try {
  await engine.fft(invalidData);
} catch (error) {
  if (error instanceof FFTError) {
    console.error(`FFT Error [${error.code}]: ${error.message}`);
    if (error.code === FFTErrorCode.INVALID_INPUT_SIZE) {
      console.error('Input size must be a power of 2');
    }
  }
}

Next Steps

• Explore specific API pages from the sidebar
• See the Quick Start Guide for getting started
• Read the Architecture Overview for internal design