This ambient audio generator synthesizes continuous sound masking noise directly in your web browser using the native Web Audio API. Rather than looping static MP3 recordings with noticeable repetitive patterns, the generator uses algorithmic mathematical filtering to produce endless, non-repeating acoustic noise across three classic spectral colors.

### Spectral Noise Colors & Acoustic Filtering The audio engine creates random Gaussian noise values and filters them according to specific spectral density curves:

- **Brown Noise (Brownian / Red Noise):** Implements an integrated low-pass filter: $y[i] = \frac{y[i-1] + 0.02 \times x[i]}{1.02}$. Power drops by $6\text{ dB}$ per octave with increasing frequency. This creates a deep, warm low-frequency rumble resembling distant ocean surf, heavy rain, or an airplane cabin, ideal for deep focus and ADHD sound masking. - **Pink Noise (1/f Noise):** Implements proportional filtering: $y[i] = \frac{y[i-1] + 0.05 \times x[i]}{1.05}$. Power drops by $3\text{ dB}$ per octave, delivering equal acoustic energy per octave band. Pink noise mirrors natural soundscapes like steady waterfalls and is clinically proven to improve sleep quality. - **White Noise (Flat Spectrum):** Uniform random distribution across all audible frequencies ($20\text{ Hz}$ to $20,000\text{ Hz}$). White noise creates a crisp, rushing hiss effective at masking sudden sharp background noises in busy office environments.

### Playback Mechanics & Gain Controls Audio buffers are allocated via `AudioContext.createBuffer` and connected to a looping `AudioBufferSourceNode` with a real-time `GainNode` volume control. The synthesized loop runs smoothly without memory leaks or buffer glitches.

### Client Synthesis Guarantee All noise textures are computed locally in your computer's audio hardware via the Web Audio API. Zero audio streams, tracking telemetry, or bandwidth usage are routed through external servers.