Slowed & Reverb Generator

The Culture of Slowed + Reverb & Nightcore Aesthetics

In recent years, the internet has witnessed a massive resurgence in altered-speed audio trends. The most prominent of these are **Slowed + Reverb** and **Nightcore**.

The slowed aesthetic traces its roots back to the late 1990s and early 2000s hip-hop culture, pioneered by DJ Screw in Houston, Texas. His technique of "chopping and screwing" involved slowing down vinyl records and layering them, creating a heavy, laid-back vibe. In the modern era, this evolved on YouTube and TikTok into "Slowed + Reverb," where pop, rock, and indie tracks are slowed down to roughly 85% speed, paired with a massive reverb tail. This modification alters the emotional tone of a track, making it feel melancholic, atmospheric, and nostalgic.

Conversely, **Nightcore** represents the opposite end of the spectrum. Originating in the early 2000s from a Norwegian DJ duo, Nightcore speeds up audio tracks (usually by 20% to 30%) and increases the pitch. This gives vocals a distinct high-pitched "anime" quality and boosts the energy of dance, electronic, and pop tracks, transforming them into high-tempo anthems.

Technical breakdown of Pitch Shifting and Reverb Convolution

Building a real-time sped-up and slowed generator inside a web browser requires leveraging the advanced capabilities of the **Web Audio API**.

In digital audio processing, speed and pitch are fundamentally linked. When you speed up playback of an audio buffer, the waveform samples are read faster, which naturally increases the pitch (vocal frequency). This is controlled via the playbackRate parameter on an AudioBufferSourceNode. To allow separate tuning, we provide a detune input. Detuning is measured in **cents**, where 100 cents equal one semitone (a single key step on a piano). By shifting detuning up or down, we adjust the playback frequency of the audio samples to fine-tune the key of the track.

**Digital Reverb** is achieved using a ConvolverNode. Convolution is a mathematical process that combines two signals: the input audio signal and an **Impulse Response (IR)**. The Impulse Response represents the acoustic characteristics of a specific space—such as a cathedral, recording studio, or cave. Instead of loading heavy external IR audio files, our engine dynamically generates a high-fidelity synthetic impulse response. We generate high-frequency white noise and apply an exponential mathematical decay formula:
y(t) = noise * e^(-t / decay_constant)
By convolving this noise envelope with your track, we create a beautiful, rich, and spacious stereo echo that blends with the original sound.

Privacy and Performance of Client-Side Processing

Processing stereo audio tracks at 44.1kHz requires processing millions of data points per second. Traditional web platforms handle this server-side because they rely on heavy command-line utilities. However, our client-side architecture decodes the compressed MP3 or WAV file directly into a raw array of floats in the browser memory. The modified audio is then processed through an OfflineAudioContext, which renders the calculations as fast as your device's CPU allows.

Because the processing runs entirely in your local browser sandbox, your personal tracks are never uploaded to our servers. This ensures 100% privacy, eliminates network upload times, and allows us to offer unlimited high-speed processing with zero costs.