A free, full-featured web-based audio and waveform editor that runs entirely in the browser — no plugins, no installs, just open and edit.
AudioMass is a browser-native audio editing tool that brings professional-grade waveform editing capabilities to any modern web browser without requiring software installation. It handles the full editing workflow: load audio from disk or record directly from a microphone, manipulate waveforms with precision cut/copy/paste operations, apply real-time audio effects, and export to your desired format — all using native Web Audio API and WebAssembly.
The editor supports an extensive range of audio formats for both import and export, including MP3 (via bundled LAME encoder), FLAC, WAV, OGG, AAC, AIFF, OPUS, and WebM. This breadth of format support, combined with the ability to encode entirely client-side using WebAssembly-compiled codecs, means no audio data ever leaves your browser unless you explicitly save it.
AudioMass has evolved well beyond a simple waveform editor. It now ships with a fully featured multitrack mode that lets you layer multiple audio tracks, drag and reposition clips on a timeline, apply crossfades where clips overlap, arm individual channels for live recording, and bounce the composite result down to a single file. The multitrack view supports beat/BPM snapping, region selection, zoom controls, and a channel strip with per-track volume and pan controls.
Under the hood, the application includes professional audio processing capabilities: a parametric equalizer with automation curves, a loudness meter implementing the LUFS standard (ITU-R BS.1770), a tempo estimator, an RNN-based noise suppression module via WebAssembly, ID3 tag reading and writing, and a proprietary session format (.amss) that preserves the full editor state — including multitrack arrangements — for later recall.
Architecture AudioMass follows a modular, event-driven architecture built around a central application object that acts as a mediator. All modules — the audio engine, UI, state manager, multitrack controller, recorder, effects chain, and keyboard handler — communicate exclusively through a lightweight publish/subscribe event bus attached to the application instance. This decoupling means each module registers listeners and fires named events without direct references to other modules, keeping cross-cutting concerns separated. The multitrack layer adds a propagation layer that intercepts certain request events before they reach the main engine, allowing it to override playback and recording behavior without forking the core event loop. The result is a genuinely layered architecture where the audio engine, UI rendering, and state history can each evolve independently.
Tech Stack The entire application is written in vanilla JavaScript (ES5-compatible) with no frontend framework or build-time transpilation required. Audio rendering uses the WaveSurfer.js library as the waveform visualization backbone, extended with a regions plugin for selection handling. The Web Audio API drives all real-time audio routing, playback, and recording via AudioWorklet nodes and ScriptProcessor fallbacks. Format codecs are delivered as WebAssembly binaries: LAME for MP3 encoding, libflac for FLAC, and a compiled RNN model for noise suppression. The server component is minimal — a single-file Go or Python static file server — reflecting that all meaningful computation happens in the browser. Deployment targets include static hosting and PWA installation via service worker and Web App Manifest.
Code Quality The codebase uses immediately-invoked function expressions (IIFEs) throughout for module encapsulation, preventing global namespace pollution in the absence of a module bundler. Code is consistently structured with private function declarations inside closures, a uniform event system, and explicit error handling surfaced through modal dialogs rather than silent failures. There are no automated test files in the repository — correctness is validated manually and through real-world browser usage. Type safety relies on JavaScript conventions rather than TypeScript, and there is no linter or CI configuration present. Comment density is moderate, with meaningful inline comments on non-obvious audio math (LUFS biquad filters, zero-crossing snap, tempo flux scoring) but sparse documentation elsewhere.
What Makes It Unique AudioMass occupies an unusual position as a fully featured, zero-dependency-on-the-user audio workstation that runs in a browser tab. Most web audio tools are either simple trimmers or wrappers around server-side FFmpeg. AudioMass compiles the entire codec stack — MP3, FLAC, noise suppression — to WebAssembly and runs it entirely on the client, meaning audio never touches a server. The multitrack mode with crossfade detection, beat-snap grids, and a proprietary binary session format (.amss with LZMA compression) goes well beyond what any comparable open-source browser tool offers. The LUFS loudness meter — a full ITU-R BS.1770 implementation with true-peak detection — is professional-grade functionality rarely found outside paid desktop DAWs, let alone browser apps.
AudioMass is released under the MIT License, which is one of the most permissive open-source licenses available. You are free to use it commercially, modify the source, redistribute it, and even embed it in proprietary products without any obligation to open-source your own code. The only requirement is that the original MIT copyright notice is retained in copies or substantial portions of the software. There are no copyleft restrictions, no contributor license agreements, and no open-core limitations — the entire feature set in the repository is available under the same license.
Self-hosting AudioMass is remarkably lightweight by web application standards. The application is a collection of static files — HTML, JavaScript, WebAssembly binaries, and CSS — that can be served from any static file host, CDN, object storage bucket, or a single-command Go or Python server. There is no database, no backend API, no authentication layer, and no runtime dependency beyond a web server capable of serving files with correct MIME types. Operational burden is minimal: updates consist of pulling new commits and re-deploying static files, there is nothing to patch at the server level, and the PWA service worker handles offline caching automatically for end users.
Because AudioMass is a pure client-side application, there is no managed cloud version with SLAs, enterprise support contracts, or high-availability guarantees from the author. What you gain from self-hosting — full data privacy, since audio never leaves the user’s browser — you trade for the absence of centralized support, automatic updates pushed to users, or a vendor to call when something breaks. Organizations that need guaranteed uptime or compliance certifications will need to manage those concerns themselves, including CDN configuration, TLS termination, and monitoring of the static hosting layer.
No Code Platforms · AI Development · Developer Tools
Visual LLM workflow platform with RAG pipelines, agent capabilities, and model management for building production AI applications.
Developer Tools · Game Development · Design Tools
Free, MIT-licensed 2D and 3D game engine with one-click multi-platform export and no royalties.
Design Tools · Productivity
The offline-first diagram editor that puts security above all else — no account, no data leaving your machine, no compromises.