Multitrack editor
Multitrack editing in the browser, with one track per speaker as soon as the session ends. Eleven effects per track, each with its visual editor.
The situation
After a session with several people, editing often starts with sorting: downloading each track, importing it into an application installed on a single computer, lining it up against the others. The project then stays on that machine, and whoever reviews it waits for an export to listen.
Tracks from the start
At the end of a recording in the live studio, the editor opens the episode with one track per speaker, plus one track for the sounds played during the session. Nothing needs importing or lining up.
Clips are moved, cut and resized with the mouse, with keyboard shortcuts to split, copy, cut and paste. Each clip carries its fade in and fade out, with their curve, and its own gain. Each track has its volume, pan, mute and solo. Normalising a clip only affects the portion it covers: one half can be raised without touching the other. Every operation can be undone and redone.
Eleven effects per track
Filter, EQ, noise gate, compressor, gain, echo, reverb, limiter, soft clipper, de-esser and stereo width. Each track takes its own chain, and the master bus takes the same set.
Each effect has its visual editor: the filter and EQ curve is visible while it is set, and the compressor's gain reduction reads live. Starting settings come with each effect, for the filter for instance: rumble cut, clean voice, hiss reduction, warmth, telephone, voice isolation.
All eleven effects and the master bus are open on every plan, the free one included.
Preview and render
Preview plays in the browser. The final render runs on the server, with the same effects in the same order: what is heard at the edit is what comes out in the file. The computer stays free during the render, and an e-mail announces the file once it is ready.
The render comes out as an MP3, with title, show, season, number and cover written into the file. Loudness is set to -16 LUFS, the level Apple Podcasts recommends, without peaks clipping.
Source versions
A source can be cleaned of room echo, or split into one source per speaker when several voices share one file. Each treatment creates a new version, and the original stays available and can be restored. Details on the AI post-production page.
Team editing
The project saves automatically during editing. A lock prevents two people from writing to the same episode at once. A reviewer opens the project and listens without being able to change it, and each person gets back their own position in the timeline. Team work is described on the media team page.
Still missing
No volume automation curves: level is set per clip and per track. The render only produces MP3, not WAV. The editor does not open on a phone. The number of tracks and renders per month depends on the plan.