Editing & color grading

The Edit page is a multi-track timeline editor: media / transcript / subtitles on the left, dual monitors in the middle, the timeline at the bottom, properties / grading on the right.

Create the main timeline and import media — the Import button sits at the top right of the media pool; video / audio / images all work, with preview proxies and thumbnails generated automatically:

Drag clips onto the track, move the playhead, press S to split at the playhead:

Timeline#

• Multiple tracks: video, audio and subtitle tracks side by side; tracks can be muted / locked.
• Tools: select / blade (split); cross-track drag with snap-to landing.
• Editing: split, duplicate, ripple delete, multi-select, speed ramps, fade in/out, picture-in-picture.
• Playback: frame stepping, loop, playback rate, volume, fullscreen; the monitor shows waveforms.
• Undo / redo: everything is undoable (Cmd/Ctrl+Z).

Transcript-driven editing#

Run ASR on an audio/video asset to get a word-level transcript:

• Deleting sentences / words cuts the corresponding segments from the source; silences and filler words ("uh", "you know") can be detected and batch-removed in one click.
• Each sentence keeps its timestamp, speaker and first text line aligned. Long text wraps in full; it is never ellipsized, and row actions do not reserve empty space while hidden.
• Transcription runs in an external interpreter; models download on first use (see Download & install).

AI assistant#

The robot button opens the same agent session pool used by AI Studio, workflows and boards. In the editor it docks as a real right-hand column by default, so the timeline and monitor resize instead of being covered; switch it to floating mode when you want an overlay. The top-left title is the current conversation — click it to search or switch sessions.

Color grading#

The Grading tab on the right:

• Curves: Luma / R / G / B channel curves with an independent undo stack (each drag is one step).
• Style presets: vivid / black & white / warm / cool / cinematic / faded.
• 3D LUT: upload a .cube to apply.
• Scopes: histogram / waveform, live.

Subtitles & filters#

• Add a subtitle at the playhead on the subtitle track and edit the text in place, or generate the whole track from the transcript.
• Filters and subtitles are burned into the exported cut.

Translation#

"Translate" works on the whole track or only the selected cues; tick "keep the original" to get a two-line "source / translation" cue. Two engines: Google is free and needs no setup, but it translates cue by cue with no context; the AI model uses the provider configured for this workspace and reads the sentence. Google is the default — it costs nothing and needs no key.

Dubbing#

Every cue carries its own dub button, and you can dub a batch. The audio lands on a dedicated dub track: the original audio is untouched, delete the track and you are back where you started, and dubbing again returns to that same track rather than stacking up a pile of one-clip tracks.

• Fit to cue length (off by default): stretch or squeeze the voiceover to that cue's duration. It uses the clip's own speed, so the render applies atempo — lossless, undoable, still adjustable in the inspector afterwards. Past ±20% the pacing starts to sound off, so whether it is worth it is yours to decide per clip.
• Bilingual cues ask which line to read. Feeding the whole cue to TTS reads the source and then the translation — a 3-second cue turns into twelve.
• Engine is either local voice cloning or a remote engine (Edge / OpenAI / Volcano). For remote engines the voice list is fetched per engine — Volcano's catalogue follows your account.

When the language does not match, the engine does not error — it reads the text with the rules it knows, produces gibberish, and reports success. So it is caught at the moment you pick the engine, with a way out, instead of after you listen to it.

Reading other languages in your own voice#

What local cloning can read depends on which weights are installed, not on the engine. The base weights are Chinese + English; Japanese, French, German, Spanish, Italian, Russian, Hindi, Arabic and Finnish each have a community finetune, and whichever is missing gets a download button right in the dub popover (~1.4 GB each).

Japanese, Korean, Russian, Arabic and Hindi are recognised from the script itself — install the weights and they just work. French, German, Spanish, Italian and Finnish share the Latin alphabet with English: no character can prove "this is French rather than English", so those are picked explicitly in the "Weights" dropdown.

Export#

Export on the timeline toolbar renders the current timeline into a new asset; once done you can head to Publishing.