Every feature, step by step

Silence, hesitation and dead air are removed on your own computer, before anything is sent. On real material, 52 s of recording yielded 23 usable seconds from this alone.

1. Open the video. The cut is already done — the preview plays without the silences.
2. Drag the Cut breaths slider left for a tighter cut, right to respect the speaker's breathing.
3. At 0.00 s it cuts every pause there is: the wall-to-wall pace social feeds like.
4. Listen before moving on. Every removed take is in the list and comes back with one click.

The AI reads the transcript and unticks false starts, repeated lines and off-topic chatter. It suggests; you decide.

1. Turn on Only what is worth it, on the left.
2. Only the audio is sent to be read — the video stays on your computer.
3. The list now shows each take ticked or unticked. Click to bring back what you want to keep.
4. Changed the breath cut afterwards? Re-read the audio brings the reading up to date.

Speech reading gets proper nouns wrong — yours, the company's, the product's. Fix it once and it holds for the whole video.

One recording becomes 9:16, 4:5, 1:1 and 16:9. The crop follows whoever is speaking, one framing per take — and it switches on the cut, where it does not show.

1. Turn on Reframe automatically, at the top of the Improvements tab. It comes on by default.
2. Click a format chip under the player.
3. Click the chip you already picked to open that format's settings: fill the screen or keep bars, and what goes in the bars.
4. Each format keeps its own settings. Going from 9:16 to 16:9 does not undo what you did in the other.

White balance, levels and contrast measured on your own footage. If the edit has two cameras, they are matched before any look goes on top.

1. Go to the Colour tab on the right and turn on Treat the colours.
2. Pick the contrast — Soft, Balanced or Strong — and the intensity.
3. Hold See the original, below the player, to compare on the same frame.

Blur the background or push it back. The installed app uses a much heavier model for this.

The three finishing switches, all on the Colour tab. Skin smooths without erasing the face; Vignette darkens the edges to pull the eye in; Grain gives texture back to an image that is too clean.

Straight from the speech, already in sync, with ready-made styles. On vertical video they are what holds people watching without sound.

Name and role of whoever is speaking, coming in and out on its own at the start.

One sentence of what was said, big on screen, exactly when it is spoken. You pick it in the text, not on the clock.

The end card with your logo, your name and the invitation to follow.

Background noise reduced, every scene at the same level and peaks safely under the ceiling. It is what separates “shot on a phone” from “shot properly”.

1. Turn on Treat the voice and open Settings.
2. Choose how much noise to remove and the loudness ceiling for where the video is going — social asks for louder.
3. Click Hear the treated voice and compare with the original.

Generated at the length of your edit — no cutting mid-bar — and it steps out of the way on its own while there is speech.

1. Turn on Soundtrack and open Settings.
2. Pick Style, Mood, Where it will run and the tempo in BPM.
3. Click Generate soundtrack and wait. Not right? Redo.
4. Set the volume and, under Music under the voice, how far it ducks.
5. Hear the mix checks voice and music together before you export.

An MP4 ready to post, in the format you picked on the chips.

1. Watch the whole preview. What is in it is what comes out.
2. Click Export video and confirm.
3. Keep this tab open and visible until it finishes — in the background the browser holds the work.
4. Need all four formats? Export one, switch the chip and export again: each format kept its own settings.

When someone else finishes the job: out goes the timeline, not the flattened video.

A project stores the decisions, not the video — that is why it is light, and why the file has to stay where it is.

Your own finished artwork on top of the video: an animated logo, a lower third you designed, anything with a transparent background.

1. Turn on Pre-render and click Choose file.
2. Send a render with alpha — ProRes 4444 or WebM with transparency.
3. Give the start and the duration. The preview uses a light version; the export uses the original file.