The timeline, the monitor,the cut, the export.

The timeline

Every project has ten video tracks and ten audio tracks. V1 is the lowest video track, right above the audio tracks; A1 is the highest audio track. The tracks the guide named for your kind of video are at the top of each group, the others arrive collapsed: one click on their header expands them. A "Music" track is orange, a title track purple, a voice track green: the color says the role.

The gestures are the ones you know: drag a clip from the bin onto a track, cut with the razor, move, trim by the edges, snap to neighboring cuts, fade by pulling the corner of a clip. On a trackpad, two fingers scroll the timeline without moving the playhead, pinching zooms it around the pointer (⌘ + two fingers with a mouse). The "Auto cut" button is in the timeline toolbar, next to the editing tools.

Every gesture is one undo step, and an agent's turn is a single one: ⌘Z undoes everything an agent just did, in one go.

The guide: tracks, pace and cuts by kind of video (in French)

The monitor

The Program monitor shows the sequence, the Source monitor shows a clip before you place it. The zoom percentage is written in the monitor bar, in figures: 100% means one project pixel for one screen pixel, and "Fit" (⌘0) shows the whole image. The zoom is kept per monitor for the session.

Pinch with two fingers

Continuous zoom, the image stays under your fingers.

⌘ + wheel

Zoom in 20% steps with a mouse, continuous on a trackpad, around the pointer.

Two fingers

Moves the image when it overflows the panel. Without a modifier, the mouse wheel moves the playhead.

Space + drag

The hand tool. Space alone starts or stops playback, as usual.

Click on the percentage

Fit, 25, 50, 100, 200, 400%.

Double-click

Full screen.

Move an element directly on the image

Select a shot in the timeline: its frame appears in the monitor, with corner handles for scale, a rotation handle, and drag to move it. No button to find, no inspector to open. As long as you move nothing, the shot carries no effect and costs nothing to render: the first drag turns the transform on, and the inspector opens on it.

Keyframes are automatic: moving the element at a time where there is no keyframe sets one. The diamond in the monitor bar shows it (blue when active) and lets you turn this behavior off. The figures stay editable in the inspector. Turning on and moving fit in a single undo step.

While dragging, the element snaps to the center of the image, to its edges, to the thirds and to your guides, within 6 px. A pink line shows the target.

Rulers, guides and platform overlays

⇧R shows two rulers graduated in project pixels, at the top and left of the monitor. Dragging from a ruler sets a guide; you move it (its position shows), you drag it back onto the ruler to remove it, a double-click lets you type its position. Guides are kept in the project, sequence by sequence.

The Guides and overlays menu (the grid button in the monitor bar) holds the classic overlays (center, thirds, diagonals, safe area, grid), the rulers, showing and locking the guides, and the platform overlays: TikTok, Instagram Reels, YouTube Shorts, Instagram and Facebook Stories, WhatsApp Status, YouTube (end screen), Facebook, LinkedIn. Each one greys out, with its label, the areas the platform's interface covers (the button column on the right, the caption at the bottom, the top bar, YouTube's end screen over the last twenty seconds), and draws the safe area in the middle.

If your project does not have the platform's ratio (a 16:9 with TikTok), the 9:16 frame centers itself in the image and what overflows darkens. When the project has a kind of video, the target platform is marked in the submenu, and the status bar reminds you the first time.

Auto cut

For a podcast, an interview or a talking-head video: Sequence > Auto cut… (also in the ⌘K palette and through the "Auto cut" button of the timeline). The window has two tabs. "Sources": the voice tracks, syncing microphones and cameras by sound, and picking the camera by who is speaking. "What to look for": silences, filler words ("um", hesitations), redone takes.

Nothing is cut without being seen. The default action is "Mark": every planned cut becomes a ranged marker in the timeline, red for a silence, orange for a filler word, purple for a repeat. You listen, you remove the marks you want to keep, then "Cut what is marked" applies the rest, on every unlocked track at once so that picture and sound stay in sync. You can also cut and close directly, cut and leave the gap, or mute.

The starting settings (automatic threshold measured on the background noise, minimum silence 0.5 s, margins of 0.1 and 0.15 s around speech) can be changed in the window. Filler words and repeats need a transcript: it is made on your machine, with a Whisper model the window offers to download the first time (145 MB). The whole operation is a single undo step.

The export, prepared for the kind of video

File > Export… (⌘⇧E) opens the export window; "Export again" (⌘⌥E) reruns the last preset without reopening it. When the project has a kind of video, it arrives already set: the preset the guide recommends (MP4 H.264, or H.265 when the kind calls for it), and the "Normalize loudness to … LUFS" box checked on the platform's target: -14 for YouTube and social networks, -16 for podcasts and training, -23 for television. You can change everything; your choice is kept in the project.

Normalization is done on the final mix, in a single pass, which matters on a modest machine: the result lands within one unit of the target, inside the platforms' tolerance. For the exact LUFS to the tenth, a true two-pass will come later.

Platform specs in the guide (in French)The Motion module: animated titles and graphics