A camera layer is a virtual camera pointed at your composition. Instead of animating zoom and pan on one clip at a time, you place a camera clip on the timeline and everything under it moves together — footage, text, shapes, mockups, sequences.
It behaves like a real camera in one important way: the project background stays put. The backdrop never slides out from behind your layers when the camera pushes in or pans.
Add a camera
Add Channel ▾ → Add camera channel. With layers selected the entry reads Add camera over selection — the camera clip spans those layers and lands right above them, the same way an adjustment layer does.
A project has one camera channel, and it is pinned to the top of the stack — it can’t be dragged down and nothing can be dropped above it, because a layer rendered over the camera would be framing the camera rather than being framed by it. Delete the clips (or the channel) to remove it. The row is drawn shorter and in a muted sand so it reads as chrome rather than another content layer.
Adding a camera again drops another camera clip onto that same channel, placed after the last one if the playhead is already inside a move — so you can chain a push-in, a hold and a pull-back along the timeline.
While the playhead is outside every camera clip the camera is neutral and nothing is transformed.
Presets
Select a camera clip and the right panel opens the Camera inspector. The Presets dropdown holds one-click moves, each with a live thumbnail showing a viewport travelling over a stand-in scene. They are grouped by intent:
Cinematic — slow, composed moves.
| Preset | Length | Move |
|---|---|---|
| Quiet Push | 6 s | A restrained forward creep that adds attention without announcing itself |
| Wide Reveal | 5.5 s | Begins close and slightly low, then opens cleanly onto the full composition |
| Lateral Glide | 6 s | A polished slider move with just enough yaw to reveal depth between layers |
| Parallax Arc | 7 s | An arcing move that lets foreground and background layers separate naturally |
| Crane Rise | 6 s | Rises through the composition while easing into a wider, level view |
Editorial — short, punchy beats.
| Preset | Length | Move |
|---|---|---|
| Soft Punch | 1.6 s | A quick emphasis beat with a controlled overshoot and a clean settle |
| Whip & Land | 1.15 s | A real reframing whip that accelerates late and settles on a new composition |
| Handheld Breath | 8 s | Small, irregular operator drift without the synthetic wobble |
| Documentary Follow | 5 s | A gently imperfect follow move that keeps finding the subject |
Focus — moves built around depth of field; each one animates focus, so they only read once DOF is on.
| Preset | Length | Move |
|---|---|---|
| Rack to Near | 3.5 s | Holds the background, then pulls focus decisively onto a near subject |
| Rack to Far | 3.5 s | Transfers attention from a close detail to the wider scene behind it |
| Focus Chase | 6 s | Moves attention across three depth beats for layered product or title layouts |
| Macro Bloom | 4.5 s | Pushes into a detail while focus and highlight bloom tighten around it |
Applying a preset also resizes the clip to the preset’s natural length, so a 1.4 s whip pan stays a whip pan instead of being stretched over a 10 s clip. Drag the clip edges afterwards to retime it.
Camera properties
The Camera Keyframes section is the same keyframe editor used everywhere else in the app — one independent track per property, with magnetic snapping, multi-select and the full easing list. See Keyframe Builder.
| Property | Unit | What it does |
|---|---|---|
| Zoom | × | Lens zoom about the frame centre. Every layer magnifies by the same amount, whatever its depth. |
| Pan X | % of frame width | Positive pans the camera right, so content slides left. |
| Pan Y | % of frame height | Positive pans the camera down, so content slides up. |
| Roll | degrees | Cants the horizon — a dutch angle. |
| Dolly | px | Moves the viewpoint in and out. Unlike Zoom this does change perspective, so layers at different depths separate. |
| Tilt X / Tilt Y | degrees | Pitches and yaws the camera for a 3D feel. |
Zoom and Dolly are deliberately separate. Zoom is a lens zoom (no parallax change); Dolly is a viewpoint move. Animating them in opposite directions is the classic vertigo effect — the subject holds its size while the background compresses.
Depth of field
The camera owns focus. Turn Depth of Field on in the camera inspector and every layer is blurred by its own circle of confusion — how far it sits from the focus plane — exactly the way a real lens behaves. A layer at the focus distance is sharp; layers in front of and behind it fall away.
This is why DOF lives on the camera and not on individual layers: focus is a property of the lens, and one camera move should be able to rack focus across the whole stack.
| Control | Unit | What it does |
|---|---|---|
| Focus distance | px | Distance from the camera to the sharp plane. |
| Focus point X / Y | % | Where the Set from preview picker last sampled — and, with a focus area, the centre of the sharp region. |
| Focus area width | % | 0 leaves the focus point a pure depth probe. Above 0 it also holds a region of the frame sharp: crisp inside, easing to the layer’s full blur by the edge of the area. |
| Aperture | px | Bigger = shallower depth of field. The panel shows the equivalent f-stop. |
| Blur level | % | Scales the computed blur — the artistic override on top of the physical result. |
| Highlight gain | % | Makes bright spots bloom into bokeh discs instead of dissolving to grey. This is most of what separates “cinematic” from “blurry”. |
Every one of these is keyframable — click the diamond beside a field to promote it into Camera Keyframes, then rack focus with the same easing curves as any other animation.
Set focus from the preview
Click Set from preview, then click a layer in the canvas. The camera measures that layer’s real distance and sets Focus distance to it — what a focus puller does, rather than guessing a number. If the focus tracks are already keyframed, the pick writes a key at the playhead instead of changing the base value.
Focus point vs focus area
They work together:
- Width 0 — the point is only where you picked. Blur is decided entirely by depth, so a layer is uniformly sharp or uniformly soft.
- Width above 0 — the point becomes the centre of a sharp region. Use this when a subject sits on the same plane as things you want soft: a face in a wide shot, one card in a row, a product against its own backdrop.
Both stay put through camera moves, so a locked-off focus area holds while the camera pushes in.
Keeping a layer out of the camera
Some layers should never move: a watermark, a fixed caption, a UI frame. Once a camera exists, every other channel header gains a small camera button — click it to lock that channel to the frame. It renders exactly as if there were no camera at all.
Saving your own moves
Under Save this move, name the current camera setup and it is stored as a reusable preset (globally, not per board) under the media folder’s camera-presets. Saved presets appear in the Presets dropdown under a Saved group and can be applied to any project, or deleted from the same list.
A preset captures both halves of the setup: the keyframed motion and the un-keyframed camera settings — aperture, focus distance, focus area, highlight gain, and whether depth of field is on. Since a DOF rig is usually dialled in as static values rather than animated, saving only the motion would lose the look. That also means a preset can be pure optics with no movement at all: switch DOF on, dial an aperture, save, and you have a reusable lens.
Auto-zoom
Screen recordings build their smart zoom as a camera clip on this channel. It is a normal camera move — swap the preset, edit the keyframes, retime it, or delete it.
Motion blur
A fast camera move (a whip pan, a snap punch) gets true accumulation blur when the project’s Blur switch is on. Blur is applied at render time only, so the preview stays responsive — see Rendering and Output Formats.
Editing layers while the camera moves
The preview’s transform gizmo follows the camera, so you can still click and drag a layer with a zoom or pan active. While Dolly or Tilt are engaged the on-screen shape of a layer is a true perspective quad that a rectangular gizmo cannot represent, so the gizmo hides itself rather than point at the wrong place — scrub to a moment where dolly and tilt are neutral, or use the inspector fields.
Leave a Reply