Appearance
Transform 3D
Moves whatever comes before it in 3D, seen through the comp's camera: a flat picture as a card, points, shapes and text themselves, or a picture of things that are already in 3D.
Category: 3D Menu path: 3D > Transform 3D
Ports
| Port | Type | Direction | Description |
|---|---|---|---|
in | imageRgba16f / shape / points / text | input | The picture, points, shapes or text to move. Like Transform2D, both ends take the type of what is wired in |
out | same as in | output | The picture where this node puts it, or the moved points or shapes |
position_in | vec3 | input | Drives Position (shown with E) |
scale_in | vec3 | input | Drives Scale (shown with E) |
rotation_in | vec3 | input | Drives Rotation (shown with E) |
orientation_in | vec3 | input | Drives Orientation (shown with E) |
A Vec2 wired into a 3D input lands on the z = 0 plane; a single number sets all three axes.
Parameters
The same rows as Transform2D, with depth added.
| Param | Type | Default | Description |
|---|---|---|---|
position | vec3 | 0, 0, 0 | Moves the card, in comp pixels. +z is away from the viewer |
scale | vec3 | 1, 1, 1 | Scales the card around the anchor. The link button sets all three together |
rotation | vec3 | 0, 0, 0 | Turns the card about X, Y and Z (degrees). Z is Transform 2D's Rotation |
orientation | vec3 | 0, 0, 0 | Turns the card (degrees), applied after the rotations |
anchor | vec3 | 0, 0, 0 | The point the card turns and scales around, in pixels from the center of the layer's content (or of the comp — see Anchor Space). Z sets it in front of or behind the card. The 9-point picker moves it to a corner, edge or the center without moving the card |
anchorSpace | string | Content | What Anchor is measured from: the center of the layer's content, or of the whole comp |
How It Works
A Transform 3D moves whatever comes before it, in 3D space. It never flattens anything, so Transform 3D nodes stack the way Transform 2D nodes do, on either side of a Draw node or a Merge:
| Before the Transform 3D | What it does |
|---|---|
| A flat picture (an image, a 2D shape drawn, a nested comp) | Makes it a card: the whole picture on a flat plane |
| Points, shapes or text | Moves them, and the Draw node after it draws them in depth |
| A picture of things already in 3D (after Draw Points on 3D dots, after a Merge of 3D things) | Moves those things, depth and all: a ball of dots turns as a ball |
| Another Transform 3D | Moves what that one placed, further |
A card is drawn by the comp's camera with true depth, so cards that cross cut through each other. With every value at its default nothing moves: a card sits exactly where the 2D layer was.
After a Draw node or a Merge
After Draw Points, Draw Shape, Draw Text or Draw Instances, a Transform 3D moves the drawn things themselves. Dots keep their depth, shrink with distance and cross other 3D things correctly, and a camera can still fly through them. A scale grows what was drawn, dots and strokes included, as scaling a card grows its picture.
After a Merge (or a Mix), it moves everything that went into it together, and the inputs still cross by depth. A 2D input of that Merge becomes a card at z = 0 and moves with the rest, so it can now pass in front of or behind the 3D inputs.
One picture wired into two Transform 3D nodes is two things, each where its own node puts it.
For a flat picture of something 3D (a turning screen that shows a ball), put the 3D thing in its own comp and bring that comp in: a nested comp arrives flat.
Points and shapes
On points or shapes, Transform 3D moves the geometry itself, the way Transform2D does in 2D: each position is moved in 3D, and its depth is kept in the z attribute (shown in Details; in a layer with a Transform 3D on its geometry, PointAttributes can set it like any attribute). Draw Shape and Draw Points then draw that geometry through the camera, at its true depth, so it crosses cards and other 3D geometry correctly and shrinks with distance.
As in 2D, it moves positions, not the things you later place on them:
- Grid → Transform 3D → Clone to Points: the grid's layout turns; each clone stays upright, facing front.
- Grid → Clone to Points → Transform 3D: the clones turn with the grid.
Points that know which way they face — a normal attribute, from a Grid's Sphere pattern or a Point Scatter on a model — have that facing turned along with their positions, so clones and images laid on a surface stay on it either way round.
A shape that turns out of the screen's plane has its curves made into short straight segments, fine enough to look smooth.
Text
Text stays text: Transform 3D places it in 3D and leaves the glyphs flat in their own plane, so TextAnimate still animates them, and a Transform 2D after it moves the text within that plane, as on a card. DrawText draws it through the camera with true depth, sharp at any distance; TextToShape gives its outlines already placed in 3D.
Effects
Where an effect sits in the chain decides what it works on:
- Before a flat picture's first Transform 3D, it changes the picture on the card and turns with it.
- After it, it works on what the camera sees, flat on the screen: a glow stays round and a blur runs across the screen however the card is turned. The picture still sits at its depth, so it crosses other 3D things correctly; anything the effect spreads past the card's edge sits on the card's plane.
- A Transform 3D after that effect still moves the card. The effect keeps working on the camera's view, so adding one (a Parent Transform, say) changes nothing about the effects ahead of it.
- On things with depth (3D dots, shapes, a model), an effect always works on what the camera sees of them, wherever the Transform 3D sits. A Glow on a ball of dots keeps its size in pixels as the ball moves away.
- A Transform 2D right after a card's Transform 3D moves the card in its own plane. One after an effect, or after a Draw node on 3D geometry, moves the flat result.
Usage Examples
Basic: tilt a title
Text → Draw Text → Transform 3D → Output, then key Y Rotation from -90 to 0 for a card that swings into view.
Creative: a hallway of images
Give several image layers each a Transform 3D at a different Z and animate a Camera layer moving forward through them.
Tips
- Consecutive 3D layers draw together through the camera; a 2D layer between them splits them into separate groups, drawn in layer order.
- A Transform 3D on a branch into a Merge keeps its depth on the way in, so the Merge crosses it with its other 3D inputs (see 3D).
- Node Preview has no camera: it shows a picture's Transform 3D unturned.
- A Transform 3D inside a Loop or Solver zone treats a picture that comes into the zone as flat.
- Blend modes on a 3D layer are ignored for now; cards draw Normal.
- An effect after the Transform 3D sees only this layer, within the comp's frame: it can't spread content from outside the frame back into view.
- The layer's 3D switch bypasses its Transform 3D nodes rather than deleting them, so switching back on restores every value and key.
- To animate one axis on its own (Z Position on its own curve, say), click the split button at the end of the Position, Scale, Rotation or Orientation row. It wires a MakeVector named after the row into the Transform 3D and moves any keyframes onto its X, Y and Z, which show in the Graph Editor in red, green and blue. Click the button again to merge them back.
- A Transform 3D with a value wired in has no handles in the Viewer (a drag would change values its wires override), unless the wire is a split row from the button above. Parenting uses this: the Parent Transform just before Output is driven by a Parent3D node, and moves everything in the layer.
- Motion blur: turn on the layer's MB switch and the card blurs as it moves through 3D, and as the camera moves past it (the comp's shutter settings).
Related Nodes
- Camera — the comp camera that looks at every card
- Transform2D — the flat transform