Aller au contenu

Transform 3D Node

By Kriss Norde

Transform 3D applies a transformation (translation, rotation, scale) to a whole 3D scene without needing a Merge 3D. Ideal for building hierarchical parenting, offsetting an object after the fact or automatically aiming at a target with Use Target.

Overview

Transform 3D is a lightweight node that takes a 3D scene as input and applies transformations to all of its content. It differs from Merge 3D (which combines several scenes and can also apply transforms): Transform 3D does only the transformation, nothing else. More concise and readable in the graph.

Chaining several Transform 3D nodes builds a parenting hierarchy: each Transform applies its values on top of the previous one.

Where to find this node

  • Effects Library: Tools > 3D > Transform 3D
  • Quick add: right-click in the Node Editor → Add Tool > 3D > Transform 3D

Inputs

  • SceneInput (orange) - required. Accepts any 3D node or 3D scene to transform.

Setup

Shape 3D → Transform 3D → Merge 3D → Renderer 3D

Useful when you want to offset/animate an object without touching the source node (which may be shared with other branches of the graph).

Controls tab

Translation

X, Y, Z Offset (in Fusion units, default: 0)
Moves the 3D element in space.

Rotation

Rotation Order
Order in which rotations are applied on the three axes: XYZ (default), YXZ, ZXY, etc. Each order gives visually different results, especially beyond 90°.
X, Y, Z Rotation (in degrees)
Rotation around the Pivot. The behavior depends on Use Target mode:
  • Use Target off → rotation around the scene's global axes.
  • Use Target on → rotation relative to the target.

Pivot

X, Y, Z Pivot (default: 0, 0, 0)
Offset of the rotation point relative to the object's center. Essential to rotate around a corner, an external axis (wheel) or a precise point.

Scale

X, Y, Z Scale (default: 1.0)
When locked (Lock X/Y/Z), a single slider drives uniform scale. Unchecked → three independent sliders for non-uniform scaling (stretched or flattened object).

Use Target mode

One of the most useful features of Transform 3D. When Use Target is checked, a set of Target X/Y/Z sliders appears. The object automatically orients itself to face that target, wherever it is.

  • The Rotation X/Y/Z values become relative to the orientation toward the target (rotation around the "facing the target" axis).
  • The target can be connected dynamically to the position of another object through Connect To - the object follows it automatically.

Typical case: a camera that must always look at a character, or a projector that follows an object. Connect the target position to the object's position, and the Transform orients itself automatically.

Q / W / E shortcuts in the viewer

Transform 3D supports direct editing in the 3D viewer with familiar shortcuts (inspired by Maya):

  • Q - Translation tool. A 3-arrow gizmo (X/Y/Z) to move directly with the mouse.
  • W - Rotation tool. A 3-ring gizmo to rotate around each axis.
  • E - Scale tool. A gizmo with 3 cube handles to scale.

Much faster than typing numbers for rough positioning. For precision, switch to the Inspector sliders afterward.

Hierarchical parenting through chaining

Chaining several Transform 3D nodes creates a transform hierarchy. Each Transform applies on top of the previous one, in order.

Shape 3D (hand)
   ↓
Transform 3D "local"   (rotation around the wrist)
   ↓
Transform 3D "arm"     (rotation around the shoulder)
   ↓
Transform 3D "body"    (translation of the whole body)
   ↓
Merge 3D

Animating the rotation of the "arm" Transform rotates the whole arm around the shoulder, and the hand follows. Classic parent-child logic in 3D animation.

Transform 3D vs the Merge Transform tab

Merge 3D has its own Transform tab. When should you use a dedicated Transform 3D?

Merge 3D Transform tab
Built-in transforms, fewer nodes in the graph. Enough for a simple offset on a group. Be careful: if you want to animate the transform, you animate directly in the Merge, which is also the hub of the scene → it can get confusing.
Dedicated Transform 3D
More readable: the node represents only a transformation. Ideal for animation (one node = one responsibility), for parenting (chaining several Transforms), and for reusing a group in several branches with different transforms.

Rule: Merge Transform for a simple static offset, a dedicated Transform 3D for animation or parenting.

Use cases

  • Animating an object - Shape 3D → Transform 3D (animated position + rotation) → Merge 3D. The source node stays "frozen", only the Transform is animated.
  • Articulated rig - chaining Transform 3D nodes to emulate a bone hierarchy (see the Parenting section).
  • Look-At camera - Camera 3D → Transform 3D with Use Target checked and Target connected to an object's position. The camera follows the object automatically.
  • Reusing an asset - the same asset Merge 3D sent into two branches, each with a different Transform 3D → two placements of the same object.
  • Custom pivot for rotation - to rotate a cube around one of its corners rather than its center, move the Pivot to the corner's location.
  • Importing external tracking - a track solved in Maya/SynthEyes/3DEqualizer can be imported through Transform 3D (.lws, .ase, .ma, .xsi formats).

Tips & pitfalls

  • Rotation Order changes the result - for combined rotations > 90°, changing the order visibly changes the object. YXZ is often more predictable for FPS-style cameras.
  • Animate X/Y/Z Rotation rather than Rotation Order - the order changes results abruptly; only change it once, at the start of the project.
  • Use Q/W/E in the viewer for rough positioning, then fine-tune in the Inspector.
  • Name your Transforms in a hierarchy: xf_body, xf_arm, xf_hand. Otherwise the graph becomes unreadable with Transform3D1, Transform3D2…
  • Use Target + Connect To = smart cameras that follow without manual keyframes. Very powerful.
  • For clean animations, keep one Transform 3D per responsibility (translation on one, rotation on another) rather than stacking everything into a single one.