DVE

DVE (Digital Video Effects) is Fusion's classic 2.5D transform for a 2D image. It does not create true scene geometry, lights or a camera, but it treats the image like a flat card that can be positioned, moved along a simplified Z axis, rotated around three axes and viewed with perspective foreshortening.

Keep Center and Pivot conceptually separate. Center places the plane in the frame. Pivot X/Y and Z Pivot define the local fulcrum used by the transform. A door-style flip therefore moves Pivot toward an edge while Center remains responsible for placement.

Z Move is not a normal Size control. It changes the plane's position along the DVE depth axis, so its apparent scale is tied to Perspective. Use Transform for a simple uniform resize; use DVE when the visual intent is a move toward or away from the viewer.

Rotation Order defines the sequence in which X, Y and Z rotations are evaluated. With only one active axis the orders may look identical, but two or three non-zero rotations can produce different final orientations and motion paths.

Rotation X tilts the card forward/backward, Rotation Y swings it like a door, and Rotation Z rotates it in screen space. Shear X/Y is a separate planar skew that creates a parallelogram-like deformation; it is not a substitute for perspective rotation.

Perspective controls the strength of foreshortening produced by X/Y orientation. Judge it on an already tilted plane rather than a front-facing card. Extreme values can stretch edges aggressively and make text or UI difficult to read.

DVE has two different mask inputs. DVE Mask is a dedicated pre-mask: it selects the source region before the transform, so the selected pixels and their matte are transformed together while the rest of the input can remain untransformed.

Effect Mask is the normal effect mask. It limits the contribution of the already processed DVE result after the transform has been calculated. The same Rectangle therefore produces different behavior depending on whether it is wired to DVE Mask or Effect Mask.

The Masking tab belongs to that pre-mask workflow. Black Background removes the region outside the DVE Mask from the transformed layer. Fill Black removes the region inside the DVE Mask from the untransformed input. Enabling both leaves only the masked, transformed region.

Alpha Mode controls the internal composite of the transformed pre-masked region over the untransformed input. Ignore Alpha makes the masked area opaque. Additive is intended for premultiplied foreground pixels; Subtractive is intended for non-premultiplied foreground pixels. Treat this as premultiplication handling, not as a creative blend-mode picker.

DVE is excellent for card flips, screens, photographs, titles, B-roll inserts and UI motion. Move to Fusion's true 3D toolset when several independent objects need real parallax, Camera3D, lights, materials, intersections or Renderer3D.

The supplied Controls capture shows Center X/Y = 0.5, Z Move = 1.0, Pivot X/Y = 0.5, Z Pivot = 0, Rotation Order = ZYX, zeroed X/Y/Z rotations and Shear, and Perspective = 1.0. These are observed capture values, not universal presets.

DVE does not convert an RGBA stream into a Fusion 3D scene. There is no output suitable for Merge3D, Camera3D, Light3D, or Renderer3D; the DVE output is still 2D and is composited using Merge.Foreground.

A genuine rectangular card needs an alpha-bounded source. For example, Background with Rectangle on its EffectMask makes a card; without that matte, DVE rotates an opaque full-frame Background.

Shear X/Y are genuinely visible in the supplied Controls capture even when older short DVE documentation omits them. Shear produces planar skew while X/Y Rotation combined with Perspective yields depth-style foreshortening.

The open Alpha Mode screenshot shows Additive selected. Ignore Alpha and Subtractive are listed but appear grey; do not promise they can be selected without a live test. The manual documents their meaning, not availability in this specific UI state.

The Settings tab is closed in all three supplied images. Shared Transform-node properties are described in the reference guide, but the exact active DVE Settings controls in this build cannot be certified from these images.

The four instructional graphs distinguish an isolated alpha card and Merge, Rotation/Z timing, a white DVEMask source-selection path, and a blue EffectMask post-effect path. These are static educational graphs, not executable .comp files or native render verification.

Extreme tilts may soften typography and fine lines during resampling. Inspect angles, perspective, output dimensions and premultiplication at 100% Viewer magnification and final delivery size.