Video & Projection

Video, Image, and Text cues all render onto a shared canvas — a virtual surface (default 1920×1080) that one or more output windows display regions of. Video is decoded by FFmpeg on a worker thread with vsync-paced backpressure; YUV→RGB conversion happens on the GPU.

The Video Output window

Window → Video Output opens a single window showing the whole canvas — the simplest setup: drag it onto the projector and fullscreen it.

Projection mapping

Window → Projection Mapping is the multi-output editor. Set the canvas size, then add outputs — one per projector. Each output has:

PropertyMeaning
Source regionThe rectangle of the canvas this projector shows (x, y, w, h in canvas pixels).
Output sizeThe projector's native resolution.
Fullscreen monitorWhich monitor the output window fullscreens onto. Assignments are remembered by monitor identity (name + geometry), so they survive unplugs and OS re-ordering.
Edge blendPer-edge (left/right/top/bottom) feathering: enable, blend width in pixels, and gamma.

Overlap the source regions of adjacent projectors by the blend width and enable edge blending on the touching edges to get a seamless wide image. The 3×1 edge-blend preset sets up a 5400×1080 canvas across three 1920×1080 projectors with the overlaps pre-configured — a good starting point to study.

Content fit

Video/Image/Text cues have a Fit mode controlling how their content maps to the canvas:

  • Fit (default) — preserve aspect ratio, letterbox/pillarbox.
  • Fill — preserve aspect ratio, cover the whole canvas, center-crop the overflow.
  • Stretch — fill the canvas exactly, ignoring aspect ratio.

The canvas is also what lighting pixel-map segments sample by default, so LEDs can mirror the projector picture.

Video status indicator

The status bar shows the current clip's decode path. Green means GPU decoding is active with no reported fallback. Amber with a ⚠ means decoding has fallen back to the CPU or a slower GPU path. Hover over the indicator for the reason and source file; Help > Status provides more detail.

The path is named as the decoder reports it: hap gpu-native, d3d12va zero-copy (<adapter>), d3d11va readback, hardware (videotoolbox), or software. The indicator stays amber after a fallback even if decoding still uses the GPU, such as when zero-copy was lost.