PorterDuff

Porter-Duff refers to a formal set of image compositing rules used to determine how a source image blends with a destination image in alpha-based rendering. Introduced by Matt Porter and Tom Duff in 1984, the framework provides a concise mathematical model for combining two RGBA pixels to produce a new color and transparency.

The model identifies a family of binary operations that depend on the source (S) and destination (D) pixels. Each operation yields a color output and an resulting alpha that describe how much of each input contributes to the result. The most common modes cover basic layering, masking, and transparency tasks used in graphics software and rendering pipelines. The Porter-Duff operators are widely implemented as compositing or blend modes in graphics APIs and editors.

The standard Porter-Duff operators (SRC/DST) include:

- Clear: result is fully transparent.

- Source (SRC): result is exactly the source.

- Destination (DST): result is exactly the destination.

- SourceOver: standard alpha over operation.

- DestinationOver: destination drawn underneath the source.

- SourceIn: keep source where destination exists.

- DestinationIn: keep destination where source exists.

- SourceOut: keep source where destination is absent.

- DestinationOut: keep destination where source is absent.

- SourceAtop: source on top of destination where both overlap.

- DestinationAtop: destination on top of source where both overlap.

- XOR: only the non-overlapping regions contribute.

Modern implementations often assume pre-multiplied alpha, simplifying calculations. Graphics APIs and tools—such as HTML canvas, SVG, and various rendering engines—expose these modes as blend or composite operations. The Porter-Duff framework remains foundational for understanding and implementing transparency and layering in digital imaging.