This project is part of the @thi.ng/umbrella monorepo.

• About
  • References
  • Status
  • Related packages
• Installation
• Dependencies
• Usage examples
• API
  • Basic usage
  • Operators
  • Custom operators
  • Additional operators / modifiers
  • Pre/post-multiplied colors
• Authors
• License

Porter-Duff operators for packed ints & float-array alpha compositing.

This package provides all 13 fundamental Porter-Duff compositing / blending operators, and utilities to pre/post-multiply alpha. All operators are available for packed ARGB/ABGR 32bit packed ints or RGBA float vectors.

• https://keithp.com/~keithp/porterduff/p253-porter.pdf (original paper)
• https://en.wikipedia.org/wiki/Alpha_compositing#Alpha_blending
• https://ciechanow.ski/alpha-compositing/
• http://www.adriancourreges.com/blog/2017/05/09/beware-of-transparent-pixels/
• http://ssp.impulsetrain.com/porterduff.html

STABLE - used in production

Search or submit any issues for this package

• @thi.ng/pixel - Typed array backed, integer and floating point pixel buffers w/ customizable formats, blitting, dithering, conversions
• @thi.ng/shader-ast-stdlib - Function collection for modular GPGPU / shader programming with @thi.ng/shader-ast

yarn add @thi.ng/porter-duff

// ES module
<script type="module" src="https://unpkg.com/@thi.ng/porter-duff?module" crossorigin></script>

// UMD
<script src="https://unpkg.com/@thi.ng/porter-duff/lib/index.umd.js" crossorigin></script>

Package sizes (gzipped, pre-treeshake): ESM: 998 bytes / CJS: 1.16 KB / UMD: 1.14 KB

• @thi.ng/api
• @thi.ng/math

Several demos in this repo's /examples directory are using this package.

A selection:

Screenshot	Description	Live demo	Source
	Pixel buffer manipulations	Demo	Source
	Port-Duff image compositing / alpha blending	Demo	Source

Generated API docs

import * as pd from "@thi.ng/porter-duff";

// packed int version (premultiplied ARGB)
pd.SRC_OVER_I(0x80800000, 0xcc0cc00)

// automatically premultiply inputs & post-multiply result
pd.porterDuffPInt(pd.SRC_OVER_I, 0x80ff0000, 0xcc00ff00);

// the above is same as:
pd.postmultiplyInt(
    pd.SRC_OVER_I(
        pd.premultiplyInt(0x80ff0000),
        pd.premultiplyInt(0xcc00ff00)
    )
)

// premultiplied float version [R,G,B,A]
pd.SRC_OVER_F([1, 0, 0, 0.5], [0, 1, 0, 0.8]);

Integer operators are suffixed with _I, float versions with _F. Consult above diagram for expected results.

• CLEAR
• SRC
• DEST
• SRC_OVER
• DEST_OVER
• SRC_IN
• DEST_IN
• SRC_OUT
• DEST_OUT
• SRC_ATOP
• DEST_ATOP
• XOR
• PLUS

New operators (e.g. for blend modes) can be easily defined via porterDuff / porterDuffInt. Both functions take 2 function arguments to extract blend coefficients from the src & dest colors:

// coefficient functions take the normalized alpha values
// of both colors as arguments, but unused here...
const customOp = porterDuffInt(() => -0.5, () => 1);

The following coefficient functions are included by default (and are used by all standard operators):

• ZERO => 0
• ONE => 1
• A => alpha of src color
• B => alpha of dest color
• ONE_MINUS_A => 1 - alpha of src color
• ONE_MINUS_B => 1 - alpha of dest color

The following modifiers are also discussed in the original Porter-Duff paper (linked above).

• darken / darkenInt
• dissolve / dissolveInt
• opacity / opacityInt

All Porter-Duff operators expect colors with pre-multiplied alpha. Premultiplication is also recommended for transparent WebGL textures (especially when using mipmaps). For that purpose the following helpers might be useful:

• premultiply / premultiplyInt
• postmultiply / postmultiplyInt
• isPremultiplied / isPremultipliedInt

Furthermore, existing PD operators can be wrapped with automatic pre/post-multiplies using porterDuffP / porterDuffPInt (see example above).

Note: HTML Canvas ImageData is using non-premultiplied colors.

Karsten Schmidt

If this project contributes to an academic publication, please cite it as:

@misc{thing-porter-duff,
  title = "@thi.ng/porter-duff",
  author = "Karsten Schmidt",
  note = "https://thi.ng/porter-duff",
  year = 2018
}

© 2018 - 2021 Karsten Schmidt // Apache Software License 2.0