Color management and transformation is a complex domain. Values are captured by cameras or generated by renderers, transferred to a storage medium, combined into frame buffers to make new images, transformed for display, and projected by devices. During each step, the color values may be transformed by photochemical, electro-optical, and digital processes.
A renderer's input working space to output working space is an interesting subset of that domain, and it's the subset that Nanocolor concerns itself with.
OpenEXR goes so far as to restrict itself to linear working spaces, and describes them completed by specifying chromaticities and an adapted whitepoint in the CIEXYZ 1931 space. Nanocolor takes inspiration from this, and uses the equations in SMPTE document RP177-1993 (reaffirmed in 2002) ~ SMPTE Recommended Practice: Derivation of Basic Television Equations.
Interesting colors do not just come from OpenEXR however, but may originate as properties in a MaterialX shade graph, vertex attributes in an OpenUSD file, or PNG or TIFF textures, to name a few common sources.
Amongst all these sources, only OpenEXR specifies colors in terms of chromaticities and whitepoint. Some sources specify the color space in documentation, whereas others name color spaces in the data itself, or refer to site specific configuration files in the OpenColorIO format, and other alternatives. This lack of broad agreement on such things as what the name of a colorspace actually refers to makes accurate and reproducible color calculations between software packages and studios challenging.
MaterialX takes an interesting perspective; it names a bakers' dozen of color spaces and defines them as OpenEXR does, and also introduces a notion of being able to remove an input transform from a color in order to accomodate image formats commonly used as texture sources for shader graphs.
Nanocolor follows this idea, and also introduces that all the color spaces must be invertible, in the sense that a value can have its input transform removed to make it a linear color value, linear values may be transformed as desired through matrix multiplication, and an output transform may be re-applied. This allows Nanocolor to transform any color space with this invertible property to any other with the invertible property.
Nomenclature shall be as per ISO 22028-1 terminology, and will occasionally spell colour with a u per common usage.
A geometric representation of colors in a metrical space.
A colour space having an exact and simple relationship to CIEXY colorimetric values.
A “spectral radiance distribution” that converts to achromatic colour signals where each component is 1.0
colorimetric colour space with three primary chromaticities, a white point, and a transfer function
“digital encoding of a colour space, including ... digital encoding method, and ... value range”
Colour space encoding plus context, including image state, intended viewing environment, and for print, reference medium
Colour space encoding is in scope.
Colour image encoding is out of scope.
Camera vendor specific colour spaces are out of scope.
Colour spaces relative to a particular encoded white (cf. LAB, LUV, etc but also IPT and other spaces optimized for gamut mapping) are out of scope.
Hybrid Log Gamma encoding is out of scope.
Refer to nanocolor.h
There are no build scripts included with Nanocolor. You may build it as a library if you wish, or you may include nanocolor.c, and optionally nanocolorUtils.c in your project.
//
// Copyright 2024 Pixar
//
// Licensed under the Apache License, Version 2.0 (the "Apache License")
//
// You may obtain a copy of the Apache License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the Apache License with the above modification is
// distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
// KIND, either express or implied. See the Apache License for the specific
// language governing permissions and limitations under the Apache License.
//Thanks to the scientists at Pixar, OpenColorIO, OpenEXR, OpenUSD, and MaterialX for the fruitful advice and feedback that helped guide the creation of Nanocolor.