What it does
Colorspacious is an MIT-licensed Python package for converting colour arrays between standard and appearance-based spaces. Its tutorials demonstrate whole-image transformations, colour differences and adjustable colour-vision simulations.
Pass a colour or image array to cspace_convert with the input and output spaces. The names sRGB1 and sRGB255 distinguish normalized channels from byte-range values. An image can be converted to JCh, have its chroma changed and then return to sRGB for display.
Conversions deliberately leave out-of-range values unclipped because intermediate calculations may need them. The tutorial shows clipping at the display stage when small overflows would otherwise create artifacts. Simulation settings select a deficiency type and a severity from zero to 100; these remain modelled views rather than an individual vision assessment.
The package depends on NumPy and is available under the MIT licence.
Strengths
- Whole arrays use the same conversion interface as single colours.
- Appearance-model components can be edited before converting back.
Watch out for
- The caller must choose the correct input range and handle output clipping.
Best for
- Python image analysis involving colour spaces or perception models
Sources
- Colorspacious supported spaces (Oct 27, 2026)
- Colorspacious conversion and simulation tutorial (Oct 27, 2026)
- Colorspacious installation, dependencies and licence (Oct 27, 2026)