Colour 0.4.7 is available!
The colour-science Developers are pleased to announce the release of Colour 0.4.7!
This release introduces new chromatic adaptation transform, colour appearance model and colourspaces, thin film optics via the Transfer Matrix Method, and significant performance optimisations. It also marks the adoption of Claude Code for development assistance.
The highlights are as follows:
Dependencies & Compatibility
Support for Python 3.14
Dropped typing-extensions requirement
Updated minimum versions: numpy >=2.0.0, scipy >=1.13.0, matplotlib >=3.9, networkx >=3.3, pandas >=2.2
Colour Adaptation & Appearance
Implemented Li (2025) chromatic adaptation transform
Added sCAM colour appearance model with conversion definitions
Colour Models
Introduced sUCS colour space with multiple conversion functions
Added 12 new RGB colourspaces from the Color Interop Forum recommendation for ColorSpace Encodings for Texture Assets and CG Rendering
Implemented Filmlight E-Gamut 2 and Fujifilm F-Gamut C colourspaces
Added Xiaomi Mi-Log Profile encoding/decoding definitions
Enhanced OSA-UCS conversion with improved accuracy and 28.5x performance gain
Optical Phenomena
Implemented Transfer Matrix Method for thin film structures
Added Snell's law and polarised light calculations
Introduced comprehensive plotting definitions for thin film analysis
Quality Metrics
Added Japanese skin complexion spectral data
Extended spectral similarity index to support
MultiSpectralDistributions
Documentation
Each documented object has been verified and processed with a LLM to improve consistency and correct typos and mistakes
Utilities
Introduced metadata system for documenting input/output value ranges via type annotations
Performance Improvements
Trilinear and tetrahedral interpolation: 1.3x faster
Robertson (1968) colour temperature conversions: 30-90x faster
scipy import optimisation for reduced load times
Please take a look at the releases page more information.
Comments
Comments powered by Disqus