-
Téo Mazars authored
- Takes 2048 samples instead of 256 to handle properly large palettes. It gives also better results with gradients. - linearly interpolate values between two samples when mapping with a gradient, to handle properly float precision.
092f239d