FASTLED_NOISE_FIXED

#define FASTLED_NOISE_FIXED   1

Use this to toggle whether to use "fixed" FastLED 8-bit and 16-bit noise functions.

The prior noise functions had some math errors that led to "discontinuities" in the output, which by definition should be smooth and continuous. The bug led to noise function output that had "edges" and glitches in it. This is now fixed, and the fix is enabled by default. However, if for some reason you wish to run with the old noise code, including the glitches, you can disable the bugfix here.

This also covers the one-dimensional gradient used by inoise8(x) and inoise16(x), corrected for FastLED#1114. It was the two-dimensional gradient with the second coordinate missing, which left a quarter of the hash table with no gradient at all, and 21 of the 256 lattice cubes flat across every one of their inputs – the longest plateau ran 1063 consecutive samples of inoise8 at a constant 128. Setting this to 0 restores that behaviour along with the rest of the old noise code.

Definition at line 73 of file fastled_config.h.