pub fn multiply_u8_color(a: u8, b: u8) -> u8
Returns a*b/255, rounding any fractional bits to nearest integer to reduce the loss of precision after multiple consequence alpha (un)premultiply operations.