gfwx::quant
pub fn quantize( image: &mut [&mut [i16]], quality: i32, min_quality: i32, max_quality: i32)