This articles deals with the reduction – or “projection” or “downscaling” – of a 3D-encoded Ambisonic sound field (“full-sphere” or “periphonic”) into a 2D representation (“horizontal-only” or “planar” or “pantophonic”). We show that the reduction operation can be equivalently achieved by 1) applying conversion formula of the normalization factors, or 2) performing a plane-wave decomposition of the original sound field and re-encoding the resulting plane waves to 2D Ambisonic. The latter approach provides greater flexibility in adapting the content to horizontal-only reproduction.