The fundamental thing future surround formats have to do is to break the outmoded, inefficient business of one channel, one speaker.

You need to encode or create the soundfield in the most efficient way, transmit that number of channels - say six or eight at most for hi-res with-height surround. Then your processor renders that for your hardware configuration to deliver the best re-creation of the original soundfield using your particular speaker layout. Like a PostScript file that describes the finest detail and renders at the resolution of the printer. MLP can already do that, I just hope that capability is in the spec.

It would put an end to people wanting "enough channels" to do "22.4" or whatever - by all means have two dozen speakers, but if the whole soundfield is encoded completely in eight channels, you don't need any more /channels/. You just tell your processor where all those speakers are and it generates the feeds to best reproduce the soundfield on that system.

--Rik

2x DSP5000, 1x DSP3200, G57 (rears), G61, G98, Explorer2, SB3, Plex, LMS, Roon in process