On 2011-11-25, Richard wrote:
I totally agree, any mention of MP3/WMA or any of the very lossy formats in the same breath as surround cannot be taken seriously.
That is very much true where it goes to ambisonic and the existing formats. But as for the formats and say 5.1, not so much. That actually works rather well. Not perfectly, but it works. And on the other hand, ambisonic could work *extremely* well when compressed by its own standards and requirements. Perhaps even bettern than pair-wise panned stuff, which the other kind always is.
I mean, I've read entire reams of papers which apply PCA and whatnot to sets of channel feeds. They always get meager results, because of temporal effects which are difficult to compensate for. I've never seen a paper applying even the most simple packing technology to a normalized B-format at any order. Still, it'd stand to reason it'd pack unreasonably well because of the temporal coincidence of the signals, the directional rolloff of the signal-set which doesn't happen with non-coincident mics, and especially because you can then do certain tricks you couldn't otherwise do in compressing the signalset.
I'm reasonably sure a dedicated ambisonic compressor could do twice as much as a discrete 5.1 one ever could, given the same physical acoustical starting point.
-- Sampo Syreeni, aka decoy - [email protected], http://decoy.iki.fi/front +358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2 _______________________________________________ Sursound mailing list [email protected] https://mail.music.vt.edu/mailman/listinfo/sursound
