On 2013-05-15, Augustine Leudar wrote:
I have used the 48 channel system at sarc on many occasions and have set up my own 32 channel one - and if I want a sound to come from a point source I would not use ambisonics- I would try and have the sound come out of only one speaker use vbap or at least ambisonics with the directivity very much reduced [...]
That ought to be very close to what higher order converges to, especially when using in-phase decoding and allowing rE to vary in directions with closely spaced speakers. If that is not enough, maybe somebody should package a decent active matrix/infinite order thingy with a decent out-of-the-box decoder, and/or take another look at 5.0 compatibility codings. There is room for that extra one channel in G-format which could be used for center panned dialog, three channel frontal stereo, a center back, or even order 1.25 encoding if you're willing to bend the rules a bit.
Speaking of dialog, one potential problem in theatrical applications is the frame of reference. Ambisonic is a sink standard which is rigidly referenced, in absolute length units, to what happens in the vicinity of the sweet spot. Current theatrical systems are relative source standards referenced to the playback rig, with that in case referenced to things like screen size and center position. Those two viewpoints don't always mesh, because even if you just work with angles, the reproduced ones don't match over different auditoria, and in particular when you have extended audiences, you get wonky parallax even at infinite order. (The same obviously goes for WFS too.) I have some trouble believing current theatrical audio people could demonstrate such effects given their relative mode of operation, but theoretically speaking that could be a bit of a snag.
Yes film producers do use surround with some effects travelling behind the listener but you will find the majority of the audio activity goes on in front of the listener (thus the middle speaker in 5.1) with much less activity and lighter mixing behind the listener (such as an aeroplane flying overhead etc) - precisely for the reasons already given.
Then you mix it accordingly, with only the portal represented by the screen reproduced holophonically.
I agree about the wife and the speakers though - it has already caused several low level marital disputes.
That's obviously one of the oldest promises of ambisonic as a distribution format: decoder adaptation to the wife. Currently the problem is "simply" that an easy to setup free decoder library isn't available. You know, the like that you can just safely link against and throw whatever audio you have at, and which after a reasonably simple setup step just Does What The User Wants (tm). Combined perhaps with a nice little Qt setup app where you can input angles and distances, or perhaps even hook up a mic for automatic guesswork. With something like that you'd have a possibility of getting accepted as the de facto audio compositing interface, and including ambisonic basically as a bonus feature.
Apropos object oriented transmission like Atmos or audio BIFS, you can't really get rid of ambisonic that way. That's because ambisonic and WFS are the only technologies out there which let you capture soundfields in addition to synthesizing them, and of the two only ambisonic scales down gracefully to something you can distribute for home use. So, what Dolby calls "bed" -- the static background -- probably should be handled as a multichannel ambisonic field even in an object based workflow, and only decoded as part of the compositing process which already has to know where the speakers are.
-- Sampo Syreeni, aka decoy - de...@iki.fi, http://decoy.iki.fi/front +358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2 _______________________________________________ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
