16 TVBEurope News & Analysis 3D sound: listening to the future
LAST FEBRUARY’S Broadcast Video Expo seemed to focus largely on leveraging and developing existing technologies rather than pushing ahead into the new frontiers. One exception was a panel discussion called ‘Immersive Audio Systems Explored — Sound for 3D’. There has been a great deal of industry discussion of the potential and application of 3D images, but very little of the potential of so called immersive or ‘3D’ audio, immersive sound that gives a viewer the sense of actually being physically within the action taking place. The goal of immersive audio is to mimic the three- dimensional sound experience of real life — being able to detect the direction and motion of objects on all three axes. The BVE panel was moderated by Peter Schillebeeckx of Soundfield and featured Chris Pike of BBC R&D, Senior Sound Supervisor Martin Black of BSkyB, Applied Research Manager John Dollin, also of BSkyB, and Mark Pascoe of Dolby Laboratories. Schillebeeckx’s opening
remarks summed up the state of technology for 3D sound: “There has been a lot of talk over the past few, well, decades about the possibility of 3D sound but a lot of it has been very much research-based. There’s a lot of exciting possibilities there, but the real question is how can we potentially translate that into the field of broadcast?” This summer, select viewers
will experience a broadcast of live 3D audio when Japanese broadcaster NHK runs demos during the Olympics of its Super Hi-Vision video technology. The presentation of 8K video on a 15m screen is accompanied by 3D audio, delivered via 22.2 channels in three spatial layers – a top layer of nine channels, a middle layer of 10 channels, a lower layer of three channels, and two LFE channels. The ultra-HD screen will give a
sensation of actually viewing the action from a stadium seat, and
The goal of immersive audio is to mimic the three-dimensional sound experience of real life — being able to detect the direction and motion of objects on all three axes. Neal Romanek reports
Frank Melchior reiterates that audience immersion is not the only important aspect of a 3D audio future — 3D audio must be easy for customers to use if it is ever to be adopted
“I think we’re at the point where we have to start thinking about revolution. I think object-oriented is the way to go” Mark Pascoe, Dolby
Chris Pike: “When people have listened to the result in our studio, you can see the smiles erupt onto their faces. They’re definitely transported to the cathedral”
the audio aims to recreate the ambience of the Olympic stadium with the sound of the athletes from below and the cheers of spectators in front and behind. Pike of the BBC described
several immersive audio experiments, including a 3D audio presentation of a concert at Manchester Cathedral. The microphone set-ups were little different from the usual configuration for 5.1 recording, but BBC R&D developed custom mixing software
Frank Melchior: “The past has shown that flexibility is a very important issue for the audience, so you can listen in different environments with different devices”
allowing them to place the microphone signals in a virtual 3D space. “When people have listened to the result in our studio,” he says, “you can see the smiles erupt onto their faces. They’re definitely transported to the cathedral.” The leap forward in 3D audio is the concept of object-oriented sound, in which separate elements on the soundtrack are transmitted and received as discrete elements that can be positioned in virtual space.
Schillebeeckx says, “Up until now we had mono, then we went to stereo, and then we went to 5.1, but they are fixed loudspeaker arrangements and the mix is flat. It always plays back the same way. “The whole concept of object-based sound is very interesting because all of a sudden you really are describing a three-dimensional space. If you then play it back over 22.2, that will work fine, but if you want to play it back over 5.1 or stereo that will also work. “
Revolution and immersion Pascoe of Dolby warned that merely adding on to the existing way of doing audio was insufficient for introducing 3D audio: “Having another technology that’s another ‘x.1’ isn’t any sort of development. That’s an evolution, rather than a
revolution. I think we’re at the point where we have to start thinking about revolution. I think object-oriented is the way to go.” But he also noted that the meaning of ‘object-oriented’ is not a fixed point. “Does it mean that we’re placing a bunch of these objects on a sphere surrounding us? Or alternatively are we — as you would in a game engine — putting together a whole bunch of objects in space and then placing ourselves at various points in that space?” How to deliver these sound ‘objects’ to the end consumer poses a huge challenge. Delivery of object-oriented 3D sound immediately brings up the question of bandwidth. Pascoe points out that variable bitrate audio is difficult to manage, and if a broadcaster is trying to deliver a number of services, each with a highly variable bitrate