Staff Writer, CNET News.com
One of the inventors of the MP3 format is back with a new technology that he hopes will revolutionize audio, creating superrealistic sound for theaters, theme parks and eventually even living rooms.
Karlheinz Brandenburg, director of the Fraunhofer Institute for Media Technology, along with a team of co-developers, is in Los Angeles this week showing off his new "" technology to representatives of Hollywood studios and giants including Disney. Brandenburg is leading to the MP3 format, also developed at Fraunhofer.
He and his team are touting their new product as true "three dimensional" audio, which can give the impression of, for example, a horse galloping through the center aisle of a movie theater, or pinpoint a noise so that it sounds exactly like a person shouting from outside theater walls. The best existing surround sound speakers can approximate this only for a small "sweet spot," perhaps a few feet wide, while the Ionsono system would create the same realistic illusion for everyone in the room.
"It was an old dream to do something like that...to do something for immersive audio, where people would feel they were in a different place," Brandenburg said in an interview. "PCs have now become fast enough that you can do the (necessary) processing in real time. It was not realistic to do that 10 years ago."
The project marks a substantial break from the way recorded sound has been replicated since Thomas Edison first began experimenting with recorded audio in the late 1870s. Just as video is being wholly transformed by digital cameras and computer processors, audio production and reproduction, too, is being transformed by the latest generation of PCs and processors.
Sound amplification has worked in much the same way for years. A sound wave is turned into an electric impulse, which is turned back into sound as it hits speakers. In the early 1930s, stereo or "binaural" sound was first patented, which used two sound sources to create the illusion that the sound was coming from a wider space.
This technique developed over time into the sophisticated audio systems of today. Sound is split between speakers into "channels," and recording tricks such as hints of echo or reverberations are added to create the increasingly realistic impression of sound coming from all around a listener.
But most of this analysis worked on the principle of a perfect listener positioned at exactly the right midpoint between the speakers, where the sound waves would meet and interact to create exactly the right illusion.
"Many of the developments in this area have been about fine-tuning frequency response and brute horsepower," said David Stump, an Academy Award-winning cinematographer and visual effects artist who has seen demonstrations of the Iosono technology. "The thing that's different about Iosono is that it just takes the approach of intelligent analysis of what sound really is and how it shows up to your ears and applies that in a different way."
Sound travels in waves, much as the waves in a pond ripple outward when a pebble is thrown in. The sounds of a busy street corner are complicated and confusing, the same way the surface of a pond would be if a handful of pebbles were thrown in at the same time.
Brandenberg's team aims to create the audio version of that pond surface being pelted by a shower of pebbles.
To do this, they use an array of small speakers, sometimes as many as 300 or 400. A complicated algorithm works out exactly what the sound waves all through a room would be if, say, the horse were galloping through the center aisle.
Each little speaker then emits whatever little piece of sound wave is necessary to create—in the 300-strong aggregate—the full virtual "picture" of the sound throughout the room.
"It's pretty amazing," said Stanley Johnston, a recording engineer for postproduction company Todd-AO, who was brought in to help with Iosono's Los Angeles demos. "It enables a sound mixer and sound designer to have the entire room to play with on a level that we've never had before."
This kind of sound is relatively easy to mix. The developers have created a console that uses a light pen to let a sound engineer specify exactly where he or she wants the sound to seem to be coming from, and the software makes the calculations on the back end. A decoder in the theater or other final venue would translate this into sounds that worked for the specific size and shape of the room.
None of this is cheap, and it isn't likely to find its way into a local multiplex soon. Fraunhofer is providing licenses to the technology itself for between $10,000 and $15,000. But it also needs powerful computing hardware and the—for today—extremely expensive array of speakers that can ring the production space.
Brandenberg and his team see the technology being adopted by theme parks and other dedicated-use spaces first, and later by movie theaters and home theater aficionados.
"At the moment, it's only for those people who can afford a home theater system that costs as much as other people's homes," Brandenberg said. "But I assume this technology will end up everywhere where you can afford to have more than a few loudspeakers."
For now, there will be a chicken-and-egg problem, however. Until there are venues that support the technology, there is little reason for sound producers to buy and work with it. And until there are movies being produced in this form, there's not much reason for theaters to do expensive upgrades—particularly when they're already facing pricy upgrades to digital projection systems.
But industry insiders who have listened to it say it might just be a matter of time.
"If technology proves itself worthwhile, issues of practicality and affordability eventually solve themselves somehow," Stump said.