Visit the site and watch the short movie (or watch it below) for some amazing audio effects. Remember to wear your headphones.

Enjoy

It is an example of what is called a ‘binaural recording‘. A dummy head, with tiny microphones stuck inside its ears, is used to record an acoustic environment such as a person cutting your hair or a band playing in a room. You then listen to the binaural recording over headphones and it is as if you were sitting at the position of the dummy head at the moment of the recording. The best way to really understand the process is to just watch and listen to these examples:

YOU MUST USE HEADPHONES FOR THE FIRST EXAMPLE!

‘Snakecharmer-Binaural‘ by Ottmar Liebert

John Q. Walker’s talk (7 minutes 45 seconds in) entitled ‘Re-creating great performances’ from the EG 2007 conference posted on TED

Don’t forget to subscribe to the blog’s RSS feed on the home page to get updates on all the latest posts and demos.

Add to: | Technorati | Digg | del.icio.us |

Email a friend

The photos are currently offline for this post. They will be back very soon!

Below is a run-down of the process normally involved, which makes the technology difficult to commercialize in its current form. Thus the push for a viable method in the current ‘listen with your own ears’ project in which a personalized 3D audio will be generated from a simple photo of the person’s head and ear.

A sound proof anechoic (no echoes) chamber is needed first.

Then tiny microphones are placed into my ears.

I am then seated in the middle of the chamber with the microphones inside my ears, and a head tracker on my head to let us know the position of my head at all times. Then sounds are played consecutively out of the three small speakers seen in the above photo on the robotic arm. The sounds are recorded with the small microphones in my ears. Then I am rotated around say 15 degrees on the chair and the sounds from the speakers are again played and recorded. This process continues for a full rotation, and then the robotic arm rises up a few degrees and a new rotation begins. This way the sound from the speakers is recorded from many positions on an imaginary sphere around my head.

The recorded signals taken from the microphones for the many positions around my head are then stored on a computer. Using these recordings we can see how my own ears filter sounds for each position in space. Using this information we can then take any audio stimulus and synthesize it so that when I listen to it over headphones I will get the impression that the sound is coming from one of the recorded positions in space. I will at this stage have a personalized 3D audio.

Thanks to Jack for the shots.

Add to: | Technorati | Digg | del.icio.us |

Email a friend

Article as pdf

Article online

THE END USER

Taking sound to a new level

By Victoria Shannon Published: March 19, 2008

PARIS: The music of my teenage years came to me mostly via AM radio. When I listen to music now, it is delivered in digital format over headphones. Compared to the tinny, static-filled radio of my youth, how could I ever want better sound than that?

It may be surprising to hear, but experts are already working on the next generation of audio. In start-up companies and research labs around the world, the reproduction of music is undergoing a fundamental transformation.

In a lab outside of Paris, for instance, researchers are working on how to recreate accurately what sound engineers intended when they were in the studio mixing musical recordings.

But the goal of the project is to reproduce that fidelity over the kinds of simple headphones that we carry around with us, not over the thousands of euros’ worth of equalizers and speakers necessary in a home stereo or cinema setup.

Brian F.G. Katz, a researcher at the Centre National de la Recherche Scientifique, said his work was based on the principle that our perception of sound is individual, determined by the size and shape of our ears and skulls – a difference lost on today’s headphones.

Today in Technology & Media

He is working with a Paris-based company, Arkamys, on a three-year effort to bring three-dimensional sound to ear buds. It is something that Arkamys is pitching as “listening with your own ears” – as opposed to listening to music over headphones that are best suited to someone else’s ears.

In a commercial sense, the idea would work like this: You use your mobile phone to take several photos of the side of your head and your ears. You send the photos over the cell network to a service that analyzes them against a collection of three-dimensional images of ears.

The service matches your ears to the closest model, and then sends back to your phone customized software that changes the music you listen to in such a way that it accurately recreates the artists’ intention.

In time, according to Philippe Tour, chief executive of Arkamys, the software would be integrated directly into mobile devices by the manufacturer, with which Arkamys is currently working, without requiring users to change or modify their headphones in any way.

Katz is presenting a study this summer that proves what we have all suspected: that there are vast differences among various kinds of headphones, from ear buds to in-ear, cheap to expensive, closed to open.

“What you find is, yes, there is quite a difference among them,” Katz said, “because the physical response of the headphone and the physics of how they function is not the same. But it also shows that if you do sufficient equalization, you can make ear bud headphones have the same 3-D quality as an expensive headphone.

“I’m not saying they have the same fidelity, but you can get rid of the defaults in terms of spacial sound,” Katz said. “There’s going to be a limit to what physically that little speaker can do.”

The benefits of improving headphone sound can extend to non-entertainment uses as well. Three-dimensional sound in headphones can be used as navigational aids for the blind, enabling mobile phones with GPS and 3-D sound to work as a guidance system.

Another non-music application is to add verisimilitude to teleconferencing with mobile phones. “In a room, the human brain is designed to separate voices – you never have problems figuring out who is talking when you’re sitting at the same table,” Katz said. “On a phone, all voices sound like they are coming from the same place.” With 3-D sound, you can recreate the whole physical layout of the meeting in front of you while you are sitting on a park bench, he said.

Today, military research is pouring a lot of money into figuring out the best uses of 3-D sound for defense purposes. The Arkamys-CNRS effort is to bring that multimillion-euro research down to a commercial, mass-market use.

—

Add to: | Technorati | Digg | del.icio.us |

Email a friend

Article as pdf (in French)

Article online (in French)

Translated in English…

Ears for everyone, a sound for each individual

Innovation

In partnership with the CNRS, a small French firm has developed a technology that personalises the listening experience on portable audio players.

What if the tiny earplugs of portable audio players and mobile phones were capable of reproducing an acoustic environment comparable to that of an auditorium? That is the objective of Arkamys, a small French firm of 15 employees who, in partnership with the CNRS, has studied the particularities of the human ear. Its project: to perfect a listening process that reproduces natural sensations and is adaptable to each individual.

Currently small earplugs that come with portable audio players give a stereo sound that appears rather flat compared to a Hi-Fi unit. The main reason is they do not produce the same spatial sensation. “Since the sounds arrive directly into our auditory canals, we lose our natural listening habits.” explains Arkamys’ MD, Philippe Tour. Several manufacturers, such as Creative Labs, Sensaura, Qsound or SRS Labs, have tried to resolve the problem by filtering the sound signal to add 3D effects. By simulating parameters such as distance, orientation and reverberation, this sort of treatment enables the listener to situate themselves at the centre of a dynamic auditory landscape, where the sounds come from above, to the right, behind or in front.

The only problem is that these systems are conceived for everybody, without taking into account the physical characteristics of each individual. “It’s as if we were listening with someone else’s ears,” concludes Philippe Tour. Each person has their own auditory perception, dependent upon their morphology and, in particular, the shape of one’s head and the form of one’s ears. Our technology takes these parameters into account in order to propose a more natural, more comfortable and more realistic hearing to the users of portable audio players.

Recreating spatial sensation

With the aid of the LIMSI (Laboratoire d’informatique pour la mécanique et les sciences de l’ingénieur), Arkamys’ engineers strive to measure and model the physical differences which determine the way in which each individual perceives sounds. “Everyone can identify the origin of a sound, situate objects in space and guess their distance,” observes Brian Katz, an acoustics researcher at the CNRS, who is part of the project. What’s more, this is done despite the fact that we have only two ears. “It is our brain which interprets the sound information and recreates the spatial sensation”, explains the researcher. “The small time delay between when a sound arrives in each ear enables the brain to deduce the spot where the sound comes from. The difference in levels between the sounds perceived by one ear and the other provides other spatial information.” In fact, all this information is modified by the attenuations and reflexions due to our height, the shape of our face and the ensemble of all the curves and roughness of our external ear. “All these parameters change the colour, elevation and tone of sounds,” emphasizes Brian Katz. “And they correspond to only one individual because each person has a different morphology.”

While current audio systems merely supply a standard acoustic profile that do not take all these differences into account, Arkamys’ technology will bring a personal quality of hearing to each user, corresponding to his own physical characteristics. “We can simulate several ear shapes to know which frequencies to filter in order to reconstitute with earphones the same auditory environment that you perceive naturally, for example, in a concert hall.” Philippe Tour explains that by the end of the year Arkamys will be able to create individualised acoustic profiles. “With your cellular phone, all you will need to do is take a photo of your face and send it to your server. We will calculate the audio correction coefficient best suited to your morphology. Then we will send it to you in the form of a parameters file. When you start up your audio player, you will only need to select this file in the options menu to have your personalised profile.”

Arkamys, who supplies notably to mobile phone and car manufacturers, intends on presenting its new technology to constructors of audio components, mobile audio players and mobile phones. Its management also envisages contacting telephone operators who could propose this service to their subscribers, either for free or as a paying service. But the future market of this technology could well be mobile television and games on mobile phones.”These services are going to take off this year,” Philippe Tour predicts. “At present, the visual comfort with small mobile screens is limited. By adding personalised 3D sound the user will have a larger spatial sensation that will give the impression of a larger image.”

Didie SRANZ

Add to: | Technorati | Digg | del.icio.us |

Email a friend