Rozenn Nicol has been investigating Spatial Audio technology at Orange Labs for more than ten years. With the publication of this monograph, she aims to promote a better understanding of how binaural technology really "works". Despite its straightforwardness the reproduction of binaural audio with headphones is always impressive - a really convincing 3D sound scene is achieved. This is possible because binaural technology merely mimics the spatial encoding that we use daily when we localize sounds in real life.
Starting from practical issues, concerning what is the real meaning of sound recording and rendering for binaural technology, the underlying theory is then progressively examined. The diffraction of the acoustic wave by the listener's body defines the key concept of binaural technology and can be represented by the associated transfer functions, which are known as Head Related Transfer Functions (HRTF). HRTFs are therefore the raw material of binaural spatialization. It is shown how the spatial information is conveyed through HRTFs, investigating both physical phenomena and auditory localization mechanisms.
Special attention is given to binaural synthesis which consists in simulating the left and right signals as they would have been recorded by a pair of microphones inserted in the listener's ear. This is one of the most well known applications of binaural technology, since it allows one, to create with full control a virtual auditory space for psychoacoustic experiment or virtual reality purposes. by straightforward filtering.
Although binaural technology is a powerful tool for sound spatialization, it should be kept in mind that, as the spatialization is determined by the listener's morphology which is unfortunately strongly individual, the spatial encoding of a sound scene is theoretically valid for one sole individual. The monograph ends with an overview of solutions for adapting the binaural spatialization process to an individual’s variability.
Posted: Monday, April 5, 2010