Photo credit: Carlos Sillero
Binaural sound is one of those tantalising terms that we hear talked about but rarely experience....yet.
It refers to reproducing sound in headphones (or sometimes speakers) as you would hear it with your own ears. If you close your eyes binaural sound will make you feel like you are in the ocean surrounded by whales, or in the middle of a concert orchestra - it's not just hearing the sounds, it is feeling like you are really there, effectively it is 3D audio.
Of course with gaming becoming more and more immersive, greater focus has been placed on developing the concept for everyday use. With the virtual reality device Oculus Rift targeting XBox compatibility from launch, the game developers naturally want the sound to seem as realistic as possible and to be able to adjust to your head movements.
But how is it possible to recreate sounds as if they were all around you?
In real life we use auditory clues to determine what's around us, including loudness, pitch, reverb rate, and their rise and fall over time. In her article for TV Technology, design engineer Mary Gruszka describes this information as HRTFs, which stands for Head Related Transfer Functions.
HRTFs can be mapped in a number of ways: firstly you can put an individual listener in an anechoic chamber with special microphones positioned at the entrance to the ear canals, though who has the time for this?
Another way is to scan a person's head to get the physical dimensions but the workload this places on even powerful computers is extreme.
The third way, and this can be used by smartphones, is for the person's measurements (eg space between the ears) to be searched against a database for the closest match.
Until recently there was no common file format to store this information which meant resulting databases would be buggy and difficult. However, now AES69 with the file extension .sofa has this year been published as the standard - the result of years of work by a group of scientists in a project called AABBA (Aural Assessment by Means of Binaural Algorithms) which was then supported by AES (Audio Engineering Society).
Piotr Majdak of the Acoustics Research Institute in Austria was one of those scientists. He said: “SOFA does not store audio content like WAV or MP3 does, SOFA describes the filters only.
"Application developers can implement spatial audio algorithms in their applications and use SOFA files for the rendering of virtual loudspeakers. Users of audio software can use SOFA files for the personalisation of audio reproduction."
It means that we might now be seeing plenty of apps including binaural sound. In addition, we've seen the new Stems format for DJing, which divides a track into percussion, bass, synth and vocals, perhaps the next step will be binaural controls.