World-leading acoustic beamforming technology
Nuance Hearing’s acoustic beamforming technology shows world-class performance in directional capabilities and audio focus.
In the field of ‘real-time’ applications, this is the most effective technology in the world. Our technology is a highly effective solution for dealing with background noise and voice distractions and is in use by the global hearing industry giants.
What is acoustic beamforming
Acoustic beamforming technology is a signal processing technique, capable of focusing on sounds (acoustic signals) coming from a certain direction, relative to the other directions.
Using an array of microphones, an algorithm detects the direction of sounds via time gaps in reception between the microphones.
The algorithm creates a gap in the volume between the sounds from the preferred direction and the background noise.
Directivity Index - DI
The reduction of the overall noise compared to the sound waves from the target direction is measured in decibels (dB), and is defined as the ‘direction coefficient’.
Nuance Hearing was able to develop products with a direction coefficient of 15dB, (the signal-to-noise ratio improves by a multiplier of 30 in acoustic energy).
For comparison, the world’s most advanced ‘real-time’ products display a directional coefficient of no more than 4-5 dB.
Our unique auto-steering technology automatically selects the preferred hearing direction from the one or two sources with the highest acoustic energy.
This allows tracking the dominant speaker even in motion (e.g., a classroom teacher), or automatically switching between speakers in a multi-participant conversation.
Within a few milliseconds, the direction of listening changes smoothly, so that the listener does not even miss the first syllable at the beginning of the sentence.
Speech intelligibility test and SNR
For each product we develop, we perform a speech intelligibility test. We use a well-known measure – SNR50, to demonstrate the effect of our technology on speech perception in noisy environments.
SNR50 is a measure of the ratio between the signal and the noise (SNR – signal to noise ratio), allowing the listener to achieve 50% word recognition. In other words, we lower the volume of the signal compared to the noise, until a listener can only recognise 50% of the words heard over the background noise. This point is their SNR50. These tests take place in a sound-attenuating room, where listeners hear target speech coming from the front, and background noise coming from each side. The background noise is comprised of two competing voice sources (one on each side) constitute the background noise. The volume of the noise is kept constant throughout the experiment, and the target signal volume is reduced adaptively to the listeners response.
Our test results
In the experiment, we tested SNR50 in two conditions – with and without our device. For example, adults with normal hearing typically show an average improvement of 8dB in SNR50 with the device. That is, if they can achieve 50% recognition of words in -4dB SNR (the noise is 4dB louder than the signal) without the device, they reach the same performance level in -12dB SNR (the noise is 12dB louder than the signal) with the device. When the speech sounds are more audible in the presence of noise, it is easier for our brain to identify and recognise the spoken message.
Ultra-fast signal processing
Nuance Hearing’s hardware and software platform captures sounds, processes them, and plays them at a record speed of 6.25 milliseconds.