Elly Yates-Roberts |
Any businessperson can tell you that conference calls are the best way to collaborate with fellow colleagues, partners and customers around the world. But when it comes to speakerphones and conferencing systems, not all devices were created equal. The quality and reliability of conference solutions is vital to support productive virtual meetings.
Today’s workforce needs audio solutions and technologies that release the potential of employees. For example, advances in microphone pick-up technology can enable conferencing tools to boost the quality and clarity of communication. And while the user simply experiences clear audio with the switch of a button, there is a lot happening behind the scenes to deliver that premium audio experience.
Before getting into the mechanics of what happens within a speakerphone, it’s important to take a step back and consider the seemingly obvious: what do ideal speakerphones do? A meeting room speakerphone has two primary functions: first, it must broadcast remote speech through a loudspeaker; then it must capture speech in the meeting room while conveying it to remote participants. Ideally, the speakerphone must convey speech as intelligibly as possible.
High performing speakerphones use ‘echo cancellation’ to prevent sound from the loudspeaker from returning to the remote listener through the speakerphone’s own microphone. They also use microphone array beamforming to further isolate voices from both ambient noise and reverberation. These arrays use multiple microphones and algorithmic signal processing to capture and convey speech to remote participants with the greatest intelligibility.
In many meeting room scenarios, the presence of background noise and reverberation – such as the moving of paper, rolling of chairs or shifting in one’s seat – can interfere with the sound of people talking, which leads to a reduction in speech intelligibility. To overcome the challenge of noise, it’s important to find a way to ensure that the ratio of speech to noise received by any remote listener is weighted in favour of speech.
When it comes to reverberation, a microphone picking up speech in a room will first pick up the voice – a sound which initially arrives at that microphone directly from the person speaking. This microphone will then pick-up reverberation as a series of additional sounds from the voice being reflected from the room’s walls, ceiling and floor.
The differences between direct and reflected signals arriving at a microphone are a question of time and strength. Reflected sounds arrive later than direct sounds, and they arrive with less energy. When a microphone receives both sounds, the result is a blurring of the signal. If you are on the receiving end when this blurring happens, it will sound as if the speaker were standing in a bathroom.
To ensure a high-quality audio experience, speakerphones need to be sensitive to the direction of speech, as opposed to the direction of the sources of noise and reverberation. This is known as a directional microphone system. These microphones are all designed with a certain ‘pick-up pattern’ which determines how sensitive it is to sound arriving from any specific direction. The most common pick-up pattern is omnidirectional. To produce a directional pick-up of sound from a series of omnidirectional microphones, you need to take advantage of differences in both the level of sound and the time at which it arrives at the different microphones. This type of directional microphone system is known as a microphone array beamformer. The delays in sound arrival make it possible to align the signals so that they are synchronised, and the subsequent summation of these signals increases the output level of the microphone array for a certain direction, while decreasing the output level for others.
An omnidirectional microphone will pick up both target speech and unwanted surrounding noise sources equally, but not when advanced beamforming is used. In this instance, speech arriving from the direction that the beam is pointing will be picked up without any change while sounds arriving from other angles will be greatly reduced.
As discussed, reverberation causes sounds to arrive at a speakerphone with additional delay and from additional angles. This results in a blurring of the signal in time, reducing the intelligibility of speech. An enhanced ratio of speech to reverberated sound is maintained throughout the use of a beam focused on the direction of the target signal. Sounds arriving at an angle, reflected from the room’s surfaces will be lessened.
With the help of microphone array beamforming, speakerphone systems such as EPOS’ EXPAND 80 Series, are equipped with a focused, steerable beam that can optimise individual voices in the meeting room for remote listeners. This level of advanced technology empowers remote teams to collaborate with the same levels of confidence and clarity as if being there in person – something that cannot be taken for granted as we head into a future of hybrid work.
Jesper Kock is the vice president of research and development at EPOS
This article was originally published in the Spring 2021 issue of The Record. To get future issues delivered directly to your inbox, sign up for a free subscription.