All conferencing is a substitute for face-to-face conversation. When talking in person, we can interject, we can analyze tone and body language, and we tend to interrupt each other. This is natural to us; this is what we expect.
Traditional telephone systems can hinder this natural style of communication. When you’re holding a handset to your ear or you have a separate earpiece and microphone that are completely sound isolated from each other, phone systems use “full multiplexing,” meaning both sides can transmit and receive sound simultaneously. With a mobile phone in speakerphone mode, when you’re speaking into it and the other person starts talking at the same time, the other person’s voice will not come through because it would cause an echo. This is the most rudimentary form of echo cancellation, and it is accomplished by disabling multiplexing. When tabletop speakerphones were first introduced, this method was used to control echo.
To allow the most natural form of communication, both the microphones and the loudspeakers must remain on at all times, with signals able to move from both sides of the conference at any time. Leaving everything on creates echo paths. An echo path is the travel of sound such that when someone speaks and it is received on the other end of the conference, it is then picked up again by that far side audio system and sent back to the person who originally spoke. This is extremely disruptive and makes effective communication virtually impossible. Gating signal paths on and off is one way to deal with this, but it is very unnatural. The best solution is one that stops or greatly reduces only the echo path, while allowing intended communication to continue unimpeded.
Acoustic Echo Cancellation (AEC) eliminates the echo path without removing other sounds that are occurring in the room. For example, if one person is in a room speaking into a microphone and communicating with a second person in a remote location, echo will happen if no mics are gated. With AEC on the line, this echo is reduced or eliminated. Fundamentally, AEC was developed to make distance communication mirror face-to-face communication as closely as possible.
AEC’s benefits can be difficult to quantify. “Improved communication: $3” does not appear on a financial statement. However, with AEC, people do not have to take turns speaking, which helps a conversation flow more naturally. Intelligibility and clarity of communication improve significantly, and mental fatigue is reduced as a result. Thus, AEC also increases productivity. All of these factors are part of AEC’s position as a value-added technology; it applies a sophisticated technological solution to a decades-old problem.
Effective echo cancellation is not an immediately obvious process inside DSP devices. When discussing what makes Biamp AEC cool, one of the major factors has been our team of audio experts who can develop sophisticated methods of echo cancellation. The members of Biamp’s DSP team have decades of combined experience. In addition to working on new projects, the DSP team continually studies AEC to develop improved functionality. Many of the elements that make AEC “cool” do not show up on a spec sheet. These factors include clean, crisp, natural sound; as well as lack of any remnant of echo getting through.
With DSP, nearly any type of filter is possible. “Filter” refers to anything that stops or reduces a specific signal from getting through. Volume control is one common example, because it reduces or increases all sound across all frequencies, thereby filtering the amplitude. With static processing, a specified parameter is set permanently – you have to manually change the setting to get a different level. Dynamic processing refers to a processing parameter that changes in real time without human intervention. In this case, if certain specified conditions are satisfied, a parameter will change from one amount to another in a predetermined way. Sophisticated dynamic processing is possible using multiple layers of different conditions and parameters that will respond. Traditional auto mixers, gating mixers, noise gates, compression, and similar tools use this type of processing.
Biamp’s AEC uses adaptive filtering technology. Adaptive processing creates an acoustic model of the room. Our processing that utilizes adaptive processing uses all active mic signals to establish that acoustic model. The model can change based on a number of factors. If there are two people in a room or three, if a door opens or closes, if another person joins, if an object or a microphone is moved, if a person speaking leans closer to a microphone, that will change the model of the room with respect to the electronic equipment that is there to capture and reproduce the audio. Adaptive processing takes all of that data, creates a picture, and processes according to the current picture. When the picture changes, the filter parameters are converged onto the best set of parameters, changing in real time in response to the model of the room for the most ideal processing.
All of these factors are important to AEC. Though the distance between the loudspeaker and the microphone is fixed, the signal from the loudspeaker is bouncing off the wall, table, ceiling, and other surfaces. When the number of people in the room changes, other sounds are added, or any other change occurs, static or dynamic AEC would likely allow echo through because the echo paths are changing and the filter can’t respond to them effectively. Our adaptive AEC is able to change its model of the room and converge again at a rate of up to 100dB of cancellation per second. It is constantly up-to-date on what is happening in the room, has an accurate model of the room, and can remove as much of the echo path as possible. There is no set standard to the definition of adaptive. However, this sophisticated method of DSP is a big part of what makes Biamp’s AEC so cool and beneficial.
To appreciate the value of AEC, one must hear it in action and compare it to other systems in order to gauge the difference. Demos are vital for appreciating this difference. Overwhelmingly, people who listen to Biamp AEC compared to other forms of AEC will choose Biamp.
One Response to The Coolness of AEC
This post is regarding the Tesira line of products.
I understand that the owners of Biamp also own Roland pro audio. We run into applications where the customer would like a to utilize an audio mixing console in larger auditoriums instead of a biamp canvas software console. Has biamp considered collaborating with Roland to come up with a series of 16, 24 and 36 channel AVB compatible desktop mixing consoles that could be used with a tesira or tesira forte’ system to provide the customer with a hands on console that would provide more of a live feel for mixing audio in their auditoriums or conference rooms?
Comments are closed.