“Barge In” is the name of the algorithm that allows understanding of voice commands on the smart speaker when it is currently playing music.

In a basic Smart Speaker system, voice commands are difficult to understand when music is playing. This is because the microphone picks up both the music and the voice. Typically, the loudspeaker is closer to the mic than the voice source. The music signal overpowers the voice signal, impeding the ability to understand the voice command.

The smart speaker market is poised to grow by 14.35 billion by 2022. The potential demand for smart speakers could be in the billions of units, placing a smart speaker in every room in a home, an office, hotel, classroom, school, and hospital bed. Juniper® research has forecasted that 220 million smart speakers will be used to make calls to landlines and mobile numbers by 2024. For all of these reasons, we developed a cost-effective method to take a new standard of perfect sound to market.


A Typical Barge In algorithm uses a DSP to subtract or cancel out the music so only voice is captured. This method works only if the music captured by the microphone perfectly matches the music delivered to the loudspeaker. In reality, the loudspeaker does a poor job of reproducing the music thus creating a modified music signal. Because of this, the cancellation of the music is poor.

typical barge in

The subVo diaphragm tracking sensor provides a voltage signal directly proportional to the diaphragm position. This allows the system to have a feedback signal that more accurately reproduces what the loudspeaker is emitting. There are two methods that use the subVo sensor to improve Barge In performance.


METHOD 1 uses the subVo sensor as the music feedback signal into the DSP. The subVo sensor accurately captures the modified music signal produced by the loudspeaker. Using the subVo sensor signal as the subtraction signal properly cancels out the music, revealing only the voice signal.


METHOD 2 uses the subVo sensor in a feedback loop around the loudspeaker to force the loudspeaker to produce the original music so cancellation can occur as expected.