Microsoft has announced a set of features that should end up improving the audio experience for your Microsoft Teams video and audio calls. High-fidelity music mode with automatic music detection is one, and the rollout of machine-learning-based noise suppression is the other.
Now available to everyone is machine-learning-based noise suppression. This feature will consider any non-speech signal picked up by a microphone as noise, and then suppress it. Related to that is the second feature, automatic music detection, which was previously teased by Microsoft on the Microsoft 365 roadmap in July.
Though it won't be coming to Teams for another few months, this feature makes sure that everyone on a Teams call can hear music clearly. For end-users, Teams will show a pop-up when it hears music to allow users to enable high-fidelity music mode and tweak the audio signal for it accordingly. The first 30 seconds in the clip below is what music sounds like in a call without high-fidelity music mode, and the last 30 seconds is what it sounds like with it turned on.
Microsoft studied more than 1,000,000 audio clips which contain speech and music as part of coming up with his feature. It also evaluated the model with an independent test of 1,000 additional audio clips crowd-sourced from a wide range of contributors. In testing, Microsoft was able to detect more than 81% of all music clips, setting a 0.1% false positive.