About This Application Note

The VAD Reusable Subsystem Application Note contains a brief description and tuning instructions for DSP Concepts’ voice activity detector (VAD) Reusable Subsystem.

VAD – Voice Activity Detector Reusable Subsystem

VAD is DSP Concepts’ proprietary voice activity detector algorithm. It is available as a Reusable Subsystem in AWE Designer.

The VAD algorithm compares a voice signal level with a stationary noise level and detects voice activity. The VAD algorithm can work with various kinds of microphone input signals, including air-conduction microphones and bone-conduction microphones. The VAD’s input signal must be mono at a 16 kHz sampling rate and with a 160 sample (10 milliseconds) block size.

Example

Figure 1 shows an example layout demonstrating the VAD Reusable Subsystem. This example layout is designed to run in Native Mode with a file input; audio signals must be assigned to appropriate channels for running the layout on a target device. VAD should work independently from the signal level, but it is recommended to adjust the signal level (using Scaler1 in this example) to about -30 dBFS for optimal performance.

Tuning

The VAD Reusable Subsystem includes an inspector which allows users to adjust the voice detection sensitivity. Figure 2 show the inspector panel of the VAD Reusable Subsystem. The VAD sensitivity ranges from 0 to 1, with 0 yielding rare voice detection and 1 yielding very frequent voice detection. DSP Concepts recommends using 0.5 for optimal detection.