The Acoustic Echo Cancellation (AEC) aims to achieve high-quality full duplex communication for mobile phones, Bluetooth earphones, hands-free kits, speakers, and other types of voice terminals.

Adaptive beam-forming is used to locate the direction of the target speaker, while protecting the main direction speech and removing non-stationary interference such as voice and television noise in the environment. The suppression of non-stationary noise is greater than 20dB, which can significantly improve the far-field speech recognition in noisy environments.

When the wind blows onto the surface of the microphone, it produces "wind noise". Wiping the surface of the microphone can also produce similar noise. Our wind noise suppression solution can automatically detect wind noise and significantly attenuate wind noise energy, while fully preserving clear speech.

Noise suppression can suppress stationary and transient noise in speech signals, thereby improving signal-to-noise ratio, improving speech intelligibility, and reducing hearing fatigue. The characteristic of Sound+ technology noise suppression technology is the ability to quickly and accurately track environmental noise levels with high speech quality.

Traditional noise suppression based on DSP is an effective method for speech enhancement, suitable for various device types and call environments. However, non-stationary noise is difficult to suppression using traditional methods, such as some obvious noise (tapping or wind), which is still difficult to suppress. The DNN denoising algorithm developed by Sound+ Technology can effectively suppress non-stationary noise while providing excellent speech quality.

The automatic gain control algorithm enables the microphone's gain to automatically adjust with the strength of the input signal, thereby maintaining the optimal level of effective speech sound. For example, when people speak loudly, the automatic gain control algorithm can adjust the output gain signal to an appropriate range, so that the output signal does not saturate. When people speak softly, output will be amplified by automatic gain control algorithms to ensure that the output signal can restore the dynamic range and details of the original sound.

Multi-band EQ adjusts output frequency response of microphone and speaker slightly to meet customer’s needs.

The adaptive de-reverberation technology of Sound+ Technology can detect and suppress reverberation speech, while retaining speech to make it easier to understand, avoiding self masking and overlapping masking effects caused by reverberation, which can affect speech intelligibility.

The Voice Activity Detection (VAD) technology of Sound+ Technology can reliably detect the presence of voice activity. The sensitivity level of VAD is adjustable, ensuring that false alarm detection is minimized in the background noise and other non speech conditions.