Speech Processing

Products and IP to acquire, reproduce and process speech signals

Communications that simply sounds clear

With a wide range of products and IP to process speech signals, improve their quality, through best-in-class acoustic echo canceller, line echo canceller, noise reduction and other specific processing blocks.

bdSound products and IPs for speech signals support the most common processing hardware platform and operating system, hence they can used on embedded systems, mobile and personal computers

A complete solution for Speech Quality Enhancement

S2C - Simply Sounds Clear

S2C is a complete product to improve quality of communications, cancelling echo and noise, processing the quality of speech signals and managing their loudness, for a crystal clear and comfortable audio experience, even in harsh and noisy environment.

bdSound’s Acoustic Echo Cancellation engine is a true full-duplex, ultrawideband up to 48kHz, high quality, robust and reliable echo cancellation engine. S2C integrates bdSound AEC engine and other processing blocks, including Noise Reduction, Dynamic Speech Processing, and more, to provide a complete and easy to use solution. 

An optional Line Echo Canceller, for analog two-wires application, can be integrated within S2C.

S2C can be integrated at application level or as ALSA plugin in Linux systems.

S2C ready to be integrated within PJSIP

S2C-PJ Simply Sounds Clear for PJSIP

S2C-PJ is a specific version of S2C designed to work easily with PJSIP, the very known SIP Stack from Teluu. S2C is the newest version of our legay product bdIMAD-pj, used everyday by thousands of end users. S2C-PJ simply replaces the PJSIP sound device with an enhanced one embedding the S2C speech processing.

Using S2C-PJ is straightforward since it is required to modify the original PJSIP file config_site.h as follows:

#include "config_site_sample.h"
#define PJMEDIA_AUDIO_DEV_HAS_PORTAUDIO 0
#define PJMEDIA_AUDIO_DEV_HAS_WMME 0
#define PJMEDIA_AUDIO_DEV_HAS_COREAUDIO 0
#define PJMEDIA_AUDIO_DEV_HAS_BDIMAD 1

 

A complete solution for Speech Quality Enhancement

S2C-A - Simply Sounds Clear for Automotive

Automotive applications are something special and require special attentention. bdSound has developed S2C-A, the Simply Sounds Clear version for automotive.

S2C-A includes all the S2C technology with full-duplex Acoustic Echo Canceller, Noise Reduction and all the other processing blocks. Also, due to the specific application, it includes a special noise reduction processing block to improve speech recognition success rate like Apple Siri® even in very noisy super sport cars. It supports also multiple microphones and beamforming.

S2C-A has successfully been certified Apple® CarPlay® and Android™ Auto on several car models, from sedane to dream sport cars.

Using multiple microphone to improve ASR performance, reject noise and reverberation

Multiple Microphones

Single microphone solution has been widely used in a lot of application and will be still used in the future. With two or more microphones it is possible to significantly improve performance of any communication application, at home, in car, at office.
Automatic Speech Recognition, like Siri or Amazon Echo, may suffer from an excess of noise and reverberation. bdSound’s technology manage multiple microphones creating a set of virtual microphones that dinamically adapt theirself to the room acoustics and to background noise and reverberation, to better record the desired talker signal and reject background noise or other noise sources.

Supported Processors

ARM Cortex A family
ARM Cortex M family (M4 and M7)
ARM Cortex R family
x86/x64
FPGA
Xilinx Zynq
Altera (Intel FPGA) SoC

Supported Operating Systems

Linux
Android
Apple iOS
Microsoft Windows
Apple OSX
RTOS
Bare Metal

Supported Platforms

Desktop
Embedded
Mobile (Android and iOS)

Our main references

 
 

Contact us