BIGSPEED Voice Compression SDK is a simple, low-complexity, low-cost voice engine for quick development of VoIP applications. It incorporates audio recording/playback and voice compression/decompresion in a single COM automation object, along with some advanced voice processing features.
The used audio format is wideband, 16kHz, 16Kbps, PCM. The compression ratio is adjustable and can vary from 4:1 to 50:1.
A lot of bandwidth can be addionally saved thanks to the silence compression. A voice activity detector (VAD) with an adjustable threshold level is used to distinguish between regions with normal speech activity and those with silence or background noise.
An optional echo suppression can be employed thanks to the robust double talk detector. When a double talk situation is detected, the output signal is faded (half-duplex).
An automatic gain control (AGC) system is implement with a dynamic range 10:1. A noise reduction is performed on the microphone signal to decrease the white noise.
BIGSPEED Voice Compression SDK utilizes an elegant event-driven paradigm for easy integration into Windows applications. A special attention is paid to provide responsive user interface while maintaining low CPU usage. Carefully designed set of properties, methods, and events offers great flexibility to develop innovative communication software. Sample applications are included in Visual Basic .NET, and Delphi 7.