Audio/video teleconferencing systems
You are here: Solutions -> Teleconferencing systems
Alango has developed a complete set of digital signal processing technologies for building different types of high quality audio teleconferencing systems.
The following problems are common when developing an audio part of a multi-microphone teleconferencing system:
- Long, sometime well above 200ms echo in large, often reverberant conference and meeting rooms. Cancelling long echoes requires long adaptive filters that are computationally expensive, demand large data memory, converge slowly, unstable in double talk and require high computational precision.
- Wide dynamic range of input signals as the distance between the user and the nearest microphone can vary significantly compared to hands-free car kit or personal speakerphone applications
- Air-conditioning, projector fan and other noises leading to a bad signal to noise ratio for distant voices
- Hybrid echo creating a potential acoustic feedback on high loudspeaker volumes. The problem is especially pronounced in integrated devices
- Several people in the same meeting room talking simultaneously posts additional challenge to the automatic microphone mixer algorithm
- Cost limitations restricting usage of powerful, but expensive floating point processors.
The problems above are additionally exacerbated by user requirements of full duplex performance, wide frequency range (8 KHz and above) and excellent sound quality. Alango solution addresses all the above issues. Its block diagram is shown in the figure below.
Alango solution uses distributed echo cancellation principal where each microphone channel is processed independently. Echo cancellation and noise reduction are first performed individually and the mixing of the output signals is done on the last stage together with the automatic gain control. All logic is integrated into one Extended Voice Communication Package (EVCP).
While being a separate product, EVCP is based on the standard Voice Communication Package implementing the same, highly efficient narrow subband processing scheme. EVCP utilizes 32 subbands with 8 KHz sampling rate and 64 subbands with 16 KHz sampling rate with the downsampling factor equal to the number of bands thus proving significant reduction of computational complexity with long adaptive filters. Downsampling also proportionally shortens the length of adaptive filters thus increasing the convergence rate and overall filter stability. Several other methods are used to further reduce the complexity thus achieving the best possible quality under the specific processor and cost limitations.
In the Alango audio-conferencing solution the standard VCP package is used to cancel the hybrid echo. The two packages (EVCP and VCP) are based on the same technology and thus cooperate seamlessly without any howling or other disturbing effects.
Alango technologies are designed to provide a good sound quality in mobile, very noisy environments. Together with the proprietary, patent pending supper high frequency resolution scheme and efficient automatic gain control technology, Alango audio conferencing solution provides excellent voice quality for large conference rooms with people talking from a distance of 3-4 meters from the nearest microphone.
Alango proprietary Intelligent Speech Mixer technology automatically detects and chooses the microphone(s) with the strongest signal-to-noise ratio. When several people in a room are talking simultaneously, the technology may mix signals from several microphones thus ensuring equal participation of all participants without switching between voices and consecutive loss of information.
Below picture shows a large conference room of an Israeli company that purchased and installed Alango multi-microphone system prototype.
The company could not find other suitable solution for its room and asked Alango to develop a proprietary system. This example system operates at 16KHz sampling rate and it integrates:
- 9 communication microphones with 250ms distributed acoustic echo cancellation and noise reductions (separately on each microphone)
- 2 PA microphones
- 1 auxiliary stereo input
- 2 telephones. The telephones can be used simultaneously calling two different numbers and providing conferencing capabilities
The system is powered by one BF533 Blackfin DSP from Analog Devices delivering the best price/performance ratio possible for systems with the above features and quality.
The following picture shows a three microphone technology evaluation system.
This smaller system uses the same technology so sound quality is not compromised. The three microphone system can be sent for evaluation by a potential customer upon signing a simple evaluation agreement. Integrated version with the microphones and the speaker in one body is also available.
Please, contact us for further technical details, licensing and pricing information.
go back to the Solutions page.