![](https://cdn.xingosoftware.com/audioxpress/images/fetch/dpr_1,w_765,h_430,c_fit/https%3A%2F%2Faudioxpress.com%2Fassets%2Fupload%2Fimages%2F1%2F20230307133436_XMOS-Voice-Processor-generic-TWeb.jpg)
The new XMOS XVF3800 voice processor was designed to deliver a high-performance, cost-optimized, rapid time-to-market solution for enterprise and consumer voice conferencing platforms. The four-microphone processor is intended to enable Microsoft Teams and Zoom certification in conferencing devices, soundbars with built-in cameras (aka videobars), portable speakerphones, PCs, gaming monitors and collaborative display solutions.
The solution offers state of the art algorithms to provide high-quality audio in challenging environments, while the processor itself enables developers to fully optimize performance based on the specific acoustics of the product at hand.
Key features include acoustic echo cancellation (AEC) to remove far-end echoes picked up by the microphones, which enables the XVF3800 to extract clear voice signals even with high-volume speaker output, combined with adaptive multiple beamformers to track both stationary and non-stationary voice sources, identifying the direction of arrival, and isolating the voice of interest, improving double talk and dereverberation performance.
As mandatory in these designs, the XMOS XVF3800 also offers noise suppression (nulls diffuse noise sources such as air conditioning, road noise etc.) to enable accurate, consistent voice capture, and automatic gain control (AGC) to ensure a consistent far-end volume for speakers at different distances from the device.
Beyond the specialized microphone interface and two supporting linear mic arrays, the XVF3800 also incorporates USB 2.0 connectivity, as well as an I2S audio interface that can sample at 16kHz or 48kHz.
![](https://cdn.xingosoftware.com/audioxpress/images/fetch/dpr_1,w_765,h_430,c_fit/https%3A%2F%2Faudioxpress.com%2Fassets%2Fupload%2Fimages%2F1%2F20230307133436_XMOS-XVF3800-evaluation-kit-TWeb.jpg)
The XVF3800 is currently available to early access customers and will be generally available, along with a development kit, the XK-VOICE-SQ66 (pictured above), in the second quarter of 2023. The new, cost-optimized processor builds on the success of XMOS’ high performance VocalFusion range.
"The modern workforce has more disparate and flexible work environments than ever before, and so has never had higher expectations for its communications tools," commented Aneet Chopra, EVP Marketing & Product Management, XMOS. "With this in mind, enterprise and consumer software providers like Microsoft Teams, and Zoom, are raising the bar to match these expectations. Turnkey solutions that are high performance and easy to integrate are needed to keep pace with new innovations and experiences."
XMOS is showcasing the XVF3800 demo during Embedded World 2023 in Nuremberg, Germany (March 14-16).
www.xmos.ai