XMOS firmware release for its XVF3510 voice processor

The two variants - XVF3510-UA for USB accessory devices, and XVF3510-INT for built-in solutions - incorporate purpose-designed algorithms that ensure a clear voice stream is captured from across the room, even in noisy environments and with audio streaming from the device itself. These include:

• Acoustic echo cancellation (AEC): enabling the detection of voice signals even through high-volume audio interference.
• Interference cancelling (IC): working intelligently to remove static point noise and ignoring audio from other devices.
• Noise suppression: removing both stationary and non-stationary diffuse noise sources.
• Automatic delay estimation control (ADEC): adjusting signal latency to guarantee optimal AEC performance.
• Automatic gain control (AGC): tuning output channel level for optimum results.
• System control and parametrisation: monitoring and adapting algorithm control, system configuration and GPIO in real-time.

Requiring just two microphones, the XVF3510 is optimised for audio visual products, smart home devices and appliances, health and fitness solutions and conference calling.

Mark Lippett, CEO, XMOS said: “The far-field voice market is expanding and evolving at great pace, and we need to be able to guarantee that our products can remain integral to rapidly diversifying use cases across the field.

“With the introduction of these two new firmware releases, the XVF3510 strikes that sweet spot between price and performance, equipping engineers with the flexibility they need to drive fresh ideas and create new value for end users.”