Voice is being rapidly adopted as a human interface technology, enabling common consumer audio devices to become “smart,” as they now have a low-cost way of providing the clearest voice signal to voice recognition services such as Google Assistant. What has been not working so well is the capability to recognize voice inputs from a distance (far-field) when the speakers are turned up to full volume. DSP Concepts’ TalkTo voice capture technology, introduced in early 2019, is built on its Audio Weaver platform, and offers advanced voice extraction capability, now powering a new wave of smart audio products, as seen at IFA 2019.
To deliver cutting edge audio features on Braun Audio’s line of hi-fi smart speakers, DSP Concepts closely collaborated with StreamUnlimited, Pure, and renowned audio engineer Karl-Heinz Fink and his team. As DSP Concepts CEO, Chin Beckmann highlighted, “This collaboration reinforces the value of the Audio Weaver platform and how proliferation of our platform is resulting in customers being able to get to market faster with best-in-class audio products.” Karl-Heinz Fink, who actively develops on Audio Weaver, stated, “Historically, audio engineers have wasted precious time jumping through hoops to get their ideas implemented. Now, thanks to Audio Weaver, we can focus on sound quality and ultimately develop a superior audio experience in significantly less time.”
Always-On Voice Neckband
DSP Concepts also announced the fruition of its partnership with Ambiq Micro, a provider of ultra low power processors for hearables and wearables, and Sensory, a pioneer in applying artificial intelligence to embedded voice recognition applications. The collaboration among the three companies enabled Shenzhen Horn Audio to create and deliver the Luduan V4 white-box headset, the world’s first Alexa Mobile Accessory (AMA) certified consumer device, which was highlighted at the Amazon Alexa exhibit, showcasing new white-box solutions, development kits, and ODM solutions.
“The TalkTo voice front end enables ODMs and OEMs to easily and cost effectively integrate a voice user interface with the best-in-class user experience,” says Chin Beckmann, CEO, DSP Concepts. “The Voice-on-SPOT reference design used by Horn Audio exemplifies our ability to address the ultra-low power market, where processing capability is limited. This technology extends beyond headphones, to billions of Internet of Things devices, where a voice interface becomes the only practical human machine interface (HMI) solution.”
Powered by DSP Concepts’ Audio Weaver platform, TalkTo is a suite of advanced microphone-processing algorithms that use advanced beamforming and dynamic, environment-aware noise-suppression to deliver a best-in-class voice experience while also providing maximum power-savings for any given acoustic environment. Audio Weaver features a unique, hardware-agnostic audio-processing engine with 400+ audio building blocks, and a powerful, graphical design environment that allows product-makers to efficiently create, integrate, and update audio features in consumer and automobile devices.
“By using a TalkTo voice-front end for Audio Weaver, ODMs and OEMs are getting more than just a state-of-the-art voice user experience – they’re also getting a highly efficient means of updating and maintaining a product’s audio feature-set over time. This capability is highly valuable in the rapidly evolving world of low-power consumer audio devices,” commented Dr. Paul Beckmann, CTO and Founder, DSP Concepts.
Braun Audio | Pure Audio | StreamUnlimited | Fink Team | Shenzhen Horn Audio | Ambiq Micro | Sensory