INDUSTRY & PRODUCT NEWS


COMSOL Conference 2017 - Call for Papers and Posters
The COMSOL Conference 2017 Boston will be held October 4-6, 2017, at the Boston Marriott Newton hotel. The program committee of the 13th annual COMSOL Conference is inviting members of the COMSOL community to submit paper and poster abstracts on their numerical simulation work and custom applications. Following the Boston, MA, event, other 2017 conferences take place in Rotterdam (Oct 18-20), Beijing (Nov 2-3), Singapore (Nov 22), Taipei (TBA), Seoul (Nov 17), and Tokyo (Dec 8).  Read More


Sonos Introduces PLAYBASE Under-TV Home-Theater Sound Solution
Sonos has introduced PLAYBASE, the latest addition to its wireless home sound system and the newest speaker in its home-theater lineup. Inspired by the simple insight that up to 70% of all TVs stand on furniture and are not mounted to the wall, PLAYBASE is a thin, yet powerful speaker that sits discreetly under the TV. Priced at $699 US and available in both black and white, PLAYBASE will be available globally starting on April 4, 2017.   Read More

New Programming-Free LCD User Interface for Embedded Applications
How many times do audio product designers hesitate to provide a display and a graphic user interface (GUI) on their new solutions, just because they fear the complexity and want to get to market faster? That's precisely why LCDTERM, a subsidiary of Best Circuits, created a solution for a programming-free LCD user interface, which is able to provide a simple connection to embedded systems, and quickly be up and running.   Read More


Bruel & Kjaer Introduces Binaural Microphone Recording Solution
With virtual reality, augmented reality, and a renewed interest in binaural audio recording for broadcast and new media applications, the audio industry has recently seen several product announcements from leading manufacturers. Now, Danish test and measurement specialists Bruel & Kjaer have announced a new binaural microphone set, which will be presented for the first time at CanJam SoCal, April 8-9, 2017.    Read More
 
WW Speaker Cabinets Introduces 21" Subwoofer for DIY Home Audio Market
WW Speaker Cabinets introduced its new X21 subwoofer to the DIY market. The X21 is the first commercially available home audio DIY subwoofer cabinet designed for 21" drivers. X21 subwoofer cabinets are lock mitered, allowing WW Speaker Cabinets to offer finished cabinets with a choice of veneers, stains, and lacquer finishes as well as raw MDF. All X21 subwoofer cabinets are customized for specific individual driver models and performance data is provided for each driver.   Read More


VB-Audio Expands VBAN Audio Streaming Capabilities with VBAN-Receptor
VBAN streaming audio protocol is an Open-Source project, developed by Vincent Burel, which allows streaming up to eight channels (7.1) of pulse code modulation (PCM) audio at up to 24-bit/96 kHz on any LAN or WLAN network, including wireless. VB-Audio already offers a family of software solutions with VBAN support and now announced availability of VBAN Receptor, a simple application to receive and listen to VBAN Audio Streams.  Read More


AES Technical Committee Group Formed to Explore Audio for New Realities
The Audio Engineering Society's Technical Council (AESTC) has formed a new group to advance the science and application of "Audio for New Realities." The group will serve the needs of the audio community working in virtual, augmented and mixed reality environments. "New realities" covers a whole gamut of application areas including film, games, music, communications, medicine, forensics, simulations, education and virtual tourism.   Read More


Sennheiser's MobileConnect brings Assistive Listening over WiFi to Students' Smartphones
As part of its ongoing commitment to inclusion, Sennheiser is advancing the availability of assisted listening solutions for universities, with its innovative MobileConnect solution (www.audioxpress.com/article/sennheiser-presents-mobileconnect-for-live-audio-streams-to-smartphones-at-infocomm-2016). By offering the ability to live stream audio via WiFi to a user's own device, the system is ideally suited to academic institutions wishing to quickly and cost effectively enable access to education to hearing impaired students.   Read More





João
Martins
Editor-in-Chief




Editor's Desk


Is Voice Our Best Friend?

One of the good things about an event such as Mobile World Congress  (also read our report on MWC 2017 from last week) is that we are able to meet with multiple technology and product development companies focusing on things such as mobile audio, wearables, and headphones and earbuds - which seem to be a more exciting trend at this formerly telecom-centric event than anything else. As I expected, following the trend from CES this year, voice personal assistants (VPAs) and voice recognition were all the rage. After all, it started with smartphones - better said, with Apple Siri on the iPhone.

During our meeting with DSP Group at MWC 2017, we got a better understanding of the multiple solutions the company is developing to implement VPAs and voice recognition in all types of devices and solutions for many different applications.

Siri wasn't the first voice interface - Apple acquired the technology in 2010. Siri includes contributions from several pioneering companies, including Nuance and One Voice Technologies, which was the first truly Artificial Intelligence (AI) based Natural Language Processing (NLP) engine, introduced in 1999. One Voice's IVAN (Intelligent Voice Animated Navigator) was part of the patent portfolio acquired by Apple in 2010.
 
Now, everyone is talking about Amazon Alexa and Google Assistant because of the notoriety of Amazon Echo and Google Home, and the fact that those supposed VPAs are integrated in what are basically... loudspeakers. Important to note, that Alexa is essentially a voice recognition system that allows commands, user interaction, and control of things, using (more or less) natural language commands in English. Google Assistant is the same, only slightly better in terms of voice recognition and natural language abilities, with an expanded range of queries, due to its direct connection with the Google search engine.
 
Alexa can compensate for its limitations with its ability to support new skills, which can be integrated by third parties using Alexa Voice Service (AVS) development tools. The only truly "intelligent personal assistant" is basically Siri, which uses artificial intelligence to generate interactions, (like IBM Watson allows), and has the ability to "learn" and improve over time. Alexa and Google have VPAs that basically recognize voice commands. There's also Microsoft's Cortana, SoundHound, and several other voice engines, all trying to become true VPAs with natural language.

The trick here was that both Amazon and Google's platforms - designed over speakers - can capture far-field voice much better than any current Apple device, including the latest generation Mac computers. And for some reason, people who use Siri, don't like having to pick up their iPhones and talk as they would do for a conversation with another person, when they want to ask Siri something. They always do it in speaker mode. Naturally, because you use voice assistants mostly when your hands are busy with something. That's something that the industry itself has learned and that Amazon was very clever to explore.
 
The other aspect of the entire voice recognition phenomena - and the one that explains why it is such a big deal at an event such as MWC - must have to do with its potential for home automation and eventually interactions with Internet of Things (IoT) devices and wearables/hearables. Some manufacturers of truly wireless earbuds are now promising users to soon integrate Alexa. The problem is that those tiny earbuds can hardly keep up with wireless transmissions, much less be able to power a voice engine that requires always-on listening for a wake-word. Of course, they will have to replace the wake system with the touch of a button, but then... that means you can also use your hands and fingers to do the same on the smartphone, smartwatch, or any other wearable.
 
A STMicroelectronics and Sensory voice recognition development kit contains up to seven microphones using beamforming and echo cancellation.

So, it is only when we meet development platform providers such as DSP Group, Cadence, Infineon, STMicroelectronics, NXP, Qualcomm, and many others, that we realize the huge challenges associated with the "voice trend." First and foremost, OEMs need to really reflect if and why they need to offer voice assistants in their products. Not surprisingly, at MWC 2017, the trend was to offer simple self-contained solutions that simply respond to basic voice commands - directly embedded in a low power platform and without the need for Internet connectivity and cloud services. And the interesting thing is that those basic and lower cost systems - which can be integrated in anything from toys to washing machines - can actually use one or more commands to send a request to another connected device (e.g., an Amazon Echo), creating a remote voice access (we've seen the demo). The DSP Group is currently developing that solution with different MEMS and signal processing partners, but basically the company showed us that it is able to offer all sorts of platforms for home, mobile, IoT, and so forth - with any available VPA service.
 
The really interesting thing, as I've noted in my previous MWC 2017 report in this space, is how this trend is actually pushing the audio technology forward. Because in order to have better Amazon Echo-type products, we need better portable speakers that are able to provide powerful sound anywhere (people will use them as a music listening devices in the kitchen and bedroom), and feature room correction DSP adaptability to compensate for variable acoustics (the DSP is needed anyway to enhance voice recognition from background noise).
 
Above all, we need much better microphones, able to capture voice from a distance, which means beamforming arrays of new generation MEMS (with digital and analog front-ends, depending on the product type) and very sophisticated technologies in terms of voice processing and audio recognition. During MWC 2017, Bragi CEO Nikolaj Hviid showed us an app that is able to "fingerprint" a person's voice and completely ignore all others. This could enable Bragi's The Dash wireless earbuds to receive commands from a single user (owner) in the middle of a busy office environment. But the potential of the technology is much larger...

Qualcomm was showing solutions for voice recognition and VPA integration at MWC 2017, and the company is currently expanding its R&D efforts in this application segment.

Infineon, NXP, STMicroelectronics, Qualcomm, and many others are currently working on all fronts of this effort because of its huge potential, which Gartner  calls "an explosive new market," already worth $360 million worldwide in 2015 but exceeding $2.1 billion in 2020. Gartner even predicts that as users discover new uses for VPAs - other than check the weather forecast or starting the music - the services could have the potential to become truly ubiquitous and serve as complete user interfaces for some apps.
 
Another topic at MWC 2017 was the use of voice recognition to power language translation and voice-to-text engines. The idea is fascinating, but (sorry guys...) it will take some time and lots of developments for that dream to become mainstream. Would you risk talking to anyone, knowing what you say is Google-translated? Guess not.
 
Also, remember, why do we need voice interactions. Certainly, not to replace things we can quickly do with a simple gesture, like turning on the lights. In fact, the on/off switch might be considered the largest obstacle in the way of voice technology :) Voice is only an attractive alternative when we are not able to move or we have our hands tied up with something else (or dirty). No wonder people use Siri the most in the car. And automotive applications are precisely where we will see all these efforts converge and dramatically evolve. For the moment, simply to interact with assistive technologies and infotainment systems. But very soon, they could also serve as an interface for autonomous vehicles. After all, what do you do when you get inside a taxi?

Fresh From the Bench
Measurement Microphones from PCB Piezotronics
By Rick Spencer
 
Stuart Yaniger reviews a complete series of professional-grade measurement microphones from PCB Piezotronics, specifically, the 378A04 free-field microphone, the 378B02 free-field microphone, the 378A21 diffuse-field microphone, and the 130E20 array microphone, all using a current source rather than a voltage source, called ICP (ICP is a registered trademark of PCB Piezotronics and is a PCB sensor containing built-in electronics that can be powered by 2 to 20 mA of constant current power.) As Yaniger explains, "It's well known that technologies designed for top-of-the-line equipment often migrate down and improve the performance of lower-cost products. What's interesting is when the opposite occurs-a less complex and expensive technology is optimized to punch out of its weight class. In this case, it's the upward migration of electret microphone technology to the hallowed halls where polarized condensers once ruled." This article was originally published in audioXpress, December 2016.   Read the Full Review Now Available Here

Voice  Coil Test Bench
Tang Band RT-2202S Ribbon High-Frequency Transducer 
By Vance Dickason
 
Tang Band (TB) Speaker sent us a ribbon transducer, a small diaphragm ribbon that incorporates a matching transformer. The device is a closed-back ribbon tweeter with a fairly small 30 mm × 10 mm ribbon diaphragm. Features include a thin aluminum foil diaphragm, an injection-molded face plate, a metal chassis, Nd-Fe-B (neodymium) magnet, a matching transformer, and a pair of gold-plated solderable terminals. The RT-2202S is a rated 80 W maximum power handling (8 W nominal). It is worthwhile to note that the assembly is not airtight and requires a small enclosure unless the device is mounted in free air. This is intended for hi-fi and high-end applications and it needs an individual enclosure space or an open air design to avoid potential damage airflow from other drivers. The driver is able to generate an SPL of 91 dB/w/m, ideally tuned to playback frequencies above 2 kHz, all the way to 35 kHz. This article was originally published in Voice Coil, April 2016.   Read the Full Article Online

AX March 2017: Digital Login
Audio Product Design | DIY Audio Projects | Audio Electronics | Audio Show Reports | Interviews | And More 

Don't Have a Subscription?
VC March 2017: Digital Login
Industry News & Developments | Products & Services | Test Bench | Acoustic Patents | Industry Watch | And More