In the world of affordable multi-room audio, Sonos is still arguably king. That’s despite increased competition from the likes of Denon’s HEOS and Yamaha’s MusicCast, as well as a whole bunch of other manufacturers all offering dedicated multi-room speakers. The company has also had to contest with the new fad on the block – smart speakers.
Apple, Amazon, Google and Microsoft have all announced smart speakers in the last year or so, and they all promise strong sound quality alongside the ability to control smart home devices using vocal commands. Amazon and Google also offer the ability to control multiple speakers in a group using nothing but voice control – seen as an encroachment onto Sonos’ patch. To help combat the growing threat from these smart speakers, Sonos is planning to add the Alexa AI to its entire range of multi-room speakers.
The Alexa AI is typically found on Amazon’s Echo line-up of speakers, but has gradually been finding its way to third-party devices. Already Alexa can be found on everything from refrigerators to lamps, but the Sonos implementation has been a long time coming.
While the Amazon Echo doesn’t exactly have the best sounding speakers in the world, users have been opting for it thanks to the smart AI that can be found within. Sonos’ thinking is that by baking in Alexa support, consumers will still opt for its speakers, which promise better sound quality, while still getting the added smarts that come with Alexa.
“As the world and our homes become more connected, we realise that our customers may prefer to control their Sonos Products by means other than their Sonos app, for example by using a voice enabled product (why not Amazon Alexa), through a home control mechanism (e.g. a Lutron Pico remote), or through the app offered by their favourite music service.
“To enable this function, you will be prompted to allow such devices to connect with your Sonos system (similar to the process you go through to connect a music service). Once this function is enabled, we collect and process the Functional Data, the Additional Usage Data, and Audio Data.
“Sonos will share a subset of this data with partners that you have specifically authorised to receive such data, in order to ensure that the voice or direct control function is working properly. If you wish to stop us from gathering this information or sharing it with the parties you have authorised, simply disable the feature (for example disable voice control) or unlink the feature (for example unlink your home automation remote) from your Sonos Products. However, the voice control function will not work unless you authorise us to collect and process the data as outlined in this section.
“Sonos is not continuously capturing sound in your home. There are two occasions when we will capture sound from within your home: (1) when you enable voice control technology on the voice enabled Sonos Product; and (2) when you utilise our innovative Trueplay room tuning technology.
“Voice control works by your voice enabled Sonos Product ‘listening’ for a key wake-up word. Please note, not all Sonos Products are voice enabled. In greater depth, this means that the voice enabled Sonos Product buffers and re-records locally, without transmitting or storing any information, until it detects the word or phrase that triggers the device to begin actively recording. If the Product does not detect the wake-up word, it continues to record over itself in a never-ending loop lasting a few seconds. This is all done locally on your Product and is not sent to Sonos or any third party. If a wake-up word is detected, the Product begins recording. In other words, it does not record or retain any audio data, or begin to transmit any data until it is ‘woken up.’ You are notified that it is recording by a visual element such as a light on the Product.
“The Product will record until your voice command is finished. The actual recording of your voice command is then send to the voice partner you have authorised to receive such recording (for example, Amazon). Sonos does not retain a copy of your voice recording.
“Concurrent with the voice recording, your Product will collect data (for example, decibel level at each frequency) about the ambient noise in the room. It is not possible to extract any speech from this type of data. We collect this data to help us improve our speech recognition technology. If you are trying to state a voice command while music is playing for example, Sonos may improve the technology by lowering the music so that your voice command can be understood properly.”
New Sonos Speakers Coming?
Sonos currently only offers one speaker with built-in mics, with that being the Play:5. Unfortunately, those mics are not the far-field array microphones that will be needed to support Alexa. That means it’s likely the company will roll-out a brand-new range of speakers, complete with in-built mics.
Those with legacy devices needn’t worry about missing out on Alexa support, however. That’s because users will still be able to send music to their Sonos speakers through voice commands, they’ll just have to own an Amazon Echo or Echo Dot in order do so.