“At Bell Labs it became apparent that music might be a significant driver for Internet use and I found myself walking several paths. One was sound field reconstruction,” Steve Crandall writes for Omenti. “Sound field reconstruction is just the fancy way of recreating the sound where it was recorded or engineered in your home. In theory you record the acoustic signatures of the recording studio and your living room and create a mathematical operation that makes your living room sound like the recording space. In practice it’s approximately hard. Sound waves are fairly large and get reflected, absorbed and transmitted by and through everything in your room. Making matters worse the sound field at different spots in the room is usually very different and things in the room move between and during listening sessions. We made serious progress, but it took exotic microphone and speaker arrays as well as state of the art (for the time) computing to make it sort of work.”

Place HomePod anywhere in the room. It automatically analyzes the acoustics, adjusts the sound based on the speaker’s location, and separates the music into direct and ambient sound. Direct sound is beamed to the middle of the room, while ambient sound is diffused into left and right channels and bounced off the wall. So your music sounds amazing, wherever you are in the room.

Place HomePod anywhere in the room. It automatically analyzes the acoustics, adjusts the sound based on the speaker’s location, and separates the music into direct and ambient sound. Direct sound is beamed to the middle of the room, while ambient sound is diffused into left and right channels and bounced off the wall. So your music sounds amazing, wherever you are in the room.

 
“The HomePod uses seven small speakers arranged in a ring,” Crandall writes. “The HomePod also has a ring shaped array of six microphones. They listen to the room (sample it) at a fairly high rate and decide what parts of the music, from its stereo signature, need to go where and how it needs be modified to interact with the room and everything in it. An Apple A8 processor, the same two billion transistor processor used in the iPhone 6 and 6 Plus, handles the computational chores. This is much more power than we had back in the day. Although I don’t know exactly what they’re working on, several people I know who are real experts at this have been at Apple for years.”

“The HomePod didn’t match the reference speakers when I was sitting at the sweet spot, but was better when I moved around the room. That is astounding,” Crandall writes. “It was creating stereo – a term has nothing to do with a pair of anything, but rather three dimensional. Moving beams of audio around it was painting a sound field in my living room.”

MacDailyNews Take: Apple has redefined stereo.

Read more in the full article – very highly recommended – (don’t miss the bits about other “smart speakers,” privacy and more) here.

MacDailyNews Take: About “stereo,” it’s time to Think Different.

Listening to something like this on just a single HomePod truly is astounding!

We cannot wait for “stereo pairing” to come to our HomePods for even more immersive sound experiences.

As Reddit user Arve stated after HomePod’s unveiling:

1. They’re using some form of dynamic modeling, and likely also current sensing that allows them to have a p-p excursion of 20 mm in a 4″ driver. This is completely unheard of in the home market. You can read an introduction to the topic here. The practical upshot is that that 4″ driver can go louder than larger drivers, and with significantly less distortion. It’s also stuff you typically find in speakers with five-figure price tags (The Beolab 90 does this, and I also suspect that the Kii Three does). It’s a quantum leap over what a typical passive speaker does, and you don’t really even find it in higher-end powered speakers

2. The speaker uses six integrated beamforming microphones to probe the room dimensions, and alter its output so it sounds its best wherever it is placed in the room. It’ll know how large the room is, and where in the room it is placed.

3. The room correction applied after probing its own position isn’t simplistic DSP of frequency response, as the speaker has seven drivers that are used to create a beamforming speaker array,. so they can direct specific sound in specific directions. The only other speakers that do this is the Beolab 90, and Lexicon SL-1. The Beolab 90 is $85,000/pair, and no price tag is set for the Lexicon, but the expectation in the industry is “astronomical”.

So yes, compared to the typical sub-$2000 speaker, the technology they apply may just as well be considered “magic”.

SEE ALSO:
Apple HomePod: The audiophile perspective plus 8 1/2 hours of measurements; HomePod is 100% an audiophile-grade speaker – February 12, 2018
Apple’s HomePod is actually a steal at $349 – January 26, 2018
Digital Trends previews Apple’s HomePod: Impressive sound coupled with strong privacy – January 26, 2018
Hands on with Apple’s HomePod: Attractive, ultra-high-quality speaker, an excellent Siri ambassador – January 26, 2018
Apple’s HomePod, the iPod for your home – January 25, 2018
One hour with Apple’s new HomePod smart speaker – January 25, 2018
Apple’s iOS 11.3 beta delivers AirPlay 2 with multi-room playback – January 25, 2018
How Apple is positioning the HomePod and why – January 24, 2018
How I got talked into buying an Apple HomePod despite my reservations – January 24, 2018
Tim Cook says audio quality puts HomePod ahead of ‘squeaky-sounding’ competition – January 24, 2018
Apple’s HomePod arrives February 9th, available to order this Friday, January 26th – January 23, 2018
Apple delays HomePod release to early 2018 – November 17, 2017
Apple reveals HomePod smart home music speaker – June 5, 2017

[Thanks to MacDailyNews Reader “Dan K.” for the heads up.]