After the presentation of Apple HomePod, I read several articles of how it can become a SONOS killer. The Apple HomePod has 7 beamforming tweeters while the SONOS play 1 has only 1. But since the high frequencies emitted by a tweeter are more diretional than high frequencies, I was wondering how it is still possible to get 360 sound out of it like Sonos Play 1.
I don't know very much about this but here is my opinion on the subject.
If you think about it, in any reverberant space, any speaker will produce 360 degrees of sound. There is a somewhat new field in audio called immersive audio. There are 3 main techniques used in this field: ambisonics, cross-talk cancellation and wavefield synthesis.
Ambisonics is when we capture sound at a point using many microphones. These microphones can have different numbers of microphones. The simplest one has 4 in a tetrahedral configuration. The higher order mics (Eigen microphone) can have dozens of microphones. The user can use a head tracker and listen to the sound on headphones. As they move the sound changes to match the direction they are facing.
Cross-talk cancellation is capturing sound with a dummy head (a mannequin with mics in its ears) and then cancelling the signal coming from both speakers (stereo) so that you feel like you are in the spot the mannequin was in. It is really complicated.
Wavefield synthesis is when we use an array of microphones and reproduce the sound with an array of speakers at those same locations.
I am sure that there is some sort of interesting DSP that the home pod uses to simulate immersive sound but I think it is mostly marketing. Since people are not attuned to the reality of these systems these big companies can get away with saying they provide immersive audio without actually doing it.
As long as the customer is happy, anything goes.