Home > Business > Google’s Nest Audio is a legit good speaker, and I have the data to prove it

Google’s Nest Audio is a legit good speaker, and I have the data to prove it


No self-respecting audio enthusiast takes ‘smart’ speakers seriously. The Amazon Echoes and Google Homes of the world are, after all, primarily designed to get you hooked on their respective voice assistants. That they happen to be able to play some tunes is little more than the means to that goal.

At least, that’s the cynical predisposition. I hope you can forgive me for being a little jaded about Google’s new Nest Audio — what I thought was simply another smart speaker destined to collect dust on a dresser. Audio is something I care deeply about, and I regularly review multi-thousand dollar speakers here at TNW. Not to say good speakers need be expensive, but the big G isn’t exactly one of the first companies I think about when it comes to sound quality.

After spending some time with the Nest Audio, I think that’s about to change — and I’ve got the data to show you why.

See, something caught my attention during Google’s Pixel 5 event and a separate briefing with journalists thereafter. Google didn’t just bore us with the usual marketing spiel about how the Nest Audio would let you hear music “as the artist intended” — sighs — but it also talked about some of the science behind the Nest Audio’s improvements.

While the original Google Home was primarily tuned through intuition — a surefire way to make mediocre speakers — the Nest Audio was tuned through research, according to Product Manager Chris Chan. The briefing was simplified for mainstream audiences, sure, but mentions of double-blind testing, maximizing dynamics, thermal dissipation, dispersion characteristics, waveguides, and internal cabinet volume reminded me more of a hi-fi briefing than something I expected for a $99 smart speaker. Chan even says Google developed its own speaker software that allows it to simulate the speaker’s sound at 2,500 listening positions.

So when I received my Nest Audio for review, I decided I would treat it like any other of my other hi-fi reviews: I was going to science it.

The Basics

Fair warning: this review will get technical, so if you’re not interested in the data behind a good speaker, there will surely be a multitude of other reviews you can choose from. I’m not going to spend much time on features and the like beyond what I wrote on announcement day.

I do have some quick impressions to share though. The Nest Audio is a cute little speaker meant to replace the Google Home before it. It obviously supports the Google Assistant and Chromecast, and it can also play audio via Bluetooth as well.

It’s smaller than I expected from photos, about the size of a paperback book. The fabric wrap looks classy, and though I wish Google had sent me one of the more colorful finishes, the Nest Audio appears to be decidedly inoffensive to most decor.

Using it is a cinch. Setup takes only a minute or two via the Google Home app. It responds rapidly to most basic voice commands thanks to local processing with a special machine learning chip. Volume and playback can be controlled via three hidden touch areas at the top of the speaker. You can connect it with another Nest Audio to create a stereo pair (it works for both Bluetooth and Chromecast), although I only received one unit for review.

The Nest Audio can automatically EQ itself and raise volume depending on ambient noise and the content that’s playing. Google says it leaves music untouched, which I appreciate, but it was pretty cool to see how the speaker could raise its volume when I turned on a faucet in the kitchen.

And oh yeah, the Nest Audio sounds really good. It’s not boomy or tizzy. It is, to my very pleasant surprise, resoundingly neutral, although you do have access to bass and treble sliders in the Google Home app. It won’t get crazy loud, but within its decibel limitations, it sounds seriously good for $99 a pop.

The measurement will show us just how good.

What I’m looking for:

As far as I’m concerned, there are two primary requirements for good sound:

  • A relatively flat anechoic frequency response. In speakers, this is perceived as neutral, which studies show usually equates to ‘good.’
  • Smooth dispersion characteristics. This means that as you move off-axis (away from the front and center of the speakers), the sound changes smoothly. This translates to a good soundstage, as these sounds eventually get reflected back to your listening position, which your brain interprets as a sense of space.

These characteristics have repeatedly been shown to correlate highly with listener preferences in double-blind tests — usually more than other aspects of sound reproduction like distortion.

How I measure

In order to perform useful speaker measurements, we need to minimize the influence of room reflections that will mask the true frequency response. Audio companies do this by measuring speakers in an anechoic chamber. I can’t fit one of those in my Brooklyn apartment, but I did just move to a place with a backyard, which helps; the farther I am from walls, the closer I can replicate an anechoic result.

I can then take measurements in software like Room EQ Wizard and, using a technique called impulse response gating or time-windowing, essentially ‘see’ when the first major reflection arrives at the microphone, usually a few milliseconds after the initial signal. I then set my ‘window’ to only include the data up to right before this reflection hits, and the reflections are effectively removed from the final frequency response.

Oh yeah, I also capture the data at 70 different angles to accurately characterize a speaker’s dispersion characteristics (as established by the CTA-2034A measurement standard). Yes, it is tedious, and I may or may not have dropped the speaker a couple of times during the process (don’t worry, the measurements didn’t change).

Google’s Nest Audio is a legit good speaker, and I have the data to prove it 2