Google Nest Audio Speaker Review

[olieb]

I'm not sure it is fair to compare Google's R&D investment and profit strategy to other audio companies. Google is an advertising company with massive funds to invest in keeping people within their ecosystem for market research. Anything they don't make, or even lose, on the price of the speaker will easily be made up with the data they gain. If that doesn't bother you, then it is a win as a consumer, but still can't really be compared to how most other companies develop and sell products.

Some of Google's ideas that have flopped would have bankrupted normal companies. They are able to take it in stride and move on to the next project. They paid $20,000,000,000.00 to keep google the default search on Safari browsers. They can hand these speakers out like Halloween candy if it gets people using them to do searches.

I'm pretty sure a fair amount of other companies R&D departments would love to spend the money and time to improve their products and go through the amount of development Google did with this, however their management has to look at risk vs reward and just how much their product will possibly make just on its selling price.

If I could use these over Bluetooth without a Google assistant app and a personal account, I would buy a pair right now. But then Google would not produce them, would they?

 

[Jimbob54]

Available on the UK Google site.

Alas not the dark /white ones. Don't really want green or pink etc.

 

[TSB]

A note on how this works on these kinds of devices:

If the microphone is enabled, all audio is processed by some kind of wake word detection engine. This engine only has one task: to check if you say "OK Google", and this is done locally, i.e. the audio is not recorded or shared with Google.

Only when the wakeword engine detects the wakeword ("OK Google") it wakes up the rest of the audio pipeline, which communicates with the internet. The speaker LED also changes colour to indicate this.

What exactly google does with what you ask the assistant is quite a lot and I can understand why somebody would not like that. It is kind of similar to using Google search from my limited understanding or other android apps like Maps.

But the claim that it is recording everything in your house is not correct.

Regarding the (audio) quality of this device, I am not surprised. I have worked with both very high-end (think 20k euro+) and more mass market products (price level a bit above this speaker). My experience is that the testing, engineering and design is usually way better for the mass market products..

The scale of the number of devices sold allows to sink more development cost on less features, and the quality of the software is therefore often a lot higher.

Simiarly, engineering for speakers is often very good on products of which millions are sold, whereas with high end speakers, it's hit and miss..

 

[JohnBooty]

I'm pretty sure a fair amount of other companies R&D departments would love to spend the money and time to improve their products and go through the amount of development Google did with this, however their management has to look at risk vs reward and just how much their product will possibly make just on its selling price.

Honest, non-rhetorical question.

Is "proper" audio engineering for a device like this really that expensive? I realize it's not pocket change but it feels like it should be in reach for most companies launching products in this space.

I am talking strictly about the audio portion... not the smart/streaming functionality or any of the other myriad costs associated with launching a physical product such as custom fabrication, packaging design, designing an aesthetically pleasing enclosure, safety certifications, etc. Those costs are huge but are constant whether you put in the effort to do proper audio engineering or not so I am ignoring them for the sake of this question.

As far as the audio engineering process goes it's a pretty known process.

1. Use off the shelf OEM drivers
2. Use off the shelf class D amplification and DSP
3. Do the basic sort of 2-way speaker design calculations that hobbyists do all the time with Thiele/Small parameters, etc
4. Measure with Klippel NFS
5. Create the appropriate DSP corrections
6. Iterate on the enclosure perhaps, etc, go back to step 3 and repeat

There are some large expenditures here (buying the Klippel NFS, paying your audio engineers) but my expectation would be that any company doing this sort of engineering is amortizing those costs across many projects. Either because they are tackling many of these products in-house (Apple, Google, etc) or because they are a smaller player outsourcing this work to OEM design partners who have themselves amortized those costs across many projects.

As they say, the quickest way to learn something is to be wrong about it on the internet. So maybe somebody with experience can enlighten me. But my overall expectation/understanding is that thanks to DSP and measurements you certainly don't have to be Google to get some nice engineering into your speaker.

 

[yo!]

Nice to see a 2W speaker is tested this good. The aluminum case is kind of high tech, the LED slot on the front causes a good amount of interference.
But never ever will I accept a google spy in here.

Still using a 1st gen ua Boombox in the workshop, EQing is mandatory. Also a pretty good sounding device and the battery pack is changeable, good old times. (The speaker is more than 10 years in use under roughest conditions and still working fine

 

[capslock]

True. While the hobby aspect -the obsessive tinkering is fading, audiophile-level sound quality is now accessible to a much broader audience. Overall, I see this as a win. If you’re after a specific sound response, you can still achieve it yourself with various DSP options and a standard passive setup.
Typically, DSP is used in less capable speakers to extend bass range and smooth the frequency response, but this effect disappears when you listen at volumes above conversational levels. At that point, you’re essentially left with a midrange + tweeter speaker, which is what it physically is.

Still, these speakers are impressive. The first DSP-driven smart speaker that truly caught my attention was the Libratone Zipp -over 10 years ago.
At this point, can we really still be surprised by this level of performance? Instead, we should be focusing on the manufacturers that persist in releasing subpar designs, often at inflated prices, while others have set the bar much higher. The technology is here, and customers deserve high-quality, thoughtfully designed products without paying a premium for outdated or poorly engineered alternatives.


I remember the fiasco when it launched. They added the switch in response.

Ideally, the switch should physically disconnect the microphone so no hack could reactivate it.
If the switch only serves as an input to the controller, it can indeed be reactivated by a hack -or by Google, which, honestly, I’d consider just as bad.

If you can take it apart, you can always cut the connection or dribble some epoxy into the mic.

 

[ta240]

Is "proper" audio engineering for a device like this really that expensive? I realize it's not pocket change but it feels like it should be in reach for most companies launching products in this space.

To be honest I don't know. But expenses are expenses and with no guarantee of profit from sales most companies are afraid of higher expenses. Google being an advertising company means they don't actually have to make money from sales.

However, when there is no guarantee that spending extra on the engineering process will equate to increased sales I can see them not wanting to spend any more on it. I haven't worked at an audio company but the larger companies that I have worked for watch every penny and there are multiple layers of people to approve any spending. I doubt the smaller players can just go "lets spend X on double blind testing and more on....". Outsourcing expenses would have to be made to seem necessary to upper management and upper management is usually only concerned with expense vs profit.

I'm just guessing that Google's business model for this is completely different from the regular company with "We have this to spend on design and the final production cost has to be X so we can wholesale it for this much. And if we sell this amount of units we will break even and everything from there on is profit". I often hear that Sears should have been what Amazon is, but Sears couldn't have spent so much more than they took in for years and years like Amazon and still been the darling of Wallstreet. They couldn't tell their investors that they would lose that much money just in the name of market share. They would have failed even sooner. Regular businesses have different rules they have to play by.

Polk audio can't say "We are going to spend more on the development of this speaker because it will keep people buying all things Polk and we will not only know what they listen to, but we will also know what questions they ask to research online and we can use and sell that data."

And after all of that there is no guarantee that in the middle of a huge Best Buy or Costco the speaker will actually sound better to the average listener.

 

[ta240]

Honest, non-rhetorical question.

Is "proper" audio engineering for a device like this really that expensive?

How much time has Rick spent on the Directiva speaker? What would that equal if he was a company that someone had outsourced their design testing to? Compared to just doing basic speaker calculations for box size and throwing an amp in it with maybe some basic DSP?

 

[Old_School_Brad]

If you can take it apart, you can always cut the connection or dribble some epoxy into the mic.

*Grabs the bottles, unscrewing the lids.*

Google Assistant chimes in, "Hmm, what are you planning to do with that resin and hardener?"

 

[JohnBooty]

management is usually only concerned with expense vs profit.

It's a crowded, competitive marketplace with lots of people reviewing products. So I think there's a clear sales benefit to making a speaker sound good.

I doubt the smaller players can just go "lets spend X on double blind testing and more on...."

But that's what I'm saying. Even a smaller player can do all of this relatively affordably now; it's even within the reach of hobbyists.

I often hear that Sears should have been what Amazon is, but Sears couldn't have spent so much more than they took in for years and years like Amazon and still been the darling of Wallstreet.

Yeah, that's certainly true.

Theoretically in the 90s Sears could have rolled existing profits (or courted new investments, or something) into a huge "bet the farm" pivot to being the next Amazon without operating in the red, perhaps. But for various reasons that kind of thing is exceedingly rare if not nonexistent.
.
I guess that's what they did with Prodigy, honestly. They certainly put a lot of effort into it. But not "bet the company" levels of effort. It was (to borrow a wack corporate-speak cliche) never really "in their DNA." Also, maybe they were just too early. The technology wasn't ready. They had a dialup service at a time when that was some deeply geeky niche stuff.

 

[oceansize]

Alas not the dark /white ones. Don't really want green or pink etc.

Fair enough. I'll take a pair in any colour - but I'm going to wait to see if the price drops further as we get into November, assuming that they are now end of line. (currently £20 discount if ordering a pair).

 

[JohnBooty]

How much time has Rick spent on the Directiva speaker? What would that equal if he was a company that someone had outsourced their design testing to? Compared to just doing basic speaker calculations for box size and throwing an amp in it with maybe some basic DSP?

Let's back up a step here.

You claimed that only Google-scale companies have the resources to put objectively competent sound into a little enclosure like this.

I questioned that claim.

That absolutely does not mean that I think high-end audio, or the Rick Sykoras of the world are obselete. Far from it. In fact, I explicitly said I (usually) still hate listening to music on these sorts of mini devices.

 

[Barrelhouse Solly]

I recently got a Nest Mini to replace an Echo Dot because I couldn't suppress the vocal response to commands using brief mode. I got tired of hearing Alexa say "OK." It doesn't verbally acknowledge commands and sounds better than the Echo dot. I use it for listening to news radio and controlling a couple of electrical outlets. Music doesn't sound any better than a $30 Bluetooth speaker but I have actual speakers in the room for that. My experience with the Amazon devices is pretty uniform bad sound. I was really surprised at how the $35 device (on sale) performed. The nice thing about the audio is that it's designed to play the part of the audio spectrum it can reproduce and not waste a lot of effort on the highs and lows it's physically incapable of handling.

 

[Cbdb2]

Enough with the soapboxing you guys. Amir showed you the switch that turns off the mic.

The switch the speaker constantly tells you to turn on?

 

[USER]

The switch the speaker constantly tells you to turn on?

It doesn't. These posts are getting dumb.

 

[KMM]

A speaker where someone listens to you.

Turn off the microphone and isolate the speaker from the web via parental controls or similar. I do that with my older iPads I exclusively use as Roon displays

 

[Cbdb2]

It doesn't. These posts are getting dumb.

Amirs review said the speaker told him twice to turn his mic on.

 

[Cbdb2]

Turn off the microphone and isolate the speaker from the web via parental controls or similar. I do that with my older iPads I exclusively use as Roon displays

Dose it even work with out wifi? Does it even work if you dont sign in to Google?

 

[USER]

Amirs review said the speaker told him twice to turn his mic on.

Amir was exaggerating. The speakers show an orange light to indicate the mic is off. (A little annoying but people would complain otherwise.) It'll mention that the mic is off once after a restart but that's it. The home program is designed to recognize google products to help people integrate them into their home group easier, as close to automatically as possible. (Again, people would complain otherwise.) Perhaps you shouldn't chime in about this if you haven't even used the product. Do we really need to make this thread about the instructions? There are better places for that.

 

[Mr. Widget]

The speakers show an orange light to indicate the mic is off. (A little annoying but people would complain otherwise.) It'll mention that the mic is off once after a restart but that's it. The home program is designed to recognize google products to help people integrate them into their home group easier, as close to automatically as possible. (Again, people would complain otherwise.)

I think the fundamental question is about personal privacy and/or trusting the motives behind Google's foray into this space.

It seems obvious from Amir's review that Google made a high investment to bring such a well designed product to market and sell it at what appears to be a loss leader price. For me, the question is why?

In the past a large company would create a loss leader to gain market dominance, but home audio is not a highly profitable area so I can't image they hope to "own" the home audio space. I also doubt they are providing such a well made and obviously well thought out product at a give away price simply because they want the world to enjoy music in their homes. What is the payback for Google?