Google Nest Audio Speaker Review

[Inertiaman]

That is exactly what I did. (Well I used 1800 Hz LR4.)
I bought several units of these as "defective", figuring that the defect is in the electronics and the drivers will be ok.
First I tried to reproduce the peaks and dips in midsummer (resonance?) but could not do it. Some hiccups (4 kHz) and hiccdowns (10.5 kHz) but not what Amir measured.
My measurements
orange: nearfield (1 cm) from woofer
with Umik in the living room (gated 6 ms) 25 cm distance
black: with grill
red: without grill (the grill/cloth seems to reduce treble quite a bit)
green: with front grill pulled out about 3 mm (there seem to be reflections)
View attachment 410186


blue: tweeter without grill, with Umik in the living room (gated 6 ms), 25 cm distance
black: woofer nearfield 1 cm, ungated, level adjusted
red: woofer 25 cm, ungated
orange: woofer 50 cm, ungated
View attachment 410187

Then I took off all the electronics (everything is nicely done with screws) and measured the drivers directly.
View attachment 410195
All the electronics inside. I only reused the white pin connector in the center to connect to the drivers.

I have an old Minidsp 2x4 that I use for EQ.
For aesthetic reasons I wanted to use the plastic panels but not accept the disadvantages.
I peeled off the cloth and cut out the grill for the tweeter.
Now it looks like this:
View attachment 410193
I might sand and paint it later. Still in progress.

This is the FR that I arrive at (gated again and without grill). In bass I adjusted for ≈60 Hz, - 3dB. But this works only for moderate levels of course.
View attachment 410189

Just for fun I made a couple of near-field measurements with the google DSP active and also with driving the speakers directly from an external amp, bypassing the DSP and internal crossover. It looks like not much is going on DSP-wise in the speakers.
It has Linkwitz transformation for the woofer below ~190 Hz. It makes the dynamic compression at low frequencies necessary, because it is pushing up the small woofers distortion considerably.
The crossover seems like a standard 12dB/octave around 2kHz for the woofer, and I guess it is the same for the tweeter.
I don't see much FR shaping otherwise.
BTW in the speaker the crossover, the Linkwitz transform and the 2 power amps are in a TAS5825M class D amp/DSP chip on the small board where the power connector is. It has i2s input, but I am not sure how much the google SW messing with the configuration of the chip after initialization.
Maybe it is easier just to connect it to two external amp and do your own digital crossover and not so aggressive Linkwitz transform.

@olieb and @fcserei , thanks for the measurements and thoughts on de-Googling these speakers.

I'm considering a similar but fully passive approach. My use case is the rear area of a camper van with 1.5m-ish listening distance. Due to various electrical and component constraints, its not practical to do an active crossover. However, I do have a fairly good EQ capability and high-pass filters from my head unit + subwoofer (and associated low-pass filter and level control).

If I high-pass these at ~ 100Hz and utilize the sub, then I won't need to boost bass. Which leaves a passive crossover in the 1500-1800 Hz range. Some questions:
1) Looking at your measurements, it seems like the woofer isn't even low-passed in the original state (response extends out near 8k). Or am I missing something (were you driving test signals directly to the isolated drivers and not a common signal via a cast "thru" the Google digital transfer function)?
2) do you think a 12db/second-order crossover will suffice?
3) I expect I'll need to make measurements of the drivers to get data necessary for a semi-optimal crossover design. Is there other work I'm overlooking?

I realize there are a million options for passive speakers in the back of my van that would take far less effort, but the size and shape of these Nest speakers is quite ideal for my situation, and seems like potential for good sound leveraging the cast waveguide-ish enclosure and decent drivers, even if a less than perfect passive crossover "lessens" their performance.

 

[olieb]

1) Looking at your measurements, it seems like the woofer isn't even low-passed in the original state (response extends out near 8k). Or am I missing something (were you driving test signals directly to the isolated drivers and not a common signal via a cast "thru" the Google digital transfer function)?

I made both kinds of measurements. First WITH the Google DSP (first picture):
A. as a gated measurement from 25cm with and without grill.
B. as a ungated measurement directly (1cm) at the woofer cone. There one can see the 12dB/oct 2kHz crossover and the tweeter is absent because the woofer is so loud and the measurement is off axis of the tweeter.

And then I measured the drivers without any electronics (second picture), just in the cabinet. The woofer (red) goes up higher than 2k of course and has some breakup peaks.

The black curve is a nearfield measurement (ungated) of the woofer that is not valid above 500 Hz but is better for EQ the bass.
[orange is a measurement with greater distance to check for artefacts from near measurement, but there are none, just the room influence if you go further away]

There are two versions of the Google Nest out there. This one (the same that Amir tested) and a (presumably newer one) with a somewhat different woofer (surround is slimmer). They do not have the same FRs.

I used 24dB/oct crossover but passive I would not do that, 2nd order should be ok. The biggest problem might be the rather peaky resonance of the tweeter, or not. I am not very experienced with passive crossover design.
I do not know why you can't go active, I ended up using this module to drive a stereo pair.

TINYSINE TSA7804C Module Amplificateur SPDIF TPA3116D2 DAC DSP 4 x50W

Le module amplificateur SPDIF Tynisine TSA7804C est un module d'amplification DSP à 4 canaux de 50W chacun.

www.audiophonics.fr

 

[silkpurse]

@olieb and @fcserei , thanks for the measurements and thoughts on de-Googling these speakers.

I'm considering a similar but fully passive approach. My use case is the rear area of a camper van with 1.5m-ish listening distance. Due to various electrical and component constraints, its not practical to do an active crossover. However, I do have a fairly good EQ capability and high-pass filters from my head unit + subwoofer (and associated low-pass filter and level control).

If I high-pass these at ~ 100Hz and utilize the sub, then I won't need to boost bass. Which leaves a passive crossover in the 1500-1800 Hz range. Some questions:
1) Looking at your measurements, it seems like the woofer isn't even low-passed in the original state (response extends out near 8k). Or am I missing something (were you driving test signals directly to the isolated drivers and not a common signal via a cast "thru" the Google digital transfer function)?
2) do you think a 12db/second-order crossover will suffice?
3) I expect I'll need to make measurements of the drivers to get data necessary for a semi-optimal crossover design. Is there other work I'm overlooking?

I realize there are a million options for passive speakers in the back of my van that would take far less effort, but the size and shape of these Nest speakers is quite ideal for my situation, and seems like potential for good sound leveraging the cast waveguide-ish enclosure and decent drivers, even if a less than perfect passive crossover "lessens" their performance.

I feel like gigaworks t40 ii would probably be better by a lot. The DSP is the trick here. SPL is way too limited without it.

 

[capslock]

This thing is finicky as hell. If your wifi connection is weak forget about it. Also if you are an iOS user, smart features are ehh useless since you have Siri. Finally its largest weakness is that it does not support Amazon Music, which limits its use to bluetooth only. I got mine for $50 and it's ok for a BT speaker at that price and sound quality, I guess.

Hmm. I got one of these and installed it in the kitchen today. Playing Amazon Music on an iphone through Chromecast right now. Have been using Amazon Music with a Chromecast Audio for about half a year now without any issues. Is this an issue that has gone away?

Side notes:
- On first impression, a bit bright but otherwise like a monitor. Bass is there but nothing like the JBL bluetooth boomstick my kids we using which I suspect has a massive boost in the upper bass.
- Switched to Amazon music from Deezer a while ago to take advantage of their trial offer and stayed there because the app was much improved compared to a few years ago when I last tried and completeness of catalogue is unmatched.

 

[capslock]

@olieb and @fcserei, thank you for posting your measurements. How did you manage to play your test signal through the google DSP, especially for the gated measurements? I imagine this would require rerouting the test signal to chromecast, and I am not aware how it can be done with Arta or REW. Thanks

 

[olieb]

I imagine this would require rerouting the test signal to chromecast, and I am not aware how it can be done with Arta or REW.

I did it with REW and Chromecast on a Macbook, but I do not recall exactly how. I remember to have played with the timing, maybe that did it. Or I routed the sound through Audio Hijack (and to Chrome or some third party tool)? I tried a few things and somehow I succeeded. I deleted everything Google immediately afterwards again.
In REW you can play the sweep from a file, that is another option (that I did not use). The gating is applied afterwards.

 

[ninetylol]

One of my speakers loses pairing every few hours suddenly. Tried resetting and setting up the whole thing several times but it wont work. Anyone found a solution to that?

 

[milus]

One of my speakers loses pairing every few hours suddenly. Tried resetting and setting up the whole thing several times but it wont work. Anyone found a solution to that?

Try finding an answer here:

Having trouble with Speaker groups?

or in its parent folder:

Google Nest Community Speakers and Displays

There has been trouble with some google products' chromecast firmware lately

DO NOT reset your speakers! If you have already done so, there is some advice there

Hope it works for you!

 

[capslock]

One of my speakers loses pairing every few hours suddenly. Tried resetting and setting up the whole thing several times but it wont work. Anyone found a solution to that?

Pairing as in bluetooth? Then it would be unrelated to the chromecast issues

 

[ninetylol]

@milus @capslock I mean the Stereo pairing. My master speakers always works but the second speaker always loses connection to the network it seems. Tried resetting Router, Speakers, Homegroups etc -> still its always the slave speaker losing connection (even if i change left and right speaker / master slave)

 

[capslock]

That should be covered in the second link provided by milus. Apparently, a firmware update should fix it.

 

[ninetylol]

That should be covered in the second link provided by milus. Apparently, a firmware update should fix it.

I will try i looked for a way to update Firmware manually but didnt see anything last time.

 

[xray]

I will try i looked for a way to update Firmware manually but didnt see anything last time.

Unless they have stopped updating that thread it looks like the firmware fix is not out yet for all devices, just some.