My Sonos Amp Review.


Userlevel 3
Badge +3
So I have hooked up the new Amp and given it a thorough listening. I am impressed.

My prior setup was B&W 683 Floorstanders into a Audiolabs MDAC Pre/DAC to a Bryston 3B-ST Amp. My streamer of choice was a RPi3B+ running PiCorePlayer and the Squeezelite software attached via USB to the MDAC.

I was a longtime Spotify user but I recently switched to using Apple Music while my wife still uses Spotify. Spotify had a squeezebox plugin so it worked well (though the Spotify Connect implementation was flaky at times). I also have over 1000 CDs ripped to ALAC on a home server.

This was a great sounding setup but I found the airplay unreliable and no native Apple Music support (other than airplay) with Squeezebox so I have been looking at other options. Since native apple music support is limited to Sonos, it was an easy choice, I just had to choose what to get.

I considered just getting the Connect and adding it into the existing system via optical to the DAC but decided to take a chance on the new Amp. Its 125W / channel vs. the Bryston's 120W so nearly identical power though the whole idea of DDFA amps had me sceptical that 125W per side could be handled by the little guy. My speakers also dip down to 3 Ohm at times so they are not the easiest to drive cleanly . The quality of the DAC was also completely unknown. That said if the Amp performance was 95% of my current setup, I could live with it. The return policy swayed me to at least give it a try. If it wasn't as good I could return it and if it does workout, I can sell the Amp and DAC and pay for the Amp.

I ordered Wednesday and it was delivered to me Friday. Unfortunately I missed the delivery guy so I picked it up at the courier depot today. Quick shipping (surprising since Im in Canada but I guess they shipped from the TO area and not from California so no border hassles).

The packaging really reminded me of Apple and I was surprised by how small the Amp actually is.

My speakers were already on banana clips so hooking it all up was quick. Basically plugging in the speakers and power cord.

Next was the app setup. The Sonos setup was easy enough. My only gotcha was that I needed to enable 802.11g on my network. I actually run two dualband routers, R7000 with a run of cat6 to an R3700 running as a Switch and Access Point. B/G clients are rare and there is a potential performance hit enabling support for it on N or AC routers so I don't normally have it on. Anyway on my 3700 2.4Ghz band I enabled G and the setup continued no issue. Added the SMB path to my ALACs and it indexed them rather fast. I also added Apple Music as a source. All very intuitive. I did have to dig a bit for EQ and Amp settings in the app but I found them. Left EQ flat, though I did turn loudness off after a quick google told me what it did (surprised it was defaulted to On on the amp...)

Anyway gave a listen to a few tracks I have locally and know well and it performed admirably. I went back and forth between the MDAC/Bryston and the Sonos Amp to A/B and it was hard to distiguish. The Bryston is an excellent amp. I would give the Bryston a slight nod on clarity, especially on lower volumes and it was actually able to drive the speaker just a bit louder despite being 5W less (to be fair I think my 3BST sample was checked out at ~155W per side on Bryston's exit QC test report more than the rated 120W). On the other hand the Sonos Amp was also able to handle the B&Ws nearly as well. It was also exceedingly quiet and also quite clear. Certainly no complaints and while it wasn't as loud, when turned up it was definitely loud.

My list of complaints about the Sonos is small.

1. It really needs a headphone jack. Its the one thing I will miss from my current DAC. I have a set of Sennheiser 650s that I use when the wife wants quiet, but now I won't have that. I can always use them with the iphone directly but the Senns need a good headphone amp to bring out the best.

2. The Front Panel UI/UX. Actual Up and Down icons for the volume to match the play / pause. Right now its 4 dots making a square. Also I am not sure what each press does but going by the slider on the ipad its not as smooth as a dial.

3. I did notice a clicking/static sound when it hit the end of a playlist. Not sure if it was related to anything else but this could get annoying as the Amp is always on. One suggestion would be to mute itself after no activity for say 5 minutes.

4. Software UI complaints. Nothing major. I am a long time user of the iPeng app that controls squeezeboxen and like its now playing screen and its swipe to left for lyrics and swipe to right to see the queue. I could find no way to see lyrics withing Sonos at all and swiping does nothing. Also need to hit edit to manipulate the queue order vs. just touching a song and moving it up or down.

Otherwise my only concern is longevity. The Bryston is a tank with a 20 year warranty. The Sonos certainly felt premium and people still use the original Connect:Amp after 10 years so. Also DDFAs are a different type of Amp. It got warm but was surprising cool all things considered. The Bryston is a heavy as it is because its a giant heatsink (not that it gets hot either...) but the size of this thing and cleaning up the wires it is crazy its actually smaller than my DAC as seen in the picture below.


This topic has been closed for further comments. You can use the search bar to find a similar topic, or create a new one by clicking Create Topic at the top of the page.

35 replies

Good review. The amp performance is what I would expect from my experience of two Connect Amps that replaced amps similar to yours. My speakers were easy loads with impedances never dropping below 5 ohms, so that allowed the Connect Amps to be more than adequate and I haven't missed my heavy amps at all.
The small footprint and elimination of cable clutter is a huge blessing; one of my amps was a two box set up with a preamplifier.
My Connect Amps are from 2011 and work fine, but are not built to last for decades with occasional service support as the earlier amps were. For the price difference, that is quite acceptable I think.
Userlevel 7
Badge +20
Very good review, thank you. Consistent with my experience, though all I did was swap a Connect:Amp for the new Amp, driving a pair of Q Acoustics 3050 floor-standers. (A completely unnecessary amplifier upgrade, but I couldn't resist it.)

1. It really needs a headphone jack. It's the one thing I will miss from my current DAC. I have a set of Sennheiser 650s that I use when the wife wants quiet, but now I won't have that. I can always use them with the iphone directly but the Senns need a good headphone amp to bring out the best.
I fully agree. In a different location I have a Connect hooked up to a 5.1 amplifier for home theatre duties, and I quite frequently use its headphone socket to listen to Sonos. It's a shame I can't do the same with the Amp.
Userlevel 1
Badge +3
Just bought an amp to replace ZP90-DAC-Yamaha amplifier, primarily so I can add on the Sonos Sub. Hooked it up and it sounds OK. Harsh at the top end even when EQ’d and sounds ‘boxy’. Lots more bass and detail though. So I added the ZP90 and DAC to the line in and bingo. Top end much better - harshness gone and sounds better that the Yamaha in pure direct. Phew! Thought I’d wasted 600 quid for a minute.
Userlevel 7
Badge +20
Just bought an amp to replace ZP90-DAC-Yamaha amplifier, primarily so I can add on the Sonos Sub. Hooked it up and it sounds OK. Harsh at the top end even when EQ’d and sounds ‘boxy’. Lots more bass and detail though. So I added the ZP90 and DAC to the line in and bingo. Top end much better - harshness gone and sounds better that the Yamaha in pure direct. Phew! Thought I’d wasted 600 quid for a minute.
I assume you realise that you're going digital -> analog in the ZP90/DAC combo, then analog -> digital -> analog again in the Amp?

Personally, I've not detected any harshness or boxiness in my use of the Amp, but it may be speaker dependent, and may also depend on whether there's a sub present or not.
Userlevel 4
Badge +3
Just bought an amp to replace ZP90-DAC-Yamaha amplifier, primarily so I can add on the Sonos Sub. Hooked it up and it sounds OK. Harsh at the top end even when EQ’d and sounds ‘boxy’. Lots more bass and detail though. So I added the ZP90 and DAC to the line in and bingo. Top end much better - harshness gone and sounds better that the Yamaha in pure direct. Phew! Thought I’d wasted 600 quid for a minute.

I would let the Sono Amp breakin with music for several hours before judging the sound quality. No harshness from mine after doing so. Your setup seems strange to me using the ZP90 along with the Sonos Amp. What purpose does the DAC serve in this setup?
Userlevel 1
Badge +3
Not sure the purpose of the DAC or anything else. Just know it sounded better. On the ‘breaking in’ front, that’s a myth.
Userlevel 4
Badge +3
Not sure the purpose of the DAC or anything else. Just know it sounded better. On the ‘breaking in’ front, that’s a myth.

I have read enough reviews of the Sonos Amp mentioning brightness in the treble to believe this is not a myth. My experience with the amp confirms that as well. Try eliminating what seems to be unnecessary equipment from your setup after your Sonos Amp has broken in and report back.
Userlevel 1
Badge +3
You mean using an electronic component for a while changes the sound? Not true. Im hoping Sonos hear comments relating to the top and correct it with a firmware update soonest.
Userlevel 3
Badge +5
I have been very close to buying the new AMP, but after reading several tests of this amplifier where everyone is mention the bright harsh treble, I decided to listen to the AMP in a HiFi shop and on several different speakers, and yes this amplifier has a problem with the treble.
Personally, I am very surpised that Sonos can bring such an amplifier to the market, I can only believe that this it is a mistake or that Sonos has deliberately attempted to voice this amplifier for an very analytical studio monitor sound.
As for my part, I await a suggestion from Sonos on how they plan to make this amplifier playing with a linear pleasing sound.
Userlevel 7
Badge +20
I have been very close to buying the new AMP, but after reading several tests of this amplifier where everyone is mention the bright harsh treble, I decided to listen to the AMP in a HiFi shop and on several different speakers, and yes this amplifier has a problem with the treble.
If you auditioned expecting to hear something, it's not really a big surprise that you did. This is known as expectation bias. I'm assuming that your listening didn't include any blind tests with other amplifiers?
Userlevel 1
Badge +3
I don’t think any statistical tests are required. It just needs the sound to be rebalanced in a more detailed way than a single slider on the app can achieve. It’s a shame that it’s not possible with my speakers in my room. Making the Amp more neutral and faithful to the source would be great!
Userlevel 3
Badge +5
I have been very close to buying the new AMP, but after reading several tests of this amplifier where everyone is mention the bright harsh treble, I decided to listen to the AMP in a HiFi shop and on several different speakers, and yes this amplifier has a problem with the treble.
If you auditioned expecting to hear something, it's not really a big surprise that you did. This is known as expectation bias. I'm assuming that your listening didn't include any blind tests with other amplifiers?


Now we start all over again, I have been working in the music industry for many years with pro audio sound, so I will not let myself drag around with your postulates.
Although I do acknowledge your very correct point of view, however, in this case, the Sonos AMP has a problem.
Userlevel 7
Badge +20
Now we start all over again, I have been working in the music industry for many years with pro audio sound, so I will not let myself drag around with your postulates. Although I do acknowledge your very correct point of view, however, in this case, the Sonos AMP has a problem.
I acknowledge your opinion on this. I'll only offer that my Amp has no such issue, and I've spent many hours listening to it now, across a wide range of sources and content. However, I guess it's possible that there are defective (or incorrectly wired) units out there.

It's been true for decades now that all reasonably designed amplifiers sound indistinguishable (i.e., neutral) when operating within their power limits and in the absence of EQ adjustments. My personal experience suggests that the Sonos Amp is no exception.
Userlevel 3
Badge +5
I don't think it's just a single amplifiers with an error, I'm pretty sure it's the programming of the DSP where Sonos mistekenly has aimed for a particular sound image.
In the old days when you tested amplifiers, you measured the frequency response of the amplifier, now no one does that,but only use their ears.
However, I think I have seen a frequency response test on AMP and that showed that the problem was most pronounced in the left channel, so perhaps the problem is not in the programming of DSP.
Userlevel 4
Badge +3
I’m using $4K + Tannoy DMT 15 Studio monitors with my Sonos Amp and it sounds great tonally. I have all tone controls set flat. It sounds tonally the
same as the amp it replaced that I had been using with those speakers for the last 8 years.

Do break it in with music for several hours as the amp will open up and improve!
Userlevel 3
Badge +5
“Do break it in with music for several hours as the amp will open up and improve! “

This is nonsense, placebo effect !
Userlevel 4
Badge +3
YMMV, based on expectation bias.
Userlevel 3
Badge +5
This is perhaps corretct with completly new unused speakers, but not for an amplifier.
Userlevel 1
Badge +3
It’s not correct for any electronic component. They don’t change with use, speaker or otherwise. It’s wishful thinking.
Strange they don’t change for the worse, isn’t it?
My own observation is that amplifier differences are obvious and important to some listeners while other listeners don't care. Note that I'm not dumping on any listeners. If amplifiers don't matter much to you, don't spend your money there -- and don't worry that someone else makes different discriminations. I will also note that any A/B comparisons must be carefully done because eliminating minor differences in level is critical. Humans tend to prefer the louder unit.

I prefer a slightly different comparison technique of 'A' [quiet period] 'B' for purchase decisions. While this might not be suitable for publication in a journal, my opinion is that if the difference is so subtle that one needs a properly managed traditional A/B/x ('x' means that no one in the room knows which unit is playing) in order to pick up any difference, there is no need to spend more money. In a research environment I would want to use a properly managed A/B/x comparison with a large number of listeners who cannot communicate with other research subjects. If a respected member of a group smiles or frowns during the session, other listeners might be influenced.

And, I have some difficulty with the A/B/x technique, even though this is the gold standard method. In A/B/x we have a box that switches components in a random sequence and listeners note 'better' or 'worse' after each switch. No one in the room knows the sequence. At the end of the test series a readout from the switcher is compared with the listener's record. If the result of 'A' vs 'B' is 50/50, there is not likely any difference. Even 60/40 is not very significant. The part that troubles me is that the A/B/x box is assumed to be "neutral" by not adding or removing anything. My question is: If we can build an absolutely neutral A/B/x box, why can't we build "neutral" equipment? At one point I built an A/B/x box and it was immediately obvious to me and everyone else familiar with the equipment that differences were more difficult to discern when the box was used.
With respect to "break-in", this is mostly nonsense, however, there can be changes in characteristics over time. For example, an amplifier's operating temperature may need to stabilize for a few hours. To me this is a design fault and this stabilization will need to be repeated after each power cycle. I can recall an electrostatic speaker that needed about 72 hours to stabilize because the membranes charged slowly. One should purchase those speakers on a Friday, apply power and go away for the weekend. Again, this was not "break-in" because if power was removed for an extended period the membranes would discharge and you'd need to repeat the 72 hours.
My own observation is that amplifier differences are obvious and important to some listeners while other listeners don't care. Note that I'm not dumping on any listeners. If amplifiers don't matter much to you, don't spend your money there -- and don't worry that someone else makes different discriminations. I will also note that any A/B comparisons must be carefully done because eliminating minor differences in level is critical. Humans tend to prefer the louder unit.

I prefer a slightly different comparison technique of 'A' [quiet period] 'B' for purchase decisions. While this might not be suitable for publication in a journal, my opinion is that if the difference is so subtle that one needs a properly managed traditional A/B/x ('x' means that no one in the room knows which unit is playing) in order to pick up any difference, there is no need to spend more money. In a research environment I would want to use a properly managed A/B/x comparison with a large number of listeners who cannot communicate with other research subjects. If a respected member of a group smiles or frowns during the session, other listeners might be influenced.

And, I have some difficulty with the A/B/x technique, even though this is the gold standard method. In A/B/x we have a box that switches components in a random sequence and listeners note 'better' or 'worse' after each switch. No one in the room knows the sequence. At the end of the test series a readout from the switcher is compared with the listener's record. If the result of 'A' vs 'B' is 50/50, there is not likely any difference. Even 60/40 is not very significant. The part that troubles me is that the A/B/x box is assumed to be "neutral" by not adding or removing anything. My question is: If we can build an absolutely neutral A/B/x box, why can't we build "neutral" equipment? At one point I built an A/B/x box and it was immediately obvious to me and everyone else familiar with the equipment that differences were more difficult to discern when the box was used.


Good points about the ABX test. Using it as part of a testing methodology is quite challenging. There are very very few peer reviewed studies and even then, not enough for a scientific confirmation of what they test. While the ABX test is not in doubt, there are too many unchecked confounding variables (and very small sample sizes of test subjects) to reach a confirmation of conclusions.

Regarding the Sonos Amp, I also found it bright when I tested it a few weeks ago (Dynaudio M10 speakers, lossless from Tidal). It was also quite boomy in the bass despite proper speaker positioning and distance from wall (well treated room). Personal preferences I am sure. I ended up dialing down the treble by 3 clicks and the bass by 2 clicks. Sounded fine then.
Fantastic convenience in a super discreet package. I did not purchase it in the end as I feel that a similarly priced AV receiver will sound just as good but also offer better room correction options.
Userlevel 3
Badge +2
An interesting review. Thanks for posting.

It is disappointing that Sonos has chosen to not address the multiple reports of brightness. In the meantime I've put my planned Amp purchase on hold.

Sonos presumably has the lab equipment available to conduct such a test, it would be great if they published some frequency response test results the community could review.
Amplifier "brightness" tends to be a subjective opinion and does not show up in a simple frequency response plot. Most amplifiers that are tagged as "bright" will have an absolutely flat frequency response. Not even the traditional THD (Total Harmonic Distortion) and entry level IMD (Intermodulation Distortion) tests can discover units that will be tagged as "bright". There are some very sophisticated tests, requiring expensive research grade instrumentation, that might be able to discover the "bright" units, but these tests are currently quoted only in academic discussions. There is no widely accepted standard method to perform these tests or discuss the results.

ings you should listen to AMP and decide for yourself. Only you know what sounds "best" (to you -- ignore the rest of us).

The flip side of "bright" is "dull". In my experience there are more "bright" units than "dull" units. If one is accustomed to the "bright" units, a not "bright" unit is often described as "dull". Phono cartridges typically exhibit a peak in their output in the 12-16KHz range (this range and magnitude varies greatly and is temperature sensitive) Anyone who can plot the frequency response of a cartridge has a lot to say about these peaks. There was one cartridge that was amazingly flat and did not exhibit this peak. After looking at the plot of this cartridge, there was not much to say -- except that it was flat. Many listeners characterized this cartridge as "dull" because they had been conditioned to accept the typical bright peak as being correct.
In my almost two decades of using decent audio kit, that includes amps from makes like
Marantz/Rotel/Arcam/ConradJohnson/Quad/Yamaha/NAD/Unison Research/Plinius and now Sonos - a good mix from different price points and geographies as diverse as Japan, Italy, US, UK, New Zealand - I still have to come across a bright sounding amp. I therefore doubt that such an animal exists outside of cheap Chinese knock offs, which are also quite decent these days within the limits of their capability. On the other hand, I have heard sound differences every time I changed speakers or moved a pair to another room, with differences ranging from barely audible to dramatically different, worthy of a range of adjectives like bright, dull, boomy, natural etc etc. Ditto from cartridge changes when I dabbled with vinyl.

It therefore remains a puzzle to me how people continue to unearth these differences where amps are concerned - and the media continues to do so from last year's model of the same brand to that released this year, but that isn't a puzzle at all once their behaviour drivers are understood. Perhaps I am a rare lucky person that never came across these mythical - to me - beasts. Or perhaps I always took the necessary care to not wire underpowered amps to difficult speakers. Or could it be that the people who find these could be wrong or suffering from the usual known biases?

I haven't even heard differences across DACs, that arguably present more complex design challenges than amps.

As to the Sonos Amp, I have not heard it in use, but if my Connect Amp experience can be used as a guide, there is no reason for me to believe that Sonos has goofed and given it a sound signature, something that would be a rookie blunder.

There are also enough user posts here that lend credence to the belief that using it as a drop in replacement would not be noticed except for higher sound levels if those are desired. And given the piss poor record of media reviews, any sound signature claims there can be very safely ignored.

On the other hand, there are many that continue to believe that the CIA killed Kennedy and the US never landed anyone on the moon.

If any sound signature is heard on the Sonos Amp, it should be returned to Sonos for a refund, claiming supply of a defective product. IMO.