Answered

SONOS ALEXA Integration?



Show first post
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.

267 replies

An update from the Sonos folks would be appreciated. I love my Sonos equipment, and continue to invest (I have 2 play 5s, 3 play 1s, soundbar, subwoofer, a connect, and a connect amp). As a paying customer whose been told this feature is coming, at least SOME sort of an update would be appropriate.

From 2 days ago:

https://en.community.sonos.com/music-services-and-sources-228994/alexa-echo-yet-6769174/index4.html#post16131181


The Alexa integration already has the highest priority here and the team's working to bring it out with all necessary fires burning under the right seats. While we're not sharing a timeline beyond the promise that it'll be out this year, that doesn't mean it's far out, just that it isn't here yet. We tend not to give timelines but wanted everyone to know this was the direction we're heading, so in this case, the company broke the usual silence and let slip some plans.

As soon as we have any further details to share we'll be letting everyone know.


If you are looking for things like exact dates or specific timelines, you aren't going to get them. Sonos does not schedule by calendar date, they schedule according to testing goals met.
Userlevel 1
Badge +1
An update from the Sonos folks would be appreciated. I love my Sonos equipment, and continue to invest (I have 2 play 5s, 3 play 1s, soundbar, subwoofer, a connect, and a connect amp). As a paying customer whose been told this feature is coming, at least SOME sort of an update would be appropriate.
Userlevel 4
Badge +2

"Indignant posters" aka frustrated Sonos customers. Frustrated customers voicing their concerns are not the problem, that's called "feedback". It's when we stop caring that it becomes an issue.


If 'we' means Sonos, then I agree to an extent. How much is up to them. "We ', meaning the other customers in this forum...I don't see why we need to care.

I personally would agree that voicing frustration and concerns is positive. But it's not productive when customers threaten to sell their stuff if demands are not met, repeating inaccurate rumors, generally assuming that this is all easy, and Sonos doesn't care.[/quote]


The "we" I'm referring to are the Sonos users. We care because we like the product and we want them to keep improving it. I, for example, had three Logitech Squeezebox Radios that frustrated me to no end. Software updates seemed to only make things worse. One of them bricked during an update with no way to fix it or replace it. That's where they lost me. I don't care what they do with that product anymore.

Sonos, i care about. Sonos has great products and, hands down, the easiest setup of any consumer electronics product I've ever seen. They are fantastic and fair with their warranty and non warranty repairs. Customer service has been great in my experience. I have no plans to sell any of my Sonos components. I love my Sonos stuff and I've told everyone I know to go with Sonos because they make the best products in the category.

That's why I'm riding their butts on a few things, I want Alexa integration, I want an Audible app and I want it all yesterday. I want the software side to improve on some things. By making it better for me, hopefully it makes things better for others. It might sound like complaining because that's exactly what it is. I'm sure it's not easy but I'm also sure they can pull it off and, until they do, I'm going to keep asking about it.

Anyway, that's where I'm coming from. Now what's the big holdup on the Alexa thing??? Let's go! Lol!
Userlevel 7
Badge +23
Where does the API reside? Is this within Alexa?

Yes: https://developer.amazon.com/public/solutions/alexa/alexa-voice-service/reference/audioplayer

To use this with Sonos today you would need to provide a "fake" silent mp3 stream to "play".
Yep, those of us who've been involved with large scale, multiple vendor systems integration projects know how complex and time consuming they can be. Software Engineering has improved a lot over the years, but any project of this scale and scope takes a huge effort just to get through the beta stages. Then there will be ongoing updates and support for years to come, on top of the existing updates and support. Keeps lots of people employed, for sure.

For those who think it's easy, here are the docs. Have at it! Lol

https://developer.amazon.com/public/solutions/alexa/alexa-voice-service/content/avs-api-overview


Involved in one right now. Multiple vendors running the same single vendor app on different POS hardware using a dozen different device drivers, connecting first to a 20 year old legacy OLTP system, then to a brand new, state of the art OLTP host. Dual site locations, 4 redundant hosting systems, countless peripheral systems for operational control, security, web hosting, mobile app data, archival data, reporting, etc. Base software already written, but the timeline is still over 2 years from design to completion for the site specific customizations.
Yep, those of us who've been involved with large scale, multiple vendor systems integration projects know how complex and time consuming they can be. Software Engineering has improved a lot over the years, but any project of this scale and scope takes a huge effort just to get through the beta stages. Then there will be ongoing updates and support for years to come, on top of the existing updates and support. Keeps lots of people employed, for sure.

For those who think it's easy, here are the docs. Have at it! Lol

https://developer.amazon.com/public/solutions/alexa/alexa-voice-service/content/avs-api-overview
Also, Alexa was designed with existing WiFi lights, thermostats, garage door openers, etc., in mind. As mentioned before, the ability to control anything other than rudimentary external audio devices was not there. Certainly not multi-room speakers with grouping, multiple sources, multiple volume controls, etc. All this functionality had to be designed and implemented, and if one thinks about, it doesn't take long to run into logistical problems that need to be solved in order to offer a seamless experience. This also has to be done in conjunction with Amazon, who may or may not be on the same set of priorities or timetable.
I believe that Philips, Logitech, Nest and the others you speak of all have a 3rd party server outside of your internal network. If you can control your lights and things while outside of your network, it almost certainly does. Notice that you cannot access your Sonos system from outside your network.

Philips and the like had good reason to setup those servers, as it was key to functionality to their products. It's way more useful to be able to turn on/off your lights when you aren't there, so it was needed. That was designed before there was any voice control. However, it wasn't as necessary for Sonos, since people don't typically don't listen to music when they aren't there to hear it. So in way, besides lights being easier then audio to control, they had a head start with the integration. Besides that, we don't know how long Philips and Amazon were working on their integration since it was there when Echo was introduced.

Then again, I imagine these smartskills work differently then what I described earlier. Seem to anyway.

When Echo controls without a hub, it's only for some basic things over wifi protocol. Many lights don't use wifi, they use zee wave, zigbee, clear connect, etc, so your Echo doesn't know how to talk to them.
Userlevel 4
Badge +2
Steve_199, I'm no expert, but I think I understand the framework of how Alexa works pretty well. What you think of is pretty easy...really isn't all that easy.

Your typical Alexa skill works as such.
1 - local Echo dot hears a command.
2 - local Echo dot sends command to Alexa servers that converts voice to text, then determines what to do with it.
3 - if the command is for a 3rd party skill, it sends the skill to the 3rd party server for processing
4 - 3rd party server processes and performs action and gives Alexa servers a response.
5 - Alexa 3rd party servers give a response to your local Echo.

So it's somewhat complicated, but beyond that...Sonos doesn't have a server out there to act as a 3rd party server in this scenario. It would have to set that up in order for a basic 'Alexa tell sonos to play xxx'. On top of that, your sonos system doesn't have a presence outside of your local network. There has never been any sort of registration process of such where the Sonos server (that doesn't yet exist) could no one zones you have in order to command it do something.

That said, there are some work arounds. Your local echo isn't completely dumb in that it can do some actions on it's own. That's why there are some bulbs that work with Echo without need for a smarthub...or work with a smarthum. Most of the time there is a smarthub. So if you have smartthings for example, your Echo knows to send a particular command to your local hub for processing. FYI, there are some hubs that can control sonos, and are setup with Alexa. Therefore you can control Sonos from an Echo by way of a hub...but the functionality is limited.

And I think your statement about simply saying "Alexa, stop" brings up another point. With all of the above mentioned, that command won't work...because you didn't tell it what to stop. You would have to say "Alexa, tell Sonos to stop". And Sonos management has seemed to acknowledged that that is far from ideal. Annoying perhaps. Hence, they want to spend the time to get tighter integration so that when you say stop, Alexa knows it's playing music on Sonos and that must be what you want to stop. Or perhaps when you ask to start music, it knows you mean in the sonos zone you're currently in, and not on the echo's speakers.

And sure, they could have gone with being able to drop the whole 'tell Sonos' at a later date, but the development work for going all the way can't be built off of going halfway. So why not just go all the way.

Again, not an expert, but that's what it seems like to me.



Thanks for that, I think you're more of an expert than you let on.

My thoughts are, Philips has a hub and different zones and Alexa works great with it. Logitech Harmony has a hub and I can control my TV through Alexa voice commands. My Nest thermostat works directly with Alexa voice commands. If Philips, Logitech, Nest and others have figured out the integration (some with a hub, others without one) then why can't Sonos? Sonos, of course, can work with a Bridge or directly through wifi so that give them two possible starting points. I'm absolutely not an expert and there's likely a great reason why they can't get this done. I just like to know what the sticking point actually is.

Btw, the "stop" command thing was for the posters saying Alexa couldn't hear them over loud music that was playing through the Alexa app. Otherwise, yes, you'd have to be more specific about what you wanted to stop.

The "Alexa, stop" would not require "tell Sonos" if Sonos uses the audio-API set, and Sonos was the most-recently used audio skill.


Where does the API reside? Is this within Alexa? I'm kind of surprised that it even exist right now, as I was not aware that Alexa was aware of any other audio then it's own. I was assuming that any APIs on the Alexa side would be rather generic, leaving the 3rd party skill to figure out what it is, or what to do with it. However, I can see where leaving off the "tell" command would prompt Alexa to assume you are referring to the same skill previously. I just figured a typical audio command like 'stop', 'pause', 'play' would be assumed to refer to the Echo's speakers, not a skill.


However, the current audio-API set has a poor extension model, so cannot add "in the xxx-room" to commands. It appears that the Sonos/Alexa features as described publically require a bunch of new features in the Alexa infrastructure itself, and Amazon are very slow at rolling out such updates.


Again, assuming you are talking about an API on the Alexa side, isn't there a generic API that Alexa doesn't try and interpret at all, just simply pass to the 3rd party skill? I don't know why you would need a specific audio-api, accept that it could make life a little easier for the 3rd party skill developers.


"Indignant posters" aka frustrated Sonos customers. Frustrated customers voicing their concerns are not the problem, that's called "feedback". It's when we stop caring that it becomes an issue.


If 'we' means Sonos, then I agree to an extent. How much is up to them. "We ', meaning the other customers in this forum...I don't see why we need to care.

I personally would agree that voicing frustration and concerns is positive. But it's not productive when customers threaten to sell their stuff if demands are not met, repeating inaccurate rumors, generally assuming that this is all easy, and Sonos doesn't care.
Reminds me of my childhood vacations in the back of the Family Truckster . . .

"Are we there yet?", "Are we there yet?", "Are we there yet?", "Are we there yet?", "Are we there yet?"

😃
Userlevel 4
Badge +2
The only bad blood I see is from a couple indignant posters. Most are waiting patiently for what was promised to be delivered 6 months from now, as should be expected.

"Indignant posters" aka frustrated Sonos customers. Frustrated customers voicing their concerns are not the problem, that's called "feedback". It's when we stop caring that it becomes an issue.

Look, I know we'll get it when we get it and customers should just shut up about it, but six more months will make it almost 3 years since it was suggested as a feature. I'm reminding Sonos that I've wanted this feature for years and I still want it. They say they've been working on it for years. The next question becomes, ok, so where is it? Maybe requesting an update is out of line, but I don't think so.

Customers that don't care don't bother posting they just move on and buy something else.
Userlevel 1
Badge
I just installed MySpeaker and its skills to control my Sonos System via Alexa. So far it works pretty good. One bad is that its application must be running on a PC or Laptop connected to the same network as Sonos.
Userlevel 1
Badge
Been waiting to hear news about when this will happen and am starting to think it never will. I have Sonos and Alexa in every room. I did this with the assurance that Sonos and Alexa are going to work together. There has been next to no news regarding integration for the past 6 months.... When can we expect to hear about Sonos and Alexa integration? Is it still happening? Please any update would be appreciated.
Userlevel 7
Badge +23
The analysis by melvimbe is pretty good.

Some code is needed on the local network, and the Smart Devices method used for a few lights will not work for Sonos. Code is needed in the Sonos players themselves, I assume this will be using a similar method to the Spotify-direct stuff. (This is also why my Skill requires a Windows 10 PC).

The "Alexa, stop" would not require "tell Sonos" if Sonos uses the audio-API set, and Sonos was the most-recently used audio skill. However, the current audio-API set has a poor extension model, so cannot add "in the xxx-room" to commands. It appears that the Sonos/Alexa features as described publically require a bunch of new features in the Alexa infrastructure itself, and Amazon are very slow at rolling out such updates.

Also Amazon are very US-only in their Alexa support, which is another problem for Sonos who are much more international-thinking in their products.
Steve_199, I'm no expert, but I think I understand the framework of how Alexa works pretty well. What you think of is pretty easy...really isn't all that easy.

Your typical Alexa skill works as such.
1 - local Echo dot hears a command.
2 - local Echo dot sends command to Alexa servers that converts voice to text, then determines what to do with it.
3 - if the command is for a 3rd party skill, it sends the skill to the 3rd party server for processing
4 - 3rd party server processes and performs action and gives Alexa servers a response.
5 - Alexa 3rd party servers give a response to your local Echo.

So it's somewhat complicated, but beyond that...Sonos doesn't have a server out there to act as a 3rd party server in this scenario. It would have to set that up in order for a basic 'Alexa tell sonos to play xxx'. On top of that, your sonos system doesn't have a presence outside of your local network. There has never been any sort of registration process of such where the Sonos server (that doesn't yet exist) could no one zones you have in order to command it do something.

That said, there are some work arounds. Your local echo isn't completely dumb in that it can do some actions on it's own. That's why there are some bulbs that work with Echo without need for a smarthub...or work with a smarthum. Most of the time there is a smarthub. So if you have smartthings for example, your Echo knows to send a particular command to your local hub for processing. FYI, there are some hubs that can control sonos, and are setup with Alexa. Therefore you can control Sonos from an Echo by way of a hub...but the functionality is limited.

And I think your statement about simply saying "Alexa, stop" brings up another point. With all of the above mentioned, that command won't work...because you didn't tell it what to stop. You would have to say "Alexa, tell Sonos to stop". And Sonos management has seemed to acknowledged that that is far from ideal. Annoying perhaps. Hence, they want to spend the time to get tighter integration so that when you say stop, Alexa knows it's playing music on Sonos and that must be what you want to stop. Or perhaps when you ask to start music, it knows you mean in the sonos zone you're currently in, and not on the echo's speakers.

And sure, they could have gone with being able to drop the whole 'tell Sonos' at a later date, but the development work for going all the way can't be built off of going halfway. So why not just go all the way.

Again, not an expert, but that's what it seems like to me.
The only bad blood I see is from a couple indignant posters. Most are waiting patiently for what was promised to be delivered 6 months from now, as should be expected.
Userlevel 4
Badge +2
I think Sonos might have been too ambitious on this idea. They seem to want to be everything to everyone which will usually leave everyone disappointed.

All I want is to say "Alexa, play my 80's rock station on my living room Sonos" and hear her say "ok". If they could have rolled out a few simple skills while they were working on their plan for world domination, they could have avoided a lot of the bad blood.

For those concerned about not being heard over the music, just say "ALEXA, STOP" in a loud voice. She will usually respond to that and stop the music. Then you can give your next request. Just don't scream it too loud or your neighbors might think someone named Alexa is trying to kill you.
How can i get on the Beta tester for Sonos and Alexa?
I have two Sonos Connect:AMPS will be adding 4 more shortly, i also just got my hands on a Amazon Dot. would really love to test this out...🆒

Maybe you have identified the niche use case where Homepod really would be a better choice. On the other hand an ipod loving apartment dweller can use Bluetooth with an Echo Look or Echo Dot connected to some nice powered speakers and save some money by ignoring the Homepod vs Sonos battle completely. It will be interesting to see what ends up appealing to this group.

Yes, it will be interesting though I can't help thinking that in addition to the above, if the Alexa integration is done well, that will be one more effective barrier to adoption of Homepods.
Userlevel 2
I do see one major benefit of Sonos over homepod. There are no walled gardens. You can build a playlist from any number of sources. Mixing amazon, NAS, Spotify. Even the Micro SD card in my Blackberry.

The question I keep asking myself is, what would I recommend to someone who loves their ipod, only cares about music in a small space like an apartment, and has no interest in tv integration.


You kind of need a whole house to get full value from a "whole house" music system like Sonos.

You can use a single Echo in an apartment but you really get more value if you have room enough for multiple Echos for intercom and "whole house" automation.

Maybe you have identified the niche use case where Homepod really would be a better choice. On the other hand an ipod loving apartment dweller can use Bluetooth with an Echo Look or Echo Dot connected to some nice powered speakers and save some money by ignoring the Homepod vs Sonos battle completely. It will be interesting to see what ends up appealing to this group.
Userlevel 1
@Pete, I think your underestimating the HomePod a bit....

Oh, I agree. I work in a company where I fully expect a HomePod on most of the Exec desks the day it's released, even more so if it's conference-call capable. And generally I recommend iPhone and iPads to family members because they just work, and I don't have constant support issues 🙂 But, most of my family don't use Apple Music, so unless they announce other integrations, that'll be a big roadblock for many people. Most of my family do have an Alexa of some sort though, so they're already on the ladder.

It's certainly a marketplace that been waiting for a decent competitor, and if nothing else, HomePod will hopefully light more of a fire under Sonos. Apple's weakness, potentially, will be its obsession with keeping people in their ecosystem. They've already precluded Android users, and potentially Pandora, Spotify, Slacker etc etc, although I'm fairly sure they'll support the other services if not from launch, very soon after.
@Pete, I think your underestimating the HomePod a bit. I imagine there are plenty of people out there who will buy Apple just because it's Apple. They may already have an Echo but wish to upgrade to what they believe will be better.

As far as the cost comparison, Apple is saying the sound is comparable to the PLAY:3, so really the pricing is around the same. But yes, you could opt for the PLAY:1 to greatly reduce cost, or you could already have Sonos or Amazon products, like much of the people on this board.

The question I keep asking myself is, what would I recommend to someone who loves their ipod, only cares about music in a small space like an apartment, and has no interest in tv integration. A part of me wants to recommend Sonos/Alexa for future proofing, but I can't say Apple isn't a bad choice for them.
Userlevel 1
There seems to be enough noise out there to reasonable assume the private beta is well underway. I've been a Sonos use since 2002, and I still have my trusty CR100 controller, and Sonos, bless them, still provide support (although the hardware replacement program has dried up). The point is, that the parallels with Sonos and Apple are grey, to say the least. I do think Sonos has made the right choice to use Alexa for voice control.

I'm one of those nerds that was messing around with voice recognition back in the late 80's and it sucked, big time. I'd say until Alexa, it still sucked, and that's including high-end systems. The go-forward battle is between Alexa (good head start) and Google (vast amounts of data). Apple have squandered a huge opportunity with Siri, and continue to do so. Even the Homepod is marketed as a music-first device, with Siri tagged on; it's not a competitor to Alexa or Home. By taking the best distributed audio system (Sonos) and meshing it with Alexa, where the entry point is $40, is a smart move. A Play:1 and an Dot is still much less than the HomePod, and more functional, and I'm sure we'll see bundles from Amazon once the integration is fully baked.

Having said all of that, I wish they'd hurry up. Assuming Apple hit their deadline this time, the holiday shipping period could be pivotal for Sonos. Since many people new to Sonos will already have an Alexa device, it becomes a $200 (Play 1) vs. $350 (HomePod), and that's a tough sell for many people. So the sooner Sonos go public with this, the better.