I know it is not the "easiest" way of doing it, but I am using a Node.js server running the Sonos HTTP API from jishi (GitHub - jishi/node-sonos-http-api: An HTTP API bridge for Sonos easing automation. Hostable on any node.js capable device, like a raspberry pi or similar.) and send HTTP commands via that API to group/ungroup Sonos players. Works great and reliably for me.
2 Likes
I do use this method also and it works but if I were to try and make rules using RM to announce family arrivals and departures from different locations, it would be easier to have them grouped in HE and select that group to speak random messages. I have a WC piston that does this but if I were to try to get rid of it, I think this would be the best way to have random messages speak on certain speakers.
Why do you need the grouping for announcements with RM? You can select several Sonos speakers as Speech Device. That would speak the message on all of the selected once and continue playing what was on them after the announcement is done.
Yes, I have tried this but it isn't simultaneous. Sounds a bit funny.
Any update on this? We would prefer to Group and Ungroup various speakers through RM if possible.
1 Like
The Sonos device driver needs to be updated to support Join capabilities and then RM could be updated to expose the drivers capability through the RM UI. Of course if the ability was there you could also create a custom command in RM too.
@mike.maxwell could join be added to the Sonos driver based on work others have done using Node.js above?
3 Likes
Any update regarding the ability to Group and Leave Sonos speakers from Hubitat?
This ability would be greatly appreciated.
1 Like
Any updates? Perhaps there is a way to use a groovy device driver to group speakers so that we could create RM rules for various presets? We prefer this over setting up another server running node.js. Hey Google, turn on the party speaker preset!
Anything?
+1 for this!! It would be awesome!
1 Like
Can you explain how you are doing this? This would be a great "app" if we could do something. Thoughts?
I use a node server with the node-sonos-http-api from jishi
It has quite a list of supported sonos actions:
- play
- pause
- playpause (toggles playing state)
- volume (parameter is absolute or relative volume. Prefix +/- indicates relative volume)
- groupVolume (parameter is absolute or relative volume. Prefix +/- indicates relative volume)
- mute / unmute
- groupMute / groupUnmute
- togglemute (toggles mute state)
- trackseek (parameter is queue index)
- timeseek (parameter is in seconds, 60 for 1:00, 120 for 2:00 etc)
- next
- previous
- state (will return a json-representation of the current state of player)
- favorite
- favorites (with optional "detailed" parameter)
- playlist
- lockvolumes / unlockvolumes (experimental, will enforce the volume that was selected when locking!)
- repeat (on/off)
- shuffle (on/off)
- crossfade (on/off)
- pauseall (with optional timeout in minutes)
- resumeall (will resume the ones that was pause on the pauseall call. Useful for doorbell, phone calls, etc. Optional timeout)
- say
- sayall
- saypreset
- queue
- clearqueue
- sleep (values in seconds)
- linein (only analog linein, not PLAYBAR yet)
- clip (announce custom mp3 clip)
- clipall
- clippreset
- join / leave (Grouping actions)
- sub (on/off/gain/crossover/polarity) See SUB section for more info
- nightmode (on/off, PLAYBAR only)
- speechenhancement (on/off, PLAYBAR only)
- bass/treble (use -10 thru 10 as value. 0 is neutral)
For grouping, you can use the join/leave commands:
Grouping
You have basic grouping capabilities.
joinwill join the selected player to the specified group (specify group by addressing any of the players in that group):
/Kitchen/join/OfficeThis will join the Kitchen player to the group that Office currently belong to.
/Kitchen/leaveKitchen will leave any group it was part of and become a standalone player.You don't have to ungroup a player in order to join it to another group, just join it to the new group and it will jump accordingly.
If you want and app, you could create an app that has a virtual child device that supports speech. The child device could send the "text" to be spoken to the parent app and the parent app groups the sonos speakers before it sends the TTS to the master in the group.
Technically, it should work, might have play around with some timing to make sure that commands don't overlap.
EDIT: There is also a "sayall" command that speaks TTS on all Sonos devices. Not sure how synchronized it is though
The node.js app does work well and I already use it. However, if we could get RM to support grouping speakers, then all of us could simply the setup by using the Sonos devices within Hubitat. We would like to see a Leave and Join command added to Sonos devices. kitchenSonos.join(livingRoomSonos).
Is this possible?
Right now it is not possible within RM and it is also not as trivial as it sounds.....
First, the sonos drivers would need to support the grouping.
Once that is done, technically you could group it then via custom commands in RM.
But I can already see the request:
I have 6 sonos devices, my door opens and I want all of them to speak a message and revert back to what they were playing before.
So you technically would need to
a. Retrieve the current status of all Sonos devices
b. stop what they are playing (if you just start grouping that would play what the master is playing)
c. group them
d. speak
e. ungroup them
f. Make them play what they were playing before (if that is even possible depending on what they were playing)
I use two different sets of Sonos speakers. One set for music, the second for dedicated notifications. I can mute the music, then play the notification, then unmute the music. A little more expensive to have a 2nd set of speakers, but much easier than trying to restore groups and previous tracks.
Again, can it be done via Hubitat Sonos device?
You can, together with the node server
You can do something like this:
Ignore the trigger event, I just selected something to have something. I also added some delays to let the joining happen.
Again, you can't do the grouping via RM directly, it is currently not supported
1 Like
I already have it setup this way, trying to simplify and reduce external dependencies. If node.js can group why can't groovy? Thanks for the suggestion but not what I was looking for.
Because the driver for the Sonos devices currently doesn't support it. It is simply not programmed. I am not saying that it is not possible to change the driver. I was trying to give you a solution with the currently available features.
You can make a feature request in the "Feature Request" category here:
https://community.hubitat.com/c/support/Requests
We are currently in the "Lounge" area and I know that the team doesn't monitor that part of the forum that closely.
Trying to expand on things a bit. I've come up with some logic but I need some help expanding it. I'd like to come up with something that resumes the SONOS after an alert has played on another TTS device. The resume ONLY happens if there was something playing before the alert. It's based on the theory that there is one speaker controlling a group.
The constraints are (for now) that the group never changes and that the alerts are being played on something else (e.g. Alexa or Google Home devices).
The variable is that the group may be or not playing something.
So far I've got something like:
Trigger : Whatever it is
Action: Condition for alert = true
(nested if) - IF (controlling media player) is playing THEN
Send GET to http://IP address:5005/pauseall
END IF
Speak alert to something else
(how do I know there was something playing on the Sonos that I paused?)
Send GET to http://IP address:5005/resumeall
END
I'm having a brain lock on testing for the active node of the SONOS playing something and then sending the GET to resume after the alert.
