HSM Audio Alert/Speach on alarm has TWO audio volumes, only one editable, effective one stuck at 50% volume

HSM Water alerts seem to have a hardcoded "50" volume value that I can't seem to get rid of that blasts my speakers when the alert goes off. I'm guessing there's a bug over the versions that caused this extra attribute as non-editable. While I can change the one that shows as "10" I can't seem to figure out how to expose that extra one. V2.3.8.125

Here's my settings in the alert editor:

Here's what I get on the rule summary screen - with a trailing volume I can't change that seems to be the actual value used:

Anyone else see this? Anyone have a hand edit in some file someplace I could attempt?

I'm seeing something similar. It looks like the first volume is from the "on this speech device" option and the second is from the "on this music device" option. You might try selecting a music device and changing the volume on that to 10 also to see if that causes the speech volume to change.

If that reduces the volume, I'd call that a bug, since the music device volume shouldn't be affecting the speech device volume.

The other thing I noticed that seems buggy is if I add a file name in the "music device" options, the file name will display instead of the text to speak, even if I un-select the music device.

I'm using a custom device driver that sends the speech/music requests to Node Red which is connected to my Google Home devices, so it's possible this is just something weird I did in that driver, but I see this:

With a speech device selected, speech at volume 40, no music device selected, music at volume 20, and no file name:
image

However if I select a music device and insert a file name it shows both devices (which are the same device in my example... "My Cast") with their volumes, but doesn't show the text to speak, only the file name:

image

If I un-check the music device but don't clear out the file name field, I get this:

image

So it is showing the speech device, the speech device volume, the music device volume, and the music device file, but not showing the text to speak.

What I would expect is:

  • If only a speech device is selected, don't show the music device volume or file name
  • If both are selected, show both devices, both volumes, the text to speak, and the file name to play

I played around with "Intrusion-Home" and I think it is doing the same thing. That section was only showing one volume before I edited it, but once i added a music device, the second volume was showing. When I then removed the music device, the second volume was still there and there is no way to get rid of it. If I blank out the volume, it automatically changes to '50'.

Tagging @bravenel as I think this area of the HSM UI might need some attention.

@Alan_F Thanks for your confirmation on what you're seeing.

The 50 comes from the music device and remains present in the display even when one isn't defined as you suspected. The inverse is true as well -0- removing speach device leaves that volume present as well.

Second apparent issue is that making a volume adjustment doesn't change the "volume" that wakes the dead on the speak moment. Set at 1 or at 50 the volume remains the same. You can hear it clear across the house. Perhaps this is a Sonos artifact vs a HSM, but there's a chance that the volume isn't being presented by HSM -- I don't have other audio or speach devices in my setup to isolate.

I can also confirm this logic, be it "feature/bug", behaves the same in other areas of HSM.

I'll +1 some eyes on HSM UI, and determining if the volume IS being passed properly by HSM (in which case it's a sonos driver? issue) or if that's a secondary bug. Fortunately, I won't miss the alert going off in the meantime. (Sadly, a gasket issue on my kitchen sink drain didn't trigger it the other day -- a hub reboot addressed the zigbee lack of response, which lead me down this path)

In my testing, the volume set for text-to-speech is working, but again I'm using a custom driver that takes the speak command, turns it into a http request to Node Red, which parses it and sends it on to my Google Home devices. That does make me think the issue may be with the Sonos driver.

My custom driver has capabilities for Notification, SpeechSynthesis, MusicPlayer and AudioVolume. On the device page I have a "Speak" button where I can set a volume and select a voice (although my driver ignores the voice selection).

image

Assuming you have the same button on your Sonos device page, have you tried sending TTS messages from there at different volume levles to see if the volume setting works?