HS4 Sonos Announcing with a different voice

[NC-Lou]

Hi,

I am trying to specify a different voice when speaking announcements. I set up a Speak Something event with the following speak text:

<voice required="Name=Microsoft David">Hello World how are things around here said David</voice>

However when the event is triggered I hear the announcement on the correct Sonos speakers but the voice is always the same. I verified that multiple voices are installed in Windows 10.

Any ideas or pointers to help me out?

Thanks in advance for any help,
Lou

Here is information about my setup:

Date/time=8/11/2020 11:47:43 AM Eastern Daylight Time
Version=HS4 Standard Edition 4.1.3.0 (Windows)
License=Registered
In Virtual Machine=No MFG: dell inc.
Antivirus=Windows Defender
OS=Microsoft Windows 10 Pro - Work Station
OS Version=10.0.19041
Uptime=0 Days 0 Hours 31 Minutes 20 Seconds
Lan IP=10.10.120.150 (HomeSeerServer)
Device Count=510
Event Count=85
Processor:=Intel64 Family 6 Model 158 Stepping 10 at 2.97 GHz
Modules/Threads=140 Modules, 102 Threads
Available Threads=1023
System Load=176 Processes, 0% Load
Free/Total Memory=12.26 GBytes / 15.81 GBytes (78% free)
Free/Total Virtual Memory=12.31 GBytes / 18.19 GBytes (68% free)
HomeSeer Memory Used=103 Mbytes
Plugin Memory Used=8 EXE Plug-Ins using 364 Mbytes
Plugins Installed=AmbientWeather 3.0.1.4, Harmony Hub 4.0.11.0, HS MyQ 4.0.2.0, Philips Hue 4.0.16.0, Sonos 3.1.0.53, Zigbee 4.0.6.0, Zodiac Iaqualink 1.2.0.9, Z-Wave 3.0.2.4

[dcorsus]

Suggest you search the forum on using other voices. This PI uses a text to file principle and I seem to recall that as off HS3, this is a direct call to a text-to-file app as opposed to going through the HS speaker client. Perhaps if you search here a bit, perhaps there are some different install guidance to make it work. I believe, installing the voices in the HS speaker client is not the way, I believe they need to be installed in the Windows TTS settings, perhaps something you did, fuzzy on details. So ..... search

[NC-Lou]

Thanks, will do.

In the meantime, I created files online and I play them directly from the HS4/Media folder. Works really well, except I have to go online and create a new file whenever I want.

Thanks again,
Lou

P.S. Cant wait for the HS4 native plugin, and I will definitely buy it.

[zwolfpack]

Windows 10 provides 64-bit and 32-bit voices. HS3 only uses the 32-bit ones. (HS4 may be a different story).

In my Windows 10 install, for the David voice the 64-bit name is "Microsoft David - English (United States)". For the 32-bit one it's "Microsoft David Desktop - English (United States)". You may need to specify this entire string for the selection to work.

My HS3 Windows setup is on Windows7 for which I purchased some Ivona voices. One of these is named "IVONA 2 Brian OEM" and an example that properly speaks with this over Sonos is

<voice required='Name=Ivona 2 Brian OEM'>Rusty.<break time="1s"/>You are a rotund kitty!

[Guest]

Thanks, will do.

Zwolfpack is correct, you have to specify the 32 bit versions of the voice file. These are the two that you should find installed already:

<voice required="name = Microsoft Zira Desktop">
<voice required="name = Microsoft David Desktop">

By going to the control panel voice configuration window, I am also able to download some (free) additional voice files, but most all are 64 bit versions. I did find a 32 bit "English" voice (nice accent) that I installed:

<voice required="name = Microsoft Hazel Desktop">

You can bring up the 32 bit voice windows configurator application to select or change the default 32 bit voice to one of the above by running "C:\Windows\SysWOW64\Speech\SpeechUX\sapi.cpl". Doing so eliminates the need for the "voice required" option.

If you go to another source for (probably paid) voices, be sure to use the 32 bit versions for Homeseer (at least HS3).