Things were working fine then audio output stopped a few days ago. In the Play Audio module the voice selection drop-down is missing, no TTS voices are available to select.
I saw there was a 2.5.0 module update so I installed it and rebooted. No change. Then tried to remove the module and reinstall but the remove took it back to 2.4.0 which could not be removed (showed 1 node active but none were).
Where can I download the TTS voice files and where do they live on the file system?
Node-Red is running on a Raspberry Pi with Raspbian and Node-Red up to date.
I did not install 2.5 until after audio stopped working. So it broke without a forced update for the Play Audio dashboard module from the admin GUI. But I do need to check the apt logs because I think there were recent Node-RED updates when I did a system update a few days ago.
In the Play Audio module the list of TTS voices is empty. I can no longer select a voice to use. There used to be 3 options to pick from.
When a message payload is sent to the Play Audio module, the Admin dashboard shows “speaking” but no sound is heard.
I do not know if the voice files are physically missing from my system or not. I could find no voice files on the system searching for tts but maybe the name doesn’t contain tts.
This appears to actually be an issue with the Chromium browser on the Pi. The Pi is in a comm closet and not setup with a display and peripherals. I mostly manage it using RDP from a desktop or laptop. What I observed this morning:
If I RDP into the Pi (this is from a Windows 10 desktop), the problem is as I described. In the Play Sound module no sound is heard and the module lists no voices to be selected. This is using Chromium on the Pi to access the Node-RED admin dashboard.
If I use the Edge browser from my Windows desktop and go into the Node-RED admin dashboard (no RDP), speech is heard and voices can be selected in the Play Audio module. (I can select from David, Zira, and Mark.)
So what I experienced as a change in behavior is actually behavior differences between 2 different methods of accessing the Node-RED admin dashboard. So I do not have an issue with Node-RED itself.
There is a third method, of course, for accessing the Node-RED admin dashboard - using a monitor and keyboard directly attached to the Pi. So I moved the Pi to a display and this yields the same results as #1 - no audio when Play Sound is speaking and no voices available for selection. I did play some wav files outside Node-RED and verified sound is working on the Pi.
The net of all this: There is an anomaly with Chromium on the Pi in both hearing sound from the Play Sound module, and in properly showing the tts voice options when editing a flow. But I can do what I need using a remote browser and will move on now...