Issue when upgrading from 0.9.3 to 0.9.5

Self Hosting
1

Hi all,
I’m testing upgrade from 0.9.3 to 0.9.5

After upgrade, on FS serveur, when using TTS, I get the following error :
“err”: {
“type”: “SpeechCredentialError”,
“message”: “Playback failed to start”,
“stack”: “Error: Playback failed to start\n at startFallback (/home/admin/apps/jambonz-feature-server/lib/tasks/say.js:266:15)\n at TaskSay.handling (/home/admin/apps/jambonz-feature-server/lib/tasks/say.js:425:21)\n at process.processTicksAndRejections (node:internal/process/task_queues:95:5)\n at async TaskSay.exec (/home/admin/apps/jambonz-feature-server/lib/tasks/say.js:124:12)”}

And
“msg”: “Say failed due to SpeechCredentialError, finished!”

Credentials are ok (Credentials didn’t change during upgrade and were ok before, Google vendor status is ok).

Another point, when using google, no audio in result, but with some others speech vendors, such as elevenlabs, same error in logs but audio is retrieved.

I’ll check deeper, but if someone has an idea, you’re welcome.

2

There was a change between 0.9.3 and 0.9.5 with regard to google tts. In 0.9.3 the google api is implemented in node.js whereas in 0.9.5 I believe we have moved this to a freeswitch module. Try verifying that mod_google_tts is loaded.

3

Hi Daveh,
After a while, i’ve worked back on this and finally activated mod_google_tts in freeswitch on feature server.
I had to define my own google credentials json file, what isn’t expected on my ming with jambonz.
Despite this, I still have the same issue :

“err”: {
“type”: “SpeechCredentialError”,
“message”: “Playback failed to start”,
“stack”: “Error: Playback failed to start\n at startFallback (/home/admin/apps/jambonz-feature-server/lib/tasks/say.js:266:15)\n at TaskSay.handling (/home/admin/apps/jambonz-feature-server/lib/tasks/say.js:425:21)\n at process.processTicksAndRejections (node:internal/process/task_queues:95:5)\n at async TaskSay.exec (/home/admin/apps/jambonz-feature-server/lib/tasks/say.js:124:12)”}

And
“msg”: “Say failed due to SpeechCredentialError, finished!”

Would you have another idea for this issue ?