Hi there,
Meanwhile I have Rhasspy (Debian CT on a NUC) running and today I created my first satellite (RPi 3, Docker and a USB soundcard).
everything seems to work, I call “Jarvis”, I hear the confirmation and he executes the command I give him BUT he doesn’t answer me.
If I ask for the time, for example, he understands it and the following line is also visible in the log:
Thank you for your prompt reply,
my settings:
master:
Dialogue management is disabled
Intent handling is disabled
satellite:
Dialogue Management: Rhasspy
Intent handling is disabled
I use FHEM, the command arrives and a response is also generated:
Thank you for your prompt reply,
my settings:
master:
Dialogue management is disabled
Intent handling is disabled
satellite:
Dialogue Management: Rhasspy
Intent handling is disabled
I use FHEM, the command arrives and a response is also generated:
lastIntentPayload:
{“Room”:“dach”,“Type”:“Temperatur”,“input”:“Temperatur im dach”,“intent”:“GetNumeric”,“probability”:1,“rawInput”:“wie warm ist es im dach”,“requestType”:“voice”,“sessionId”:“288ff6a9-b064-4b36-9b60-817372ee6d05”,“siteId”:“master”}
lastIntentTopic:
hermes/intent/de.fhem:GetNumeric
voiceResponse:
Die Temperatur von dach beträgt 19,1 Grad
But nothing arrives at the loudspeaker.
P.S. the beeps after the hotword are output.
When I ask for the time, NOTHING comes to the speaker even though it is generated by Rhasspy, right?
I’m a bit confused because you’re first talking about asking for the time and now you give an example of the temperature. So which script is subscribed to the GetTime intent and responding with a reply to end the session? I don’t see this in your post.
I’m sorry I messed up something myself.
At first I thought that Rhasspy could tell me the time by default. But then I tested it with the temperature above FHEM, change apparently is given back to Rhasspy.
The fact is that there is no speech output.
I have never used the remote HTTP way (I’m an MQTT aficionado), but I noticed the site ID of the master in the logs you showed. I would expect this to be the site ID of the satellite if you’re speaking to your satellite.
Had the same question amd got it working after great help from romkabouter on here, with MQTT though. Messed arond with the HTTP posting method as well but didn’t get it to work. Probably was my own fault, though I can’t reproduce it, so not idea what went wrong.
In my (working) set-up he server is running Home Assistant (hass.io) in a VM with the mosquitto add-in. The sat is a RPi4. Rhasspy sends events to HA, which are processed there and a response is sent back to the sat. The server doesn’t have a mic.
Thank you, I´ve tried the same settings but i doesn´t work.
Here is my complete summary including screenshots:
The master (192.168.66.51) runs on a NUC without audio output or input.
The satellite-dach (192.168.66.240) runs on a RPi 3 with a respeak2 and should be responsible for the audio output and input.
FHEM is connected to the master and also executes commands that I enter via text or voice input on the satellite-dach or via text on the master.
What I absolutely cannot get to work is that I can get voice output on the satellite-dach.
The audio itself works, I get the “beeps” after entering the hotword and even if I enter a text in TTS, I get it over the speakers on the satellite-dach.
Ah I missed that you’re using FHEM, I have no idea how FHEM works,
But I think you want to point to the same external MQTT server in both Rhasspy instances and enter the user/pwd. Is there a MQTT server running @.51?
I HA, the MQTT Broker has to be set in up with an add-in, as OC2019OC explaines here thankfully.