Ah, I’m running on a separate Pi in Docker. It looks 2.5.7 is still the latest.
@jonaspaulo I’m on 2.5.6. Do the Satellite Site ID’s work properly in 2.5.5?
on 2.5.5 i have server +satellites working ok
@synesthesiam and others. Do you think there is some way how to use Google home mini as a wireless mic? I know it is strictly chained with google but… maybe to record sound and let it upload somewhere to analyze it by STT? Seems like Google has really best platform for commands recognition (mean HW).
Which STT do you have in mind? Google or Rhasspy?
The idea behind Rhasspy is private and local.
If you upload it to Google STT to analyse, you might just as well use Google Home 
I royally messed up the dialogue manager in 2.5.6 and 2.5.7, and am working on fixes now. Stand by 
I’ve seen others ask about this, and so far no one has come up with a solution. I don’t have one, but I’ve heard the hardware is quite good. I guess that’s how they “get you”, by tying it to Google only.
In this particular case, @Jiran can’t because there isn’t Google support for Czech
Updated to 2.5.7.2 and back to the business. THANKS!
1 Like
Yes, exactly, no STT by google. I wonder if it is possible to use this PS4 mic (eye camera) using raspberry zero or nodemcu, just PS4cam>wireless board>Rhasspy. Would this be usable? (talking about latency)
Ah ok, I could not find a way to use the Google Home as a remote mic.
I have been searching for that as well
I do not know how fast a pi zero is, but using that as a satellite might work ok
Please what do you think? I have my PI4 quite far from the place where I want to talk to Rhasspy. So I search for a solution of a distant microphone (maybe wireless). One solution is to use my old PI2mod.B or PI3 with directly attached PS3 mic.
In this case. Is it better to use this pi3 as microphone only (which sw to use?). Or run Rhasspy directly on this PI3 and send command to HASS on another PI4 via mqtt or other way?
Of course I would prefer to have both HASS and Rhasspy on single PI4 but not sure if separated mic via PI3 will be acceptable because of low latency…
Hi @Jiran, I agree with @Romkabouter that a satellite would be the way to go for now.
I’d specifically recommend the “Remote HTTP Server” option, since that’s the easiest to set up. This will do wake word detection + recording on the Pi3, and then pass the recorded audio to the Pi4 for speech/intent recognition.
Later on, you might try the “Shared MQTT Broker” option, since it’s a bit faster with Kaldi. This is because once the wake word is recognized on the satellite (Pi3), audio is streamed directly into the base station (Pi4) and processed as you’re speaking it.
First of all, thanks for this project - it would really round up many things once I can get it up and running 
I am trying to run it within a docker container, running on a rpi4 within a Kodi/libreelec installation.
Tts, intent recognition and handling is running on a different server and is connected to a HA installation. I somehow managed that part using various docker container successfully.
But now i am stuck at the audio playback and recording. I can see a long list of devices in audio in- and output settings but none seem to work. I have tried to use a usb sound/microphone adapter, hmdi tv, as well as the re-speaker 4mic. But I cannot get any sound in or out.
Any help is appreciated.
Ps: In case you are not familiar with it, Kodi is somewhat limited to what you can install on it (no apt-get).
1 Like
I assume you’re adding --device /dev/snd:/dev/snd to the docker run command, right? If so, there may be a conflict between Rhasspy and Kodi. I’d try these two things:
- Switch between PyAudio and ALSA in Rhasspy for audio input
- Create virtual sound devices for both Rhasspy and Kodi based on this issue response
Just realised that i never responded here. Sorry 
I’ve managed to get the audio (using a different microphone) in the meantime and i have an “almost” working satellite setup so far. Still struggling with this error: https://community.rhasspy.org/t/nluexception-not-enough-values-to-unpack-expected-2-got-1/1880
Great 
all is working fine, but only one small problem. I have no sound. I have tried the aplay option for audio-out, and at he “Available Devices:” option i tried JACK and DEFAULT (when i use default the line above starting with “Device” is empty)
I want my audio get out of the jack-output not HDMI.
I already using Mopidy and sound is working nice. Maybe there is a conflict i created when installing Mopidy? Maybe screwed-up some settings? I don’t know where to start.
Also… when i use the aplay command on my Raspberry Pi, i see it is playing, but i hear nothing!
Would be great if someone can help me or can tell me how to bypass this problem. I only need the audi-out for the “Wakeword” .
Thanks
@synesthesiam just installed your addon, can see the Web interface, good start! Now my problem is that the HA server is in the office and I was just wondering what was the easiest way to have a remote microphone. Now there is no possibility to use the mic of the phone, with the HA app installed? Or do I need to install a Rhasspy with a microphone and HP in the living room? Thanks.
Is there a way to play voice announcements through browser instead of speakers or Rpi’s internal audio jack ?.
Here is the post url for details of where i am stuck at
Any help will be highly appreciated as this is driving me nuts
You must have a dedicated raspberry pi then as a satellite. I am actually setting mine up. Satellite works well now intents are needed .is there a custom component for it ?
Did you add something like this for the microphone?
Ok, same brand, Seeed, did you need a special driver? Happy with?