-
Notifications
You must be signed in to change notification settings - Fork 65
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Help needed with AudioKit #121
Comments
D4 and D5 are for stream and wifi, so yes :)
This is correct, the software cannot do local hotword detection, so streams the audio to Rhasspy server over MQTT
In general I think red is not good, but the audiostream seems to work. Is there anything in the log of Rhasspy? |
Hi, thanks for the quick response. This is in my rhasspy logs:
Could this be the issue?
P.s Just restarted Rhasspy and now I don't see it in the logs anymore |
I also noticed that the option 'Activation' 'Key press only' does not work. If I use that option I do notice that the satellite stops sending audioframes over MQTT. But I would expect to se audioframes again if I press the KEY1 button. I do not see anything happen if I hold that button down or press it only once and start talking. Only 'Remote Hotword detection' seems to send audioframes. The log which is streamed live from my Rhasspy server is also showing nothing. I only see debug messages when I am using the Atom Echo M5. I am using Rhasspy version 2.5.11 Also things like setting volume using Rhasspy does not seem to work. I can see that the command is send both in logs and in the MQTT explorer:
But if I check the satellite I still see the volume on 1. |
@Romkabouter I have a ES8388_V2. Could it be that it detected the wrong version and that this could be the reason that my KEY1 is also doing nothing? |
@rcjkierkels I did not create that device nor the algorithm, but if I check the code KEY1 will always be GPIO_NUM_36:
So key_listen will always be GPIO_NUM_36. |
@Romkabouter I investigated the audio transmitted by the AudioKit and it makes sense that Rhasspy does not detect any wakewords because its rubbish what it transmits. I only hear some noise. I used the code here to capture the audio: When I test this with the atom m5 I do clearly hear me speak. Could there be something else wrong? |
Ok. Saved hotword detection again (no changes) in Rhasspy and now hotword detection also seems to work. |
I have create a new release Please test and reopen if not fixed :) |
Hi,
I previously got this working with an Atom Echo M5 but because of the poor recognition I am now trying an ESP32-AudioKit V2.2 (ES8388)
Unfortunately I am not getting it to work. I successfully flashed the device using
device_type=2
. After rebooting the device (siteId: satellite2) the LED D4 and D5 is permanently ON. Is that correct?I am using MQTT Explorer to check the MQTT traffic and see constantly audio frames coming in:
This is my satellite config:
This is what I configured in Rhasspy:
I does not matter if I press KEY1 or I just try to say the wakeword. Nothing happends and the D4/D5 leds never change. They are always ON (red).
Does anybody know what could be wrong?
The text was updated successfully, but these errors were encountered: