Hi,
I have configured my M5 Atom Echo by following the instructions on $13 voice remote for Home Assistant - Home Assistant and it works like a charm.
But I would also like to use the device as media_player in addition. How can this be configured? Is it possible to use it as voice assistant AND as media_player?
Thanks,
Riesi
Should be possible, but I wonder what Atom Echo you have, that you want something like music played on that speaker…? It barely gets out voice in a usable volume, how should that work?
Anyway, the configuration should be straight forward, add the media_player
to the code and flash it with the new firmware.
Thanks. My usecase be speech output based on “say …”. That should be ok - it also works when using assist.
Unfortunately I installed my Atom Echo only according to the link like a recipy. How can I change the code of the existing assist firmware and flash it? Is there some guide I can use?
I was hoping for a more specific link, but thanks anyways. What I found out so far is that this: https://github.com/esphome/firmware/blob/3462e80829099395613b550dbfa28518292ae01e/voice-assistant/m5stack-atom-echo.yaml is the configuration file that I guess I have to change.
And then flash it by using GitHub - esphome/esphome-flasher: Simple GUI tool to flash ESPs over USB
Flashing is possible over the air.
We need a little more information about your setup.
Do you have ESPHome installed? If so, how (as an Add-on, stand alone install or …)? What HA are you running (HA-OS, core, venv, …)?
If you have ESPHome running, is the device added to your dashboard?
This all depends on what you already have in terms of installations.
My best guess would be, to re-flash the device with the firmware from your ESPHome dashboard and work your way up from there. I personally find the guide a little incomplete, as it doesn’t state how to get the device into ESPHome afterwards. This is intended in this guide, but it doesn’t help if one wants to go further.
Let us know the answers to the questions above, and we’ll find a good solution for you.
Thanks. I have HA running on a raspberry pi in a docker environment. HA Version 2023.7.
During the installation following the guide the Echo was recognized and the ESPHome integration was installed (ESPHome - Home Assistant). There is no ESPHome addon or standalone running.
In addition, I have on the same raspberry, but in a different docker-compose, whisper and piper running and integrated via wyoming protocoll integration.
I have also attached some screenshots of the configuration.
You’ll need to install esphome.
I’m writing a small guide what to do. I’ll post a link, when ready.
EDIT:
Here we go:
It looks like you cannot have it as media player on the esp-idf
framework that the m5stack-atom uses when configured as a voice assistant - you get a compatibility error message if you try to add a media_player
section.
If the voice assistant pipeline knows how to address it for speech output, isn’t there really a way to just trigger it with a TTS engine, preferably Piper.
It is really a bummer that the media_player component is not available when using esp-idf. Not using an Atom Echo here but another esp32 with a bit more of a decent speaker. Hopefully this gets possible sometime.
There is an issue open on the github page of esphome to extend that. I think the more “like” it, the more the probablity it gets realized. It’s here: Add media player to echo voice assistant · Issue #77 · esphome/firmware · GitHub