I recently bought the Muse Luxe v2 with the expectation that it could work as both a Home Assistant voice assistant and a music playback device at the same time. I already use the Home Assistant Voice Preview Edition, but that device cannot handle music playback, so this was the main reason for choosing Muse Luxe v2.
With the latest official firmware, voice assistant functionality works fine. However, when I try to play music through it, the playback is choppy and stuttering. Based on logs, it seems like the hardware might not be able to handle both tasks simultaneously.
Interestingly, if I disable/remove the voice assistant functionality, music playback works without issues.
So my question is:
Is anyone successfully using Muse Luxe v2 as both a voice assistant and a music player at the same time?
I’ve been trying to create a custom firmware where the voice assistant is disabled during music playback to free up hardware resources, but so far I haven’t achieved a stable result.
Any experiences, suggestions, or confirmations would be appreciated.
Therefore, you can trigger the player either directly from the Home Assistant web interface or by creating an automation that can be triggered by a combination of buttons.
But I’m not sure I fully understand your problem! Could you give me some more details?
The initial problem I had was that in the latest version of Music Assistant, I simply cannot find a way to force the output codec to WAV for this specific player. Maybe I’m just missing it, but it doesn’t seem to be exposed in the UI anymore.
However, I realized that even if I could force it, pushing a continuous, uncompressed 1.4 Mbps WAV stream over 2.4GHz Wi-Fi to a tiny ESP32 chip just to save CPU cycles for the voice pipeline is an architectural nightmare. Even on a robust home network, relying on an IoT Wi-Fi chip to handle that much sustained traffic while juggling background tasks is begging for packet drops and buffer underruns.
I actually ended up solving it by modifying the ESPHome firmware logic. I didn’t remove the voice assistant – it runs perfectly fine when idle. But the moment a media stream starts, the firmware forcefully kills the voice pipeline to free up all hardware resources (CPU and RAM). This allows the ESP to flawlessly decode standard compressed streams (MP3/FLAC) natively, saving network bandwidth. The only catch is that I have to stop the music via a physical button or HA dashboard, but the second the playback stops, the voice assistant boots right back up.
It’s a trade-off I’m perfectly happy with, as I get great audio and reliable voice control without stressing the ESP’s network stack. Thanks anyway for your time.
A lot of work is being done on MA due to sendspin, eventually things will start to work for wrover (stuttering playback is known issue) devices too.
i do not understand why you say the VoicePE is not good for music?, it’s a better device for the job than wrover is and has audio jack out
In its original condition, the speaker is unusable. It is true that I did not try connecting the audio output to any other speaker because I do not have one and, most importantly, it would not look as good.
