Muse Proto Repeating TTS messages

RASPIAUDIO
1

I’ve been using the Muse Proto since its release to allow Homeassistant to make TTS announcements at home. It’s been very successful, although sometimes updating it via ESPhome has thrown up errors, they haven’t caused problems.

In the last month (probably as Homeassistant did its monthly update) any TTS messages work, BUT repeat about 80 seconds later, and keeps repeating.

Sending a MP3 to the unit only produces one playback. Sending TTS to other playback devices is fine as well. So it’s just TTS messages to the Muse Proto that is the problem.

Homeassistant version 2025.4.2 running HAOS on a x86 PC, OS15.1

ESPhome version 2025.3.3

Raspiaudio version is the most current, succesfully uploaded via your website I think 2024.9.xx

I have seen one other person report this so far.

Hope you can come up with a solution, as the Muse Protos ar so useful!

Andy

2

Hi, thanks for signaling this is scheduled for this week, we will keep you posted here.

3

Hi,

There a new version on the store Proto-HomeAssistant

yaml is here

Let me know if there is still an issue

thx

4

Hi!
And thanks for looking in to this: On compilation I get this…

Failed config

psram: [source /config/esphome/raspiaudio-muse-proto-381c08.yaml:95]

Octal PSRAM is only supported on ESP32-S3, not ESP32.
mode: octal
speed: 80000000.0
enable_ecc: False

Thanks again! (and sorry…)
Andy

5

More info from later:
Compile with OCTAL - fail
Replace OCTAL with QUAD - many immediate python looking errors
Power cycle, retry - SUCCESS in compilation in 600 seconds. Many Yellow warnings, BUT…
TTS speech perfect and clear, no repeating.
MP3 clip played
Then somehow it went into an error loop.
Power cycled.
Back to original poor audio with repeat.
Will try again tomorrow.

BTW I am a hardware engineer, not a coder!!!

Andy

6

Hi Andy
You can delete lines 90,91,92 (psram:) from the proto_microWW.yaml file. They are there by mistake… (And esphome reports an error since the very last version (2.4.0))… Sorry
For the other difficulty, we can’t reproduce it here!
Maybe, you could check your use of Home Assistant and Assist…
Keep me informed…
LR

7

Hi,
My setup is NOT to have a full voice assistant, but just a way to make TTS announcements and play notification MP3s. So, I removed from your code all references to Voice Assistant, Microphone, and Wake Word, as I thought that could be part of the problem (my HA setup not configured for VA)

So, with simplified code, and the PSRAM bit removed, it’s now working well as I require. I therefore can’t say the full Voice assistant mode is Ok.

Perhaps it’s worthy you offering a simpler more compact code option for people who just require the audio output side?

Thanks again for your help!
A