Hello, I'd like to use several Voice PEs as a synchronized multiroom players. In the presentation video there was mentioned that the Voice PE has slimproto player embedded, but I didn't discovered it. So is there possibility to stream synchronized music from Music Assistant to several Voice PEs? Either slimproto or snapcast?
#Synchronized multiroom music with Home Assistant Voice PE
devout remnant
river moss
Not at the moment.
devout remnant
Is it planed in near future? Or, could I try to update esphome yaml? My main intent is to have voice assistant in each room which is also capable to play music which must be synchronized somehow. At least slimproto. Snapcast is preferable.
If I'd like to play with it, which yaml is the main one? I see several ones in repo https://github.com/esphome/home-assistant-voice-pe?tab=readme-ov-file, but README is not too chaty yet.
What's the difference bwtween home-assistant-voice.yaml and home-assistant-voice.factory.yaml for example? Do I guess correct if I think that the voice-kit.yaml is the one I could start to tweak?
And if I break something, then reflashing by https://esphome.github.io/home-assistant-voice-pe/ should revert to the factory settings?
river moss
There's no way to use slimoroto/squeezelite features in ESPHome.
IDK if it's considered as feature at all.
devout remnant
I found several forks of snapcast client for esp32, so I'd like to try to incorporate it into the Voice. But I don't understand differences between yaml esphome files in voice repo. Why there are 4 files for single device?
Actually I prefer snapcast over squeezelite because of theoretical better sync. On the other side I didn't try squeezelite yet. Perhaps it's enough...
river moss
I found several forks of snapcast client for esp32, so I'd like to try to incorp...
I have several ESP32-s3 with squeezelite. They work good. Sync works with MA. But this is not ESPHome.
Merging dedicated media player firmware with ESPHome, that is built for multi-purpose usage of ESP chips, will lead to many problems with productivity.
devout remnant
I understand this. But I'd like to try. That was my main reason to not buy additional esp32 with DAC, but wait for Voice. To have a voice assistant capable of synced audio. Otherwise I could buy either hifiberry for my RPis or try to setup esp32s with i2s. But trying to test snapcast player with esphome worth the effort for me.
river moss
I understand this. But I'd like to try. That was my main reason to not buy addit...
Then try it. 🙂
devout remnant
I'm just confused by Voice yamls. Why there are 4 files? With which one I should start to customize the Voice?
river moss
They're adding up. The base one is https://github.com/esphome/home-assistant-voice-pe/blob/dev/home-assistant-voice.yaml
Just check contents, and you will figure it out.
devout remnant
I noticed that one is just a placeholder, but 8mb and factory seems to include others which confuses me. So the home-assistant-voice.yaml is the one that is flashed into the Voice? What is the purpose of home-assistant-voice.factory.yaml and home-assistant-voice.8mb.yaml?
river moss
Probably this is a bit complicated to understand ... 8mb version is for ESP chips with 8MB flash. .factory file is for flashing devices, that can be adopted without explicit WiFi settings in YAML (because of Improv), and updated from centralized location.
They both include base file, as it contains main logic - and they change/add to it.
devout remnant
So my understanding is that I play inside the home-assistant-voice.yaml, but I'll flash the 8mb file, right? Just double checking to not brick my $59 device right after purchasing 😉
Now I see it. The Voice has 8MB RAM, not Flash. So I'll flash the base file with 16 mb flash home-assistant-voice.yaml, right?
river moss
Yes
devout remnant
Thank you for confirmation. So it's time to play 😉
river moss
Please keep us posted! It's always interesting to add functionality! 🙂