Yes, exactly. 
I’m stll waiting for the parts (speaker and USB female port) from Ali, will start 3D print the parts tomorrow.
The speakers looks like the one you link to but are blue instead of red.
https://www.aliexpress.com/item/1005006267193796.html?spm=a2g0o.order_list.order_list_main.10.300c1802oLpfdM
1 Like
Yes they will fit.
FYI, in case anyone of you following this thread missed it, “the biggest announcement of the year has yet to come though” as Nabu Casa will launch some kind of first-generation Home Assistant Assist Satellite voice smart speaker hardware product during the the live stream of Voice: “What’s new in voice - Chapter 8” on the 19th of December 2024 (which now state “Join us as we announce the voice assistant hardware we’ve been working hard on all year and see what’s new since our last Voice chapter!”), so be sure to check that out as this scene about to explode when those ship and arrive:
They mentioned during the last Home Assistant release party that it will have immediate availability and that they have already manufactured A LOT of them so should not be a problem for loads of users to get it. A lot more info purposely “leaked” about that voice hardware in the Home Assistant 2024.12 release notes, read → 2024.12: Scene you in 2025! 🎄 - Home Assistant
From previous leaks we know that the hardware from Nabu Casa is based on the same XMOS XU316 DSP chip for advanced audio processing + ESP32-S3 running their fork of ESPHome in the Home Assistant Voice PE repo which is still seeing a lot of activlity in the form of new commits coming in via pull reqiests, see and follow → GitHub - esphome/home-assistant-voice-pe: Home Assistant Voice PE
Yup, waiting for the release in a week - but I personally think I won’t buy it, at least not right away. Respeaker does it for me. Especially due to the fact we’re using basically same hardware, and same firmware.
2 Likes
Diversification is important too because it can drive innovation, both for software and hardware, which is why I think it so great that most of these projects are fully or partially open source, and why it cool to also see hardware products like ReSpeaker Lite and FutureProofHomes Satellite1 coming to market relativly early to compete with Nabu Casa’s hardware products and software projects.
I do however really hope Nabu Casa will try to to push all code with new features and functions as well as new all audio related custom components to the upstream ESPHome mainline repositories to mainstream most of that so easier for other projects to use it (instead of only keeping that in their thier forked “home-assistant-voice-pe” repository and other repositories:
2 Likes
Are there any new findings concerning the usage of two I2S modules in the XMOS firmware of Nabu Casa instead of using only one as it seems to be the case with the default respeaker lite I2S firmware?
Is it beneficial to use two I2S modules?
Nabu casa ESP32-S3 yaml:
formatBCE project yaml:
Yes, competition is great.
And yes, Nabu devs are working hard to put everything from PE repo to the ESPHome. MWW, Nabu media_player are there already.
1 Like
This is guided by hardware difference. Nabu PE device has separate i2s channels for mics/speaker. Respeaker Lite de s (and FPH SAT1 devs too) decided to use duplex I2S for both. Similar pinout is used in ESP32-S3-BOX.
Single I2S saves GPIO pins. But it has limitation: you can’t use different sample rate for speaker and microphone. It’s either 16kHz, or 48kHz for both.
16kHz is enough for speech, so Nabu devs made all the parts of input pipeline work on 16kHz. But for decent sound 48kHz are needed on speaker. So Nabu code is using 48kHz for output. In their case it’s easy, because they have separate I2S and can assign different sample rates.
But for Respeaker Lite there was a tough time. First we had 16kHz firmware exclusively. It was fine, but the sound was not so good and unacceptable for music playback. Then I asked guys from Seeed for 48kHz firmware. Eventually they delivered it. But microphone wasn’t working, because MWW and VA components of Nabu code are expecting 16kHz.
Then Respeaker community made fork for nabu_microphone repo, and reduced sample rate there with simple resampling. So now my code is using that fork.
In conclusion, we have difference between PE and Respeaker, but both working good.
2 Likes
Got mine today, and I´m pretty impressed by the range and accuracy, I´m using vosk on a crappy NUC clone with low specs and the response comes almost immediately! If someone needs a minimalist enclosure for only the module (i.e. response played by external player) I made this. You can find it on thingiverse.
Unlight
1 Like
Are you running vosk on docker? If so can you point me to install instruction. It’s driving me crazy.
1 Like
Is the new enclosure ready for release?
Pretty much. Will finish the tutorial this week.
1 Like
It will make a great Christmas present. Thanks!
It will make you buy rotary encoder and solder a lot of wires… 
no problem. Just so I can get a head start on this, which rotary encoder do you recommend?
I use KY-040 in the enclosure, it fits well. I guess some others would fit too, but this one is meant for it.
P.S. also I use BTF lighting WS2812B LED strip 144 diobes per meter. One foot will be enough roughly (43 LEDs).
Ok, thanks, I’ll put in an order.
Created new repository.
Meet Koala!
After Voice PE release, and before FPH release, this one sits right in the middle, i guess. 
Better speaker, fancy LEDs, distinctive design - and completely open source. You can assemble it from parts yourself.
If you don’t want to tinker yourself - i will be selling ready-made devices.
But it’s completely unnecessary.
4 Likes
Very nice project. Can you share about its circuit? I’m abit confuse about how we can connect the respeaker lite to the esp32-s3 devkit