Year of the Voice - Chapter 2: Let's talk

I do have default_config: in my configuration, But I fail to see how HA config will have impact on completely separate containers being started up

I just realised… the error in my log mentions Illegal instruction
Does this not relate to an incorrect architecture or something… Is there a chance that docker could be downloading the arm64 image instead of amd64?

here is my host CPU info:

Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Byte Order:                      Little Endian
Address sizes:                   48 bits physical, 48 bits virtual
CPU(s):                          2
On-line CPU(s) list:             0,1
Thread(s) per core:              1
Core(s) per socket:              2
Socket(s):                       1
NUMA node(s):                    1
Vendor ID:                       AuthenticAMD
CPU family:                      16
Model:                           6
Model name:                      AMD Turion(tm) II Neo N54L Dual-Core Processor
Stepping:                        3
CPU MHz:                         800.000
CPU max MHz:                     2200.0000
CPU min MHz:                     800.0000
BogoMIPS:                        4392.56
Virtualization:                  AMD-V
L1d cache:                       128 KiB
L1i cache:                       128 KiB
L2 cache:                        2 MiB
NUMA node0 CPU(s):               0,1
Vulnerability Itlb multihit:     Not affected
Vulnerability L1tf:              Not affected
Vulnerability Mds:               Not affected
Vulnerability Meltdown:          Not affected
Vulnerability Mmio stale data:   Not affected
Vulnerability Retbleed:          Not affected
Vulnerability Spec store bypass: Not affected
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Retpolines, STIBP disabled, RSB filling, PBRSB-eIBRS Not affected
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Not affected
Flags:                           fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm 3dnowext 3dnow constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid pni monitor cx16 popcnt lahf_lm cmp_le
                                 gacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt nodeid_msr hw_pstate vmmcall npt lbrv svm_lock nrip_save

So how does one go about adding further TTS and STT variations such as English (New Zealand) for local processing as shown in the release video yesterday (May 4) when they were connected to HA Cloud?

Presently only English UK & US is available for TTS and only American English for STT. I have looked through all the documentation I can find but no references to this.

Interesting that I don’t have default config but I also didn’t have assist pipeline and it is working albeit slowly in the STT

@synesthesiam - just for clarity…

We need to install Whisper and Piper containers… that’s it? What about one of the Rhasspy containers? Any advantage to doing one of those (even if early/developer)?

Thanks!

Due to experimenting with frigate and compreface, I have included a dedicated ubuntu/docker PC with an old Cuda GPU into my setup. I still run my Home Assistant on a Raspberry Pi, but try to shift any heavy lifting where I can.

So I have added wyoming-whisper and wyoming-piper onto this PC using the above docker images.

I have since read that piper can be offloaded to a Cuda GPU. Has anyone yet added this configuration to the wyoming-piper image? Should I attempt this and push the change somewhere?

1 Like

Another likely higher-ended M5 Stack option, although ESPHome support might be lacking currently

1 Like

Thanks to @Cadster I was also able to use it on the android app mizudroid, but my goal is to install it on my asterisk server.
Do you happen to know how to configure the sip.conf and extensions.conf to be able to call using Asterisk?

For the cool VoIP solution, would any other ATA (or IP phone even) work, or is there a specific reason we need the grandstream ht801?

Any chance you’ve figured this out yet? Because I would like to know this as well.

Unfortunately no, I have not

Wow the voice assistant really sounds amazing (pun intended)! I’m really looking forward to add some esphome based assistants to some rooms as soon as I can get my hands on the hardware. Unfortunately the atom seems to be sold everywhere.

Has anyone already figured out if there’s a way to tell Home Assistant which room the request is coming from? It feels counter-intuitive to say “Turn on the lights in the living room” when I’m standing in the living room, maybe even pressing a button.

1 Like

Add the docker container multiple times to your compose file, one for each voice.

I run HA in a container and I’ve set up extra containers for piper & whsiper and configured them in HA. I chose “en-gb-southern_english_female-low” for the voice. However, if I click “try voice” for piper, it sounds like mostly gobbledegook. e.g. “Hello. How can I assist?” sounds something like “Feerv. How can my cyst?”

Is this voice package just broken or have I configured something incorrectly?

Thought of that… wasn’t sure it was necessary, thought I’d ask.

Great work! What’s the process of adding new local TTS languages in Piper?

I currently lack Swedish

1 Like

I mean.
There are 2 options, local processing and cloud processing.
Both options are available.

Have you just intentionally overlooked all the information, videos and documentation on Whisper and Piper which run locally ?

Why on earth did you reply to me directly? I was simply helping another user explaining their issue – regardless of my own ideas around voice control (hint: I don’t use it, except in very few select cases).

I have 2 Rhasspy satellites running with Google AIY kits. They have 2 microphones and work from across the room with TV or music playing.
They do have occasional false wake word triggers.

1 Like

Stop pulling me into your conversation and click the right buttons on the forum. You replied to a post of mine that had nothing to do with what you’re saying.

I have no idea what’s wrong, but I’m using HA on Rpi4 and the HA app on my android phone and the NabuCasa cloud.
If I have an external Nabu address on my phone and an internal LAN network, it is not possible to access the phone’s microphone when it is at home because it needs https.
Does this mean that it is not possible to use assist and other devices in the local network? I don’t want to run all devices through the cloud. I assumed that the assist will also connect at home via the cloud and the other devices will remain in the LAN network or have I set something wrong?