installation issue: PyAudio

[vadalp]

Hi,
I started with a new install of Raspbian Stretch full. Followed your YouTube installation procedure. I do have an AIY kit.

I got an error of PyAudio failed due to BDist Wheel issue. As everything is handled by the script, i saw the error and it flew by. End result is that Alexa worked. Google Assistant didn't. Google Assistant complained the IS_MUTE flag was set and the input device was busy.

So, i decided to install PyAudio using pip

pip install PyAudio.

It went fine but, I guess the whole thing broke after a reboot.

Companion service didn't even start. Once again, the status report complained about input devices being busy.

do you have any idea what is going on?

Did you install with Jesse or Stretch?

V

[shivasiddharth]

The build failure is a warning nothing to be worried about. Just after that, you will see a line that says pyaudio installed. So you need not install it again. The reason for google assistant's misbehaviour is due to the audio configuration. AIY script deletes the .asoundrc and Raspbian auto-populates the .asoundrc file. I have just updated the git to address that. After installing the audio configs and before starting the installation, just make sure that the contents of .asoundrc are same as asound.conf Please try the installation once again using a new copy of OS.

[vadalp]

So I tried a clean install.

The audio-test.sh failed. but it is not clear exactly what is going on. May be you can clear it up.

So, the Assistants-Pi/Startup.wav and /usr/share/sounds/alsa/ files dont play. Give out the following error.

Playing WAVE '/usr/share/sounds/alsa/Front_Left.wav' : Signed 16 bit Little Endian, Rate 48000 Hz, Mono aplay: set_params:1299: Sample format non available Available formats:

• S32_LE

But it plays the mic_test.wav that was recorded. Playing WAVE 'mic-test.wav' : Signed 32 bit Little Endian, Rate 48000 Hz, Stereo

Only format it plays is S32_LE. I am going to proceed with the installation. Sounds more like some settings that can be changed later.

[vadalp]

So after completing the final steps. I still got the same errors as before this time i captured the error for google assistant.

ALSA lib pcm.c:8403:(snd_pcm_set_params) Rate doesn't match (requested 16000Hz, get 48000Hz) [16693:16708:ERROR:audio_input_processor.cc(755)] Input error ON_MUTED_CHANGED: {'is_muted': False} ON_START_FINISHED ON_ASSISTANT_ERROR: {'is_fatal': True} [16693:16709:ERROR:audio_input_processor.cc(755)] Input error ON_ASSISTANT_ERROR: {'is_fatal': True}

[shivasiddharth]

Copy the contents of asound.conf into asoundrc and try running the assistant.

[vadalp]

I copied the asound.conf content from this site into .asoundrc.

Reboot resulted in the following: Now Google assistant works perfectly but Alexa wakeword service keeps restarting. Hence,it says "Hello" every few minutes.

[shivasiddharth]

Did you authenticate the Alexa, before restarting? It is easy to miss out on that ! If not done, then stop the alexa related services using:

  sudo systemctl enable companionapp.service
  sudo systemctl enable client.service
  sudo systemctl enable wakeword.service

and manually run the following scripts. companionstart.sh clientstart.sh They are inside the Assistants-Pi scripts directory

[vadalp]

Yes. Alexa was authenticated at the time of install. Infact before updating .asoundrc, Alexa services were running fine. but no voice interaction could happen.

[shivasiddharth]

Then please post the outputs of:

sudo systemctl status companionapp.service
sudo systemctl status client.service
sudo systemctl status wakeword.service

[vadalp]

● companionapp.service - Companion Service Loaded: loaded (/lib/systemd/system/companionapp.service; enabled; vendor preset: enabled) Active: active (running) since Thu 2017-11-16 10:04:30 EST; 55min ago Main PID: 372 (bash) CGroup: /system.slice/companionapp.service ├─372 /bin/bash /home/pi/Assistants-Pi/scripts/companionstart.sh ├─398 sudo npm start ├─445 npm ├─774 sh -c node ./bin/www └─775 node ./bin/www

Nov 16 10:04:30 raspberrypi systemd[1]: Started Companion Service. Nov 16 10:04:31 raspberrypi sudo[398]: pi : TTY=unknown ; PWD=/home/pi/alexa-avs-sample-app/samples/compan Nov 16 10:04:31 raspberrypi sudo[398]: pam_unix(sudo:session): session opened for user root by (uid=0) Nov 16 10:04:39 raspberrypi bash[372]: > alexa-voice-service-sample-companion-service@1.0.0 start /home/pi/alexa Nov 16 10:04:39 raspberrypi bash[372]: > node ./bin/www Nov 16 10:04:39 raspberrypi bash[372]: This node service needs to be running to store token information memory a Nov 16 10:04:41 raspberrypi bash[372]: Listening on port 3000 Nov 16 10:59:32 raspberrypi bash[372]: Successfully retrieved access token for session id:

xxxxxxxxxxxxxxxxxxxxx

● client.service - Client Service Loaded: loaded (/lib/systemd/system/client.service; enabled; vendor preset: enabled) Active: active (running) since Thu 2017-11-16 10:04:30 EST; 56min ago Main PID: 373 (bash) CGroup: /system.slice/client.service ├─ 373 /bin/bash /home/pi/Assistants-Pi/scripts/clientstart.sh ├─ 406 sudo mvn exec:exec ├─ 443 /usr/lib/jvm/jdk-8-oracle-arm32-vfp-hflt/bin/java -classpath /usr/share/maven/boot/plexus-clas └─1004 java -Xbootclasspath/p:/root/.m2/repository/org/mortbay/jetty/alpn/alpn-boot/8.1.6.v20151105/a

Nov 16 10:59:32 raspberrypi bash[373]: Logged in. Nov 16 10:59:32 raspberrypi bash[373]: Welcome to AVS! To interact with this app, type one of the following comm Nov 16 10:59:32 raspberrypi bash[373]: Login/Logout: login or logout Nov 16 10:59:32 raspberrypi bash[373]: To listen: /l/listen, or just press enter! Nov 16 10:59:32 raspberrypi bash[373]: Next song: n/next Nov 16 10:59:32 raspberrypi bash[373]: Play/Pause: p/play/pause Nov 16 10:59:32 raspberrypi bash[373]: Previous song: b/back Nov 16 10:59:32 raspberrypi bash[373]: Switch locale: en-US/en-GB/de-DE Nov 16 10:59:32 raspberrypi bash[373]: Quit: q/quit/exit Nov 16 10:59:32 raspberrypi bash[373]: Display this help text again: h/help

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

● wakeword.service - Wakeword Service Loaded: loaded (/lib/systemd/system/wakeword.service; enabled; vendor preset: enabled) Active: active (running) since Thu 2017-11-16 11:01:20 EST; 44s ago Main PID: 1155 (bash) CGroup: /system.slice/wakeword.service ├─1155 /bin/bash /home/pi/Assistants-Pi/scripts/wakeword.sh └─1156 sleep 45

Nov 16 11:01:20 raspberrypi systemd[1]: Started Wakeword Service.

The last service restarts every 45 seconds.

[shivasiddharth]

your companionapp and client service are running normally as they should.. Which is the actual alexa itself So there is no issue with Alexa, its the just the snowboy which needs to be troubleshooted.

stop the snowboy service using: sudo systemctl status wakeword.service open a terminal run the following and getback to me with the output:

cd /home/pi/alexa-avs-sample-app/samples
cd wakeWordAgent/src && sudo ./wakeWordAgent -e kitt_ai

[vadalp]

So, I stopped the service. The directory changes to ~/alexa-avs-sample-app/samples/wakeWordAgent/src but complains sudo: ./wakeWordAgent/: command not found.

if you are looking for snowboy directory in kitt_ai directory. It does exist.

[shivasiddharth]

Can you run the installer script again. this time choose to install Alexa alone. Since most of the stuff has already been made it should breeze through this time. During installation watch for the part when it starts cloning snowboy.

[vadalp]

After Alexa authorization, i tried running the wakeword command. I get the following error:

ALSA lib pcm.c:2495:(snd_pcm_open_noupdate) Unknown PCM cards.pcm.phoneline ALSA lib confmisc.c:1281:(snd_func_refer) Unable to find definition 'defaults.bluealsa.device' ALSA lib conf.c:4528:(_snd_config_evaluate) function snd_func_refer returned error: No such file or directory ALSA lib conf.c:4996:(snd_config_expand) Args evaluate error: No such file or directory ALSA lib pcm.c:2495:(snd_pcm_open_noupdate) Unknown PCM bluealsa ALSA lib confmisc.c:1281:(snd_func_refer) Unable to find definition 'defaults.bluealsa.device' ALSA lib conf.c:4528:(_snd_config_evaluate) function snd_func_refer returned error: No such file or directory ALSA lib conf.c:4996:(snd_config_expand) Args evaluate error: No such file or directory ALSA lib pcm.c:2495:(snd_pcm_open_noupdate) Unknown PCM bluealsa ERROR:Failed to open PortAudio stream.Device unavailable ERROR:KittAiSnowboyWakeWordEngine: Initialization error:Failed to open PortAudio stream. ERROR:WakeWordAgent: exception in constructor: Failed to open PortAudio stream. ERROR:main: Exception happened: Failed to open PortAudio stream. INFO:Shutting down Wake Word Agent

[shivasiddharth]

Good !! We are almost there. I really appreciate your patience. Sometimes, raspbian modifies the audio configs, so please post the outputs of following:

sudo nano /etc/asound.conf
sudo nano ~/.asoundrc

[vadalp]

sudo nano /etc/asound.conf no such file exists.

sudo nano ~/,asoundrc options snd_rpi_googlemihat_soundcard index=0

pcm.softvol { type softvol slave.pcm dmix control { name Master card 0 } }

pcm.micboost { type route slave.pcm dsnoop ttable { 0.0 30.0 1.1 30.0 } }

pcm.!default { type asym playback.pcm "plug:softvol" capture.pcm "plug:micboost" }

ctl.!default { type hw card 0 }

[shivasiddharth]

Paste the contents of asoundrc into asound.conf. Type the follwoing: sudo nano /etc/asound.conf paste the following into that and save:

options snd_rpi_googlemihat_soundcard index=0

pcm.softvol {
type softvol
slave.pcm dmix
control {
name Master
card 0
}
}

pcm.micboost {
type route
slave.pcm dsnoop
ttable {
0.0 30.0
1.1 30.0
}
}

pcm.!default {
type asym
playback.pcm "plug:softvol"
capture.pcm "plug:micboost"
}

ctl.!default {
type hw
card 0
}

this should fix your audio issues and snowboy should run now.

[vadalp]

Thanks for your patience. Finally it worked!!!

may be you need to tweak the script to copy the asound.conf file to /etc directory.

[shivasiddharth]

its already done.. I have intimated the AIY guys also about the .asoundrc issue (16000hz, 48000Hz). https://github.com/google/aiyprojects-raspbian/pull/172