Hello-Lucy using Adafruit Voice Bonnet

sinbad339

I’m trying to bring up Hello-Lucy on a “new” Magic Mirror. I’m running on a Pi 3B+, and installed the Adafruit Voice Bonnet for audio.

I can do
arecord --format cd --duration 7 --channels 1 test2.wav
then
aplay test2.wav
and I hear the recording from the Voice Bonnet speaker. I don’t have to specify he device for these to work.

When I list my audio devices, I get:
pi@Magic:~/MagicMirror/config $ aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: Headphones [bcm2835 Headphones], device 0: bcm2835 Headphones [bcm2835 Headphones]
Subdevices: 8/8
Subdevice #0: subdevice #0
Subdevice #1: subdevice #1
Subdevice #2: subdevice #2
Subdevice #3: subdevice #3
Subdevice #4: subdevice #4
Subdevice #5: subdevice #5
Subdevice #6: subdevice #6
Subdevice #7: subdevice #7
card 1: vc4hdmi [vc4-hdmi], device 0: MAI PCM i2s-hifi-0 [MAI PCM i2s-hifi-0]
Subdevices: 1/1
Subdevice #0: subdevice #0
card 2: seeed2micvoicec [seeed-2mic-voicecard], device 0: bcm2835-i2s-wm8960-hifi wm8960-hifi-0 [bcm2835-i2s-wm8960-hifi wm8960-hifi-0]
Subdevices: 1/1
Subdevice #0: subdevice #0
pi@Magic:~/MagicMirror/config $ arecord -l
**** List of CAPTURE Hardware Devices ****
card 2: seeed2micvoicec [seeed-2mic-voicecard], device 0: bcm2835-i2s-wm8960-hifi wm8960-hifi-0 [bcm2835-i2s-wm8960-hifi wm8960-hifi-0]
Subdevices: 1/1
Subdevice #0: subdevice #0

Card 2 for both play and capture is the Voice Bonnet.

When I restart MMM with Hello-Lucy, I hear the default greeting, but always get the microphone shown with ERROR. Even though the lists indicate card 2, I’ve tried 0 and 1 since card 2 is the only thing listed, but still no microphone.

Here’s my config.js file:
/* Magic Mirror Config Sample
*

By Michael Teeuw https://michaelteeuw.nl
MIT Licensed.

let config = {
address: “localhost”, // Address to listen on, can be:
// - “localhost”, “127.0.0.1”, “::1” to listen on loopback interface
// - another specific IPv4/6 to listen on a specific interface
// - “0.0.0.0”, “::” to listen on any interface
// Default, when address config is left out or empty, is “localhost”
port: 8081,
basePath: “/”, // The URL path where MagicMirror is hosted. If you are using a Reverse proxy
// you must set the sub path here. basePath must end with a /
ipWhitelist: [“127.0.0.1”, “::ffff:127.0.0.1”, “::1”], // Set [] to allow all IP addresses
// or add a specific IPv4 of 192.168.1.5 :
// [“127.0.0.1”, “::ffff:127.0.0.1”, “::1”, “::ffff:192.168.1.5”],
// or IPv4 range of 192.168.3.0 --> 192.168.3.15 use CIDR format :
// [“127.0.0.1”, “::ffff:127.0.0.1”, “::1”, “::ffff:192.168.3.0/28”],

    useHttps: false,                // Support HTTPS or not, default "false" will use HTTP
    httpsPrivateKey: "",    // HTTPS private key path, only require when useHttps is true
    httpsCertificate: "",   // HTTPS Certificate path, only require when useHttps is true

    language: "en",
    locale: "en-US",
    logLevel: ["INFO", "LOG", "WARN", "ERROR"], // Add "DEBUG" for even more logging
    timeFormat: 24,
    units: "imperial",
    // serverOnly:  true/false/"local" ,
    // local for armv6l processors, default
    //   starts serveronly and then starts chrome browser
    // false, default for all NON-armv6l devices
    // true, force serveronly mode, because you want to.. no UI on this device

    modules: [
            {
            disabled: false,
            module: "Hello-Lucy",
            position: "top_center",
            config: {
                keyword: 'HELLO LUCY',              // keyword to activate listening for a command/sentence
                timeout: 15,                        // timeout listening for a command/sentence
                standByMethod: 'DPMS',              // 'DPMS' = anything else than RPi or 'PI'
                microphone: "1,0",                  // run "arecord -l" card # and device # mine is "0,0"
                sounds: ["1.mp3", "11.mp3"],        // welcome sound at startup. Add several for a random greetings
                confirmationSound: "ding.mp3",      // name and extension of sound file
                startHideAll: true,                 // All modules start as hidden EXCEPT PAGE ONE
                // *** Page One is your default startup page *** This overrides startHideAll: true,
                pageOneModules: ["Hello-Lucy","MMM-EasyPix"],                     // default modules to show on page one/startup
                pageTwoModules: ["Hello-Lucy", "MMM-BMW-DS", "MMM-EventHorizon"], // modules to show on page two
                pageThreeModules: ["Hello-Lucy", "MMM-Lunartic"],                 // modules to show on page three
                pageFourModules: ["Hello-Lucy", "MMM-PC-Stats"],                  // modules to show on page four
                pageFiveModules: ["Hello-Lucy", "MMM-Searchlight"],               // modules to show on page five
                pageSixModules: ["Hello-Lucy", "MMM-NOAA3"],                      // modules to show on page six
                pageSevenModules: ["Hello-Lucy", "MMM-Recipe"],                   // modules to show on page seven
                pageEightModules: ["Hello-Lucy", "MMM-rfacts"],                   // modules to show on page eight
                pageNineModules: ["Hello-Lucy", "MMM-History"],                   // modules to show on page nine
                pageTenModules: ["Hello-Lucy", "MMM-HardwareMonitor"]             // modules to show on page ten
                }
            },
    ]

};

/*************** DO NOT EDIT THE LINE BELOW ***************/
if (typeof module !== “undefined”) {module.exports = config;}

Any help would be appreciated.

sinbad339

Well, I gave up on the Adafruit Voice Bonnet, and instead installed a USB microphone with output (Samson GoMic). Now, mic and speakers work every time whether MM/Hello-Lucy starts up automatically at boot, or if I start it using npm start.

I’ll probably mark it resolved, even thought the fundamental issue is not.

A BIG thanks to @sdetweil for all of his patience and help.

sdetweil

@sinbad339 said in Hello-Lucy using Adafruit Voice Bonnet:

  microphone: "1,0",

so arecord says card 2,0

 arecord -l
**** List of CAPTURE Hardware Devices ****
card 2: seeed2micvoicec [seeed-2mic-voicecard], device 0: bcm2835-i2s-wm8960-hifi wm8960-hifi-0 [bcm2835-i2s-wm8960-hifi wm8960-hifi-0]
Subdevices: 1/1
Subdevice #0: subdevice #0

so it should be

 microphone: "2,0",

sinbad339

@sdetweil said in Hello-Lucy using Adafruit Voice Bonnet:

so it should be
microphone: “2,0”,

Thanks for replying. This config is what I had when I posted. As I mentioned, I’ve tried “0,0”, “1,0”, and “2,0”; all result in the microphone error.

sdetweil

@sinbad339 can u try this

arecord -D default -q  -r 16000 -c 1 -t wav -f S16_LE >test.wav

the replace default with

arecord -D hw:2,0  -q -r 16000 -c 1 -t wav -f S16_LE >test.wav

sinbad339

@sdetweil
With MM running, if I issue

arecord -D default -q  -r 16000 -c 1 -t wav -f S16_LE >test.wav

I get a good .wave file that I can play.
With MM still running, specifying the hw explicitly results in an error.

arecord -D hw:2,0  -q -r 16000 -c 1 -t wav -f S16_LE >test2.wav
arecord: main:830: audio open error: Device or resource busy
pi@Magic:~/MagicMirror/config $ arecord -l
**** List of CAPTURE Hardware Devices ****
card 2: seeed2micvoicec [seeed-2mic-voicecard], device 0: bcm2835-i2s-wm8960-hifi wm8960-hifi-0 [bcm2835-i2s-wm8960-hifi wm8960-hifi-0]
  Subdevices: 0/1
  Subdevice #0: subdevice #0

sdetweil

@sinbad339 said in Hello-Lucy using Adafruit Voice Bonnet:

With MM still running, specifying the hw explicitly results in an error.

ok, so MM is using that device

while mm is running do

ps -ef | grep arec

u should see the arecord process and the parms passed
from one of my apps that uses arecord

arecord -D default -q -r 16000 -c 1 -t wav -f S16_LE -

the trailing dash means stream content to stdout (which is captured by the upstream library)

sdetweil

@sinbad339

arecord -D default

the question is, what is ‘default’

sinbad339

@sdetweil
So the arecord -l lists card2 as my expected capture device (seeed-2mic-voicecard). I find it interesting that it enumverates as #2 with no #0 or #1.

Anyway, I ran your ps command; not sure what this output means (as I’m sure you’re aware by now, I’m a Noob on Linux).

ps -ef | grep arec
pi        1953  1927  0 11:04 pts/0    00:00:00 grep --color=auto arec

sdetweil

@sinbad339 hm…

should have used arecord…

try

ps -ef | grep rec

(there are only 2 apps… arecord or rec )

the output from ps

pi        1953  1927  0 11:04 pts/0    00:00:00 grep --color=auto arec

user that started process
the process id
the parent process id +
(i don’t know, 0)
the time the process was started
(i don’t know, pts/ )
how much cpu time has been used since started, in seconds
the string used to start the process, pgm name and then parms

sinbad339

@sdetweil

ps -ef | grep rec
root        76     2  0 11:02 ?        00:00:00 [vchiq-recy/0]
pi        2320  1927  0 11:24 pts/0    00:00:00 grep --color=auto rec

Not sure what vchiq-recy is…
But no arec or arecord.

sdetweil

@sinbad339 said in Hello-Lucy using Adafruit Voice Bonnet:

But no arec or arecord.

well, that explains the error…

you did run the installation dependencies script, right?

cd Hello-Lucy/installers
bash dependencies.sh

sinbad339

@sdetweil
Yes, I did run it. Its been a while, and I remember warning about depreciated versions, but nothing that I thought could be an issue. OK to run it again?

sdetweil

@sinbad339 said in Hello-Lucy using Adafruit Voice Bonnet:

OK to run it again?

i don’t think there is a need

add debug:true to the module config section

module: 
    config:  {
        debug:true,
   .... whatever else was there
    }

and then

pm2 stop all 
cd ~/MagicMirror
npm start >somefile.txt

when error reported

alt-spacebar, n to minimize mm
ctrl-c to stop mm
and look at somefile.txt

sinbad339

@sdetweil
WELL, now it seems to work. Running from command line w/ debug, I don’t get microphone with ERROR, and if I say “Hello Lucy”, it prompts for a command, then does it, and issues the “ding”. I’ve done this three times now, and its worked every time. Here’s the debug outout:

> magicmirror@2.18.0 start /home/pi/MagicMirror
> DISPLAY="${DISPLAY:=:0}" ./node_modules/.bin/electron js/electron.js

[05.01.2022 11:55.24.384] [LOG]   Starting MagicMirror: v2.18.0
[05.01.2022 11:55.24.394] [LOG]   Loading config ...
[05.01.2022 11:55.24.404] [LOG]   Loading module helpers ...
[05.01.2022 11:55.26.860] [LOG]   Initializing new module helper ...
[05.01.2022 11:55.26.864] [LOG]   Module helper loaded: Hello-Lucy
[05.01.2022 11:55.26.866] [LOG]   All module helpers loaded.
[05.01.2022 11:55.27.040] [LOG]   Starting server on port 8081 ...
[05.01.2022 11:55.27.066] [LOG]   Server started ...
[05.01.2022 11:55.27.069] [LOG]   Connecting socket for: Hello-Lucy
[05.01.2022 11:55.27.071] [LOG]   Starting module helper: Hello-Lucy
[05.01.2022 11:55.27.073] [LOG]   Sockets connected & modules started ...
[05.01.2022 11:55.27.482] [LOG]   Launching application.
[05.01.2022 11:55.37.133] [LOG]   Hello-Lucy has recognized: ALL IS THE
[05.01.2022 11:56.16.829] [LOG]   Hello-Lucy has recognized: NINE

sinbad339

The difference is how MM is started. If I do it the manual way, the microphone works (with or without debug enabled). If MM does its normal autostart, using mm.sh that includes DISPLAY=:0 npm start, then the microphone doesn’t work.

sdetweil

@sinbad339 said in Hello-Lucy using Adafruit Voice Bonnet:

its normal autostart,

how is that done? pm2 or did u do some other setup

sounds like mm is not running as user but as root?

sinbad339

@sdetweil
I followed the instructions from https://www.magicmirrorcentral.com/complete-raspberry-pi-magic-mirror-tutorial/.
It included

type in the following command:

pm2 startup
 
A different command will appear. Run that command also.

Now create a shell script for startup. Open up the built-in text editor on the Pi by typing in:

cd ~
nano mm.sh
The lower line basically means, “open up a file called mm.sh in the nano text editor.”

When the text editor has opened, add this to the mm.sh file:

cd ~/MagicMirror
DISPLAY=:0 npm start
Make your shiny new script executable by typing in:

chmod +x mm.sh
Now start up Magic Mirror and make it start up on boot:

pm2 start mm.sh

pm2 save
Now, reboot your Pi by typing in

sudo reboot

sdetweil

@sinbad339 hm… my install script does all that.

mm.sh already exists in the installers folder.

try my fixuppm2.sh from here
https://github.com/sdetweil/MagicMirror_scripts

sinbad339

@sdetweil
Let me start by saying I really appreciate your help with this. You’ve given me a lot of your time, and clearly I couldn’t have made any progress without you. Thank you.

Running your fixupppm2 script fixes the microphone on a restart, but then there is no speaker output. If I stop MM, then manually restart it with npm start, the mic and speaker work fine, as before the fixup.

sdetweil

@sinbad339 that makes no sense to me.
could u start it w pm2, and then ps -ef >somefile.txt

and email me thevfile some file.txt

same userid as here at gmail.

I want to see if there is some permissions problem

oh, check the file contents to make sure no userid or password is exposed. shouldn’t be but u never know.