MMM-AssistantMk2

A Former User

@cr4z33 Details are dependent with also your asound config. But,
Can you record and play some audio in Terminal?

arecord -t 5 test.wav & aplay test.wav

Can you execute this? tell me the result.

(I haven’t had voiceHAT, so cannot guaranteed, but same name could be available.)

Cr4z33

@sean
I get an error and please note that I’ve tried to translate to English. :winking_face:

arecord -t 5 test.wav & aplay test.wav
[1] 2911
arecord: main:574: unrecognized file format 5
test.wav: File or directory does not exist
[1]+  Output 1                arecord -t 5 test.wav

A Former User

@cr4z33
Try this.

arecord -t 5 -D hw:0,0 test.wav

Does this make test.wav?
If so, try this again.

aplay -D hw:0,0 test.wav

Cr4z33

@sean
Nope. :pensive_face:

arecord -t 5 -D hw:0,0 test.wav
arecord: main:574: unrecognized file format 5

A Former User

@cr4z33

cat ~/.asoundrc

Tell me the result.

How you can setup previously? (I think you purchased that device for Google AIY project. Was there any instructions?)

A Former User

@cr4z33
Oops ! my mistake!

arecord -d 5 -D hw:0,0 test.wav

Cr4z33

@sean said in MMM-AssistantMk2 & MMM-GAction:

@cr4z33
cat ~/.asoundrc
Tell me the result.

How you can setup previously? (I think you purchased that device for Google AIY project. Was there any instructions?)

pi@raspberrypi:~ $ cat ~/.asoundrc
cat: /home/pi/.asoundrc: File or directory does not exist

It’s the so- called Google Voice Kit AIY (https://aiyprojects.withgoogle.com/voice-v1/).

I used a ready meal Raspbian image made by Google.
Apart from creating the JSON file everything else consists of ready scripts (i.e assistant_library_demo.py is a script to runs and uses Google Assistant):
https://github.com/google/aiyprojects-raspbian/tree/voicekit

A Former User

@cr4z33
Hmmm… As I know,

arecord -d 5 test.wav
aplay test.wav

This should work because there is no other suspicious things. your card and device no is 0,0 and that is the default.

Cr4z33

@sean
Again I made a noob error… I ran both commands at once lol! :smiling_face_with_open_mouth_cold_sweat:

Running them separately worked like a charm! :thumbs_up_medium-light_skin_tone:

I am going now to retry with the module addition.

Cr4z33

@sean
So after following the installation instructions I ran a test by using the ‘Very simple version but limited’ module.

As a result all I get now from MM is a black empty screen (no config.js message and no trace of other modules can be seen).

sgopala

@cr4z33 There is a small typo in the config file, please correct as under:

Incorrect: [replace this]
play: { // Full values are in FOR EXPERTS section.
playProgram: “mpg321”, // recommended.
}

Correct: [with this]
play: { // Full values are in FOR EXPERTS section.
playProgram: “mpg321”, // recommended.
},

A Former User

@sgopala Thank you!! I’ve fixed. I think I had assumed that was the last field of ESSENTIALS parts, so I might have missed the last comma.

sgopala

@sean Thanks a ton for your modules sir, you are a rockstar. But somehow I liked the previous version because of the following reasons:

(i) there is a delay after the hotword detection in the new version
(ii) the hotword detection hangs after 1st interaction with the assistant
(iii) the output window (iframe) was better sized and placed, too small on this version

Request to please see if these are feasible.

Once again, thanks a lot for this lovely module.

A Former User

@sgopala
You can always checkout the older version from github (use 1.x.x-final-snap)
And for the issues;

some delay is added intentionally to avoid possible collision. I’ll modify it could be customizable in next patch.
output size could be controllable with screenZoom field.

Thanks.

Cr4z33

@sgopala Thanks for pointing this out. ;)

@sean I keep getting either a black MM screen or the ‘config.js’ message.

At this point I prefer to work with a config with MMM-AssistantMk2 only and no other modules until everything is working 100%.

If you could help me by giving please a look at the below config? :)
Please note that I am trying to use

0:0

as device Id.

https://pastebin.com/ePfrtR75

A Former User

@cr4z33
Try this.

{
  module: "MMM-AssistantMk2",
  position: "top_right",
  config: {
    record: {
      recordProgram: "arecord",
      device: null        // or device: "hw:0,0"
    },

    play: {
      playProgram: "mpg321",
    },
  }
},

citizendevpi

@sean sweet i added a bunch of outher modules and now get a memory error how do i update to the newer version so hopefully these errors sort themselvs out

citizendevpi

!!! { audioFormat: 1,
endianness: ‘LE’,
channels: 1,
sampleRate: 16000,
byteRate: 32000,
blockAlign: 2,
bitDepth: 16,
signed: true }
mmap() failed: Cannot allocate memory
Failed to create permanent mapping for memfd region with ID = 3246537711
Ignoring received block reference with non-registered memfd ID = 3246537711
ALSA lib pulse.c:243:(pulse_connect) PulseAudio: Unable to connect: Protocol error

[…/deps/mpg123/src/output/alsa.c:165] error: cannot open device default
Error: open() failed: -1
at Speaker._open (/home/pi/MagicMirror/modules/MMM-AssistantMk2/node_modules/speaker/index.js:106:13)
at Speaker._write (/home/pi/MagicMirror/modules/MMM-AssistantMk2/node_modules/speaker/index.js:183:23)
at doWrite (/home/pi/MagicMirror/modules/MMM-AssistantMk2/node_modules/readable-stream/lib/_stream_writable.js:428:64)
at writeOrBuffer (/home/pi/MagicMirror/modules/MMM-AssistantMk2/node_modules/readable-stream/lib/_stream_writable.js:417:5)
at Speaker.Writable.write (/home/pi/MagicMirror/modules/MMM-AssistantMk2/node_modules/readable-stream/lib/_stream_writable.js:334:11)
at Reader. (/home/pi/MagicMirror/modules/MMM-AssistantMk2/node_helper.js:287:17)
at emitNone (events.js:111:20)
at Reader.emit (events.js:208:7)
at /home/pi/MagicMirror/modules/MMM-AssistantMk2/node_modules/wav/node_modules/readable-stream/lib/_stream_readable.js:934:16
at _combinedTickCallback (internal/process/next_tick.js:131:7)
Some error happens. Try again.
Whoops! There was an uncaught exception…
TypeError: this.sendSocketNotification is not a function
at Reader. (/home/pi/MagicMirror/modules/MMM-AssistantMk2/node_helper.js:296:20)
at emitNone (events.js:111:20)
at Reader.emit (events.js:208:7)
at /home/pi/MagicMirror/modules/MMM-AssistantMk2/node_modules/wav/node_modules/readable-stream/lib/_stream_readable.js:934:16
at _combinedTickCallback (internal/process/next_tick.js:131:7)
at process._tickCallback (internal/process/next_tick.js:180:9)

ya looks like speaker error

Cr4z33

@Sean
If I use null I get a black MMM screen.

If I use hw:0,0 I get the config.js message.

Finally if I use "hw:0,0" (with quotes) I get a black MM screen again.

:smiling_face_with_open_mouth_cold_sweat:

A Former User

@cr4z33
Ok… Are you using Telegram?
I can give more direct help on Telegram. My id is @eouia.
Here is not so good place to talk these things.