[WIP]

Installation (WIP)

On mac:

git clone https://github.com/medialab/murmures.git
cd murmures

# dependencies
brew install sox # audio recording
brew install git-lfs # git large files
git lfs install

# install whisper
mkdir -p models/Xenova
git clone https://huggingface.co/Xenova/whisper-small models/Xenova/whisper-small
git clone https://huggingface.co/Xenova/whisper-small models/Xenova/whisper-base
git clone https://huggingface.co/Xenova/whisper-small models/Xenova/whisper-tiny

npm cache verify
export npm_config_arch=arm64
export npm_config_target_arch=arm64
export npm_config_build_from_source=true
npm install

On Linux (including WSL) platform:

# clone repo
git clone https://github.com/medialab/murmures.git
cd murmures

# install dependencies
sudo apt-get install git-lfs # git large file system
# install whisper
mkdir -p models/Xenova
git clone https://huggingface.co/Xenova/whisper-small models/Xenova/whisper-small
git clone https://huggingface.co/Xenova/whisper-tiny models/Xenova/whisper-tiny
git clone https://huggingface.co/Xenova/whisper-base models/Xenova/whisper-base

# install nvm – comment if you already have it
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.5/install.sh | bash
nvm install 20
nvm use 20
# keyboard
sudo apt-get install -y libxkbcommon-x11-0
sudo apt install -y evtest
# cups related
sudo apt install build-essential python3 make g++ pkg-config cups libcups2-dev
sudo apt install cups
sudo cp /etc/cups/cupsd.conf /etc/cups/cupsd.conf.original
sudo chmod a-w /etc/cups/cupsd.conf.original

# audio recording
sudo apt install -y alsa-utils sox libsox-fmt-all

# install dependencies
npm i
# on debian rebuild printer
npm rebuild printer --build-from-source

After installation

Find the name of the printer you are going to use. On Debian:

lpstat -p -d

Copy config file:

cp config.sample.json config.json

Then in config.json modify printerName accordingly. Adjust other settings as you which

On debian/raspberry PI, you may also have to set your keyboard capture mode to X11:

sudo raspi-config

Then:

Advanced Options
→ Wayland
→ X11
→ Reboot

Then when running echo $XDG_SESSION_TYPE you should get x11.

You should also on debian identify the proper audio output:

aplay -l

Then modify config.json audioCard by setting the proper number. You can try different cards with the script tests/play.js.

Same thing for the recorder:

arecord -l

Then modify config.json recordCard by setting the proper number. You can try different cards with the script tests/record.js.

If sound is too low:

alsamixer

Then F6 > select your output (eg USB audio) > augment the sound > esc.

Then:

sudo alsactl store

Usage

npm start # default script, starts the app
npm html # generate html paginated document of public stories
npm server # serve webserver with html paginated document of public stories

Setup services for launching at startup on a raspberry pi

Paste the following line in your terminal (replace replace_with_your_username with your linux username):

nano /home/replace_with_your_username/start-murmures.sh

Then paste start-murmures.sh file content into it (replace replace_with_your_username with your username – whoami if you're not sure).

Then give the script file execution rights:

chmod +x /home/replace_with_your_username/start-murmures.sh

Then create an autostart script:

mkdir -p /home/replace_with_your_username/.config/autostart
nano /home/replace_with_your_username/.config/autostart/murmures.desktop

Then paste the content of autostart-murmures.desktop (replace replace_with_your_username with your linux username).

Then give the autostart script execution rights:

chmod +x /home/replace_with_your_username/.config/autostart/murmures.desktop

You're done!