litexa

# Appendix: WAV Audio Conversion

The @litexa/assets-wav module performs automatic conversion of your WAV asset files to Alexa-compatible MP3 files upon deployment.

Alexa skills require all audio asset files to follow this format (opens new window). If your skill will involve heavy or continuous audio production and assets, the @litexa/assets-wav extension module may be worth installing in your Litexa project.

# Installation

You will first need to install node gyp (opens new window). This supports native add-on modules for Node.js, which is required because @litexa/assets-wav depends on the lame (opens new window) module.

npm install -g node-gyp

Additionally, you will need to install build tools to compile the 'lame' dependency:

DANGER

As of July 2019, there is a known issue (opens new window) with installing lame on Node version 12+. To use this extension, we recommend using an LTS version of Node (either 8 or 10).

  • Windows:
npm install -g windows-build-tools
  • Linux:
# apt-get (Ubuntu/Debian)
sudo apt-get install build-essential

# yum (RHEL/Amazon Linux)
sudo yum groupinstall "Development Tools"
  • Mac OS:

Install Xcode Command Line Tools (opens new window)

Next, the extension can be installed locally with the following command:

npm install @litexa/assets-wav

Global installation

You can choose to install the extension globally instead, by running:

npm install -g @litexa/assets-wav

Doing so, however, will automatically convert all WAV files to MP3 files in all Litexa projects upon deployment. We do not recommend globally installing this conversion if you plan to use other asset converters in your projects because it may produce conflicting or indeterminate output.

If you want to track the extension as a dependency, just run the following inside of your Litexa project's root directory:

npm install --save-dev @litexa/assets-wav

This should result in the following directory structure:

project_dir
├── litexa
└── node_modules
    └── @litexa
        └── assets-wav

# Usage

# Automatic conversion (default)

Put your WAV files in the assets folder, in the same way you would with all other skill assets.

In your Litexa code, refer to the above sound files as if the MP3-converted version existed in your skill. For example, if you had introMusic.wav and meow.wav:

launch
  soundEffect introMusic.mp3
  say "Welcome to the Ultimate Cat Show. <sfx meow.mp3>"

# Manual conversion (optional)

For debugging purposes, you can perform an explicit conversion of a WAV file to the Alexa-supported MP3 by using the conversion script convert.coffee located in the package directly.

Using the script requires you to have CoffeeScript (opens new window) installed. You can install CoffeeScript with the command:

npm install -g coffeescript

Then, you can use its command line to execute the script. For example:

coffee convert.coffee meow.wav ./testAssets/meow.mp3

The first argument is the path to the WAV file you want to convert. The second argument is optional and specifies a destination path for the converted MP3 file. If omitted, it defaults to the same location as the WAV file path.

Only use manual conversion for debugging

If you have a WAV file and MP3 file of the same name in your assets directory, this extension will halt deployment because it would produce a duplicate MP3 file. We recommend only using the script to test what your audio files would sound like after the conversion, and let the extension handle the conversion automatically during skill deployment.

# What happens on deployment

Upon deployment, Litexa will automatically convert any WAV files in the assets folder and locally store the conversion in .deploy/converted-assets. It will also create a hash file alongside the converted file for caching purposes - it does not convert files when they haven't changed since the last deployment.

If you are using the @litexa/deploy-aws module for deployment, it will deploy them to your assets S3 bucket, following the same steps of non-converted assets. See the Deployment section for details.

← Language Reference →