The Playa works with WebAudio leveraging on Howler using a technique called Audio Sprites, this ensure that the audio can be reproduced on devices that can't reproduce more than one file simultaneously. This technique is similar to its Computer Graphic counterpart, a single file contains multiple tracks and metadata describe when a track begins and its lenght.
Clone or download this repository.
The software is intentionally kept as simple as possible.
Open a Terminal, navigate to project folder, launch a webserver and visit it from a browser.
For example from a Mac is possible to launch a webserver with this command:
python -m SimpleHTTPServer 8080
Then visit http://localhost:8080 from your browser.
To prepare an audio track that can be reproduced as sprite is necessary some assistance, the project contains a script (only Mac and Linux) that use Audiosprite to compose tracks in a single file and saves its metadata. Audiosprite uses ffmpeg and convert the audio track in multiple formats to ensure compatibility across many devices and browsers.
Before using the script is necessary to install Audiosprite, Audiosprite depends on ffmpeg and is written in Node.js, to install ffmpeg is strongly suggested to use Brew, to install AudioSprite will be necessary to install Node.js, is possible to install Node.js using Brew.
To install Brew open a Terminal and execute
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install.sh)"
On Mac open a terminal and execute these commands:
brew update
brew install node@10
On Mac open a terminal and execute these commands:
brew install ffmpeg
npm install -g audiosprite
Now you'll be able to run convertAudio.sh and convert folders and folders of samples into sprites.
./convertAudio.sh ./path/to/your/audio/files/ theNameOfYourOutput
The script will output directly in audio/ folder some sound files in different formats, and a json file.
The conversion script will recursively traverse all files and folders in the passed path, this permits to keep audio sprites samples well organized, for example my folder is organized this way:
SAMPLESWAP
|-BASS_LOOPS
| |-sample1_BASS_.wav
| |-sample2_BASS_.aif
| |-...
|-DRUM_LOOPS_and_BREAKS
| |-sample1_DRUM_.aif
| |-sample2_DRUM_.wav
| |-...
When I convert files I execute this command:
./convertAudio.sh ./SAMPLESWAP the-playa-demo
The player does not know your folder structure, to keep audio samples organized in multitrack is necessary to define tags in config.json file and use those tags to name your samples, this example uses _BASS_ and _DRUM_, check config.json and samples in SAMPLESWAP to understand better what's going on.
Is possible to add more tags to config.json in order to add new tracks to multitrack player The Playa.