bbc/audiosprite

Name: audiosprite

Owner: BBC

Description: Modified version of audiosprite

Created: 2016-08-19 15:31:30.0

Updated: 2016-09-14 09:45:58.0

Pushed: 2016-09-23 11:33:58.0

Homepage:

Size: 108

Language: JavaScript

GitHub Committers

User	Most Recent Commit	# Commits

Other Committers

User	Email	Most Recent Commit	# Commits

README

What?

This is a ffmpeg wrapper that will take in multiple audio files and combines them into a single file. Silent gaps will be put between the parts so that every new part starts from full second and there is at least 1 second pause between every part. The final file will be exported in mp3, ogg, ac3, m4a and caf(IMA-ADPCM) to support as many devices as possible. This tool will also generate a JSON file that is compatible with Howler.js or zynga/jukebox framework.

Why?

iOS, Windows Phone and some Android phones have very limited HTML5 audio support. They only support playing single file at a time and loading in new files requires user interaction and has a big latency. To overcome this there is a technique to combine all audio into single file and only play/loop certain parts of that file. zynga/jukebox is a audio framework that uses this technique. digitalfruit/limejs is a HTML5 game framework that includes Jukebox and lets you add audio to your games using audio sprites.

Installation

install -g audiosprite

You can install FFmpeg and the ogg codecs on OSX using brew:

 install ffmpeg --with-theora --with-libogg --with-libvorbis

Hints for Windows users

• You need to install Node.js
• Use Git Bash instead of Command Line or Powershell
• Download ffmpeg and include it in your path export PATH=$PATH:path/to/ffmpeg/bin
• IMA-ADPCM(the fastest iPhone format) will only be generated if you are using OSX.

Usage

diosprite --help
: Usage: audiosprite [options] file1.mp3 file2.mp3 *.wav
: Options:
output, -o      Name for the output files.                                   [default: "output"]
path, -u        Path for files to be used on final JSON.                     [default: ""]
export, -e      Limit exported file types. Comma separated extension list.   [default: "ogg,m4a,mp3,ac3"]
format, -f      Format of the output JSON file (jukebox, howler, createjs).  [default: "jukebox"]
log, -l         Log level (debug, info, notice, warning, error).             [default: "info"]
autoplay, -a    Autoplay sprite name.                                        [default: null]
loop            Loop sprite name, can be passed multiple times.              [default: null]
silence, -s     Add special "silence" track with specified duration.         [default: 0]
gap, -g         Silence gap between sounds (in seconds).                     [default: 1]
minlength, -m   Minimum sound duration (in seconds).                         [default: 0]
bitrate, -b     Bit rate. Works for: ac3, mp3, mp4, m4a, ogg.                [default: 128]
vbr, -v         VBR [0-9]. Works for: mp3. -1 disables VBR.                  [default: -1]
samplerate, -r  Sample rate.                                                 [default: 44100]
channels, -c    Number of channels (1=mono, 2=stereo).                       [default: 1]
rawparts, -p    Include raw slices(for Web Audio API) in specified formats.  [default: ""]
help, -h        Show this help message.


diosprite --autoplay bg_loop --output mygameaudio bg_loop.wav *.mp3
: File added OK file=bg_loop.wav
: 1.25s silence gap added OK
: File added OK file=click.mp3
: 1.70s silence gap added OK
: Exported caf OK file=mygameaudio.caf
: Exported ac3 OK file=mygameaudio.ac3
: Exported mp3 OK file=mygameaudio.mp3
: Exported m4a OK file=mygameaudio.m4a
: Exported ogg OK file=mygameaudio.ogg
: Exported json OK file=mygameaudio.json
: All done


t mygameaudio.json

esources": [
"mygameaudio.caf",
"mygameaudio.ac3",
"mygameaudio.mp3",
"mygameaudio.m4a",
"mygameaudio.ogg"

pritemap": {
"bg_loop": {
  "start": 0,
  "end": 3.75,
  "loop": true
},
"click": {
  "start": 5,
  "end": 5.3,
  "loop": false
}

utoplay": "bg_loop"

Setting autoplay track

You can use --autoplay option to set a track that will start playing automatically. This track is then marked as autoplay and looping in the JSON. This syntax is Jukebox framework specific.

Custom silent track

On some cases starting and pausing a file has bigger latency than just setting playhead position. You may get better results if your file is always playing. --silence <duration> option will generate extra track named silence that you can play instead of pausing the file.

Usage with zynga/jukebox framework.

Generated JSON file can be passed straight into jukebox.Player constructor. Check out Jukebox documentation/demos for more info.

settings = {/* JSON generated by audiosprite*/};

his part needs to be in user event callback.
myPlayer = new jukebox.Player(settings);

ayer.play('click');

Usage with digitalfruit/limejs framework.

First generate LimeJS asset from the JSON file that you can require inside your code.

n/lime.py gensoy path/to/mygameaudio.json

Then use AudioMap class to play the file.

.require('lime.audio.AudioMap');
.require('lime.ASSETS.mygameaudio.json');

audio = new lime.audio.AudioMap(lime.ASSETS.mygameaudio.json);

o.play('click');

Don't forget to use the --rawparts=mp3 option to benefit from the LimeJS feature to automatically switch to Web Audio API when it's supported by the client.