recordrtc

RecordRTC: WebRTC audio/video recording

RecordRTC Documentation / RecordRTC Wiki Pages / RecordRTC Demo / WebRTC Experiments

RecordRTC is a JavaScript-based media-recording library for modern web-browsers (supporting WebRTC getUserMedia API). It is optimized for different devices and browsers to bring all client-side (pluginfree) recording solutions in single place.

Please check dev directory for development files.

1. RecordRTC API Reference
2. MRecordRTC API Reference
3. MediaStreamRecorder API Reference
4. StereoAudioRecorder API Reference
5. WhammyRecorder API Reference
6. Whammy API Reference
7. CanvasRecorder API Reference
8. GifRecorder API Reference
9. Global API Reference

Browsers Support:

Browser	Support
Firefox	Stable / Aurora / Nightly
Google Chrome	Stable / Canary / Beta / Dev
Opera	Stable / NEXT
Android	Chrome / Firefox / Opera
Microsoft Edge	Normal Build

How RecordRTC encodes wav/webm?

Media File	Bitrate/Framerate	encoders	Framesize	additional info
Audio File (WAV)	1411 kbps	pcm_s16le	44100 Hz	stereo, s16
Video File (WebM)	60 kb/s	(whammy) vp8 codec yuv420p	--	SAR 1:1 DAR 4:3, 1k tbr, 1k tbn, 1k tbc (default)

RecordRTC Demos

1. RecordRTC to Node.js
2. RecordRTC to PHP
3. RecordRTC to ASP.NET MVC
4. RecordRTC & HTML-2-Canvas i.e. Canvas/HTML Recording!
5. MRecordRTC i.e. Multi-RecordRTC!
6. RecordRTC on Ruby!
7. RecordRTC over Socket.io
8. ffmpeg-asm.js and RecordRTC! Audio/Video Merging & Transcoding!
9. RecordRTC / PHP / FFmpeg
10. Record Audio and upload to Nodejs server
11. ConcatenateBlobs.js - Concatenate multiple recordings in single Blob!
12. Remote stream recording

How to link?

npm install recordrtc

# you can use with "require" (browserify/nodejs)
var RecordRTC = require('recordrtc');
var recorder = RecordRTC(mediaStream, { type: 'audio'});

or using Bower:

bower install recordrtc

To use it:

<script src="./node_modules/recordrtc/RecordRTC.js"></script>

<!-- or -->
<script src="//cdn.WebRTC-Experiment.com/RecordRTC.js"></script>

<!-- or -->
<script src="//www.WebRTC-Experiment.com/RecordRTC.js"></script>

There are some other NPM packages regarding RecordRTC:

• https://www.npmjs.org/search?q=RecordRTC

How to capture stream?

<script src="https://cdn.rawgit.com/webrtc/adapter/master/adapter.js"></script>

<script>
function successCallback(stream) {
    // RecordRTC usage goes here
}

function errorCallback(errror) {
    // maybe another application is using the device
}

var mediaConstraints = { video: true, audio: true };

navigator.mediaDevices.getUserMedia(mediaConstraints).then(successCallback).catch(errorCallback);
</script>

Record audio+video in Firefox

You'll be recording both audio/video in single WebM container. Though you can edit RecordRTC.js to record in mp4.

var recordRTC;

function successCallback(stream) {
    // RecordRTC usage goes here

    recordRTC = RecordRTC(MediaStream);
    recordRTC.startRecording();
}

function errorCallback(errror) {
    // maybe another application is using the device
}

var mediaConstraints = { video: true, audio: true };

navigator.mediaDevices.getUserMedia(mediaConstraints).then(successCallback).catch(errorCallback);

btnStopRecording.onclick = function () {
    recordRTC.stopRecording(function (audioVideoWebMURL) {
        video.src = audioVideoWebMURL;
        
        var recordedBlob = recordRTC.getBlob();
        recordRTC.getDataURL(function(dataURL) { });
    });
};

Demo: AudioVideo-on-Firefox.html

Record only Audio

var recordRTC = RecordRTC(mediaStream);
recordRTC.startRecording();
recordRTC.stopRecording(function(audioURL) {
    audio.src = audioURL;
   
    var recordedBlob = recordRTC.getBlob();
    recordRTC.getDataURL(function(dataURL) { });
});

Echo Issues

Simply set volume=0 or muted=true over <audio> or <video> element:

videoElement.muted = true;
audioElement.muted = true;

Otherwise, you can pass some media constraints:

function successCallback(stream) {
    // RecordRTC usage goes here
}

function errorCallback(errror) {
    // maybe another application is using the device
}

var mediaConstraints = {
    audio: {
        mandatory: {
            echoCancellation: false,
            googAutoGainControl: false,
            googNoiseSuppression: false,
            googHighpassFilter: false
        },
        optional: [{
          googAudioMirroring: false
        }]
    },
};

navigator.mediaDevices.getUserMedia(mediaConstraints).then(successCallback).catch(errorCallback);

• Constraints Reference

Record Video

Everything is optional except type:'video':

var options = {
   type: 'video'
};
var recordRTC = RecordRTC(mediaStream, options);
recordRTC.startRecording();
recordRTC.stopRecording(function(videoURL) {
    video.src = videoURL;
   
    var recordedBlob = recordRTC.getBlob();
    recordRTC.getDataURL(function(dataURL) { });
});

onAudioProcessStarted

Useful to recover audio/video sync issues inside the browser:

recordAudio = RecordRTC( stream, {
     onAudioProcessStarted: function( ) {
         recordVideo.startRecording();
     }
});

recordVideo = RecordRTC(stream, {
    type: 'video'
});

recordAudio.startRecording();

onAudioProcessStarted fixes shared/exclusive audio gap (a little bit). Because shared audio sometimes causes 100ms delay... sometime about 400-to-500 ms delay. Delay depends upon number of applications concurrently requesting same audio devices and CPU/Memory available. Shared mode is the only mode currently available on 90% of windows systems especially on windows 7.

Record animated GIF image

Everything is optional except type:'gif':

// you must "manually" link:
// https://cdn.webrtc-experiment.com/gif-recorder.js

var options = {
   type: 'gif',
   frameRate: 200,
   quality: 10
};
var recordRTC = RecordRTC(mediaStream, options);
recordRTC.startRecording();
recordRTC.stopRecording(function(gifURL) {
   mediaElement.src = gifURL;
});

Record a Webpage

You can say it: "HTML/Canvas Recording using RecordRTC"!

<script src="//cdn.WebRTC-Experiment.com/RecordRTC.js"></script>
<script src="//cdn.webrtc-experiment.com/screenshot.js"></script>

<div id="elementToShare" style="width:100%;height:100%;background:green;"></div>
<script>
var elementToShare = document.getElementById('elementToShare');
var recordRTC = RecordRTC(elementToShare, {
    type: 'canvas'
});
recordRTC.startRecording();
recordRTC.stopRecording(function(videoURL) {
    video.src = videoURL;
   
    var recordedBlob = recordRTC.getBlob();
    recordRTC.getDataURL(function(dataURL) { });
});
</script>

See a demo: /Canvas-Recording/

API Reference

initRecorder

It is a function that can be used to initiate recorder however skip getting recording outputs. It will provide maximum accuracy in the outputs after using startRecording method. Here is how to use it:

var audioRecorder = RecordRTC(mediaStream, {
  type: 'audio',
  recorderType: StereoAudioRecorder
});

var videoRecorder = RecordRTC(mediaStream, {
  type: 'video',
  recorderType: WhammyRecorder
});

videoRecorder.initRecorder(function() {
  audioRecorder.initRecorder(function() {
    // Both recorders are ready to record things accurately
    videoRecorder.startRecording();
    audioRecorder.startRecording();
  });
});

After using stopRecording, you'll see that both WAV/WebM blobs are having following charachteristics:

1. Both are having same recording duration i.e. size
2. Video recorder is having no blank frames
3. Audio recorder is having no empty buffers

This method is really useful to sync audio/video outputs.

setRecordingDuration

You can ask RecordRTC to auto stop recording after specific duration. It accepts one mandatory and one optional argument:

recordRTC.setRecordingDuration(milliseconds, stoppedCallback);

// the easiest one:
recordRTC.setRecordingDuration(milliseconds).onRecordingStopped(stoppedCallback);

Try a simple demo; paste in the chrome console:

navigator.mediaDevices.getUserMedia({
    video: true
}).then(function(stream) {
    var recordRTC = RecordRTC(stream, {
        type: 'video',
        recorderType: WhammyRecorder
    });

    // auto stop recording after 5 seconds
    recordRTC.setRecordingDuration(5 * 1000).onRecordingStopped(function(url) {
        console.debug('setRecordingDuration', url);
        window.open(url);
    })

    recordRTC.startRecording();
}).catch(function(error) {
    console.error(error);
});

clearRecordedData

This method can be used to clear old recorded frames/buffers. Snippet:

recorder.clearRecordedData();

recorderType

You can force any Recorder by passing this object over RecordRTC constructor:

var audioRecorder = RecordRTC(mediaStream, {
  type: 'audio',
  recorderType: StereoAudioRecorder
})

It means that ALL_BROWSERS will be using StereoAudioRecorder i.e. WebAudio API for audio recording.

This feature brings remote audio recording support in Firefox, and local audio recording support in Microsoft Edge.

You can even force WhammyRecorder on Firefox however webp format isn't yet supported in standard Firefox builds. It simply means that, you're skipping MediaRecorder API in Firefox.

disableLogs

You can disable all the RecordRTC logs by passing this Boolean:

var recorder = RecordRTC(mediaStream, {
  disableLogs: true
});

audioChannels

You can force StereoAudioRecorder to record left-audio-channels only.

var audioRecorder = RecordRTC(audioStream, {
  type: 'audio',
  recorderType: StereoAudioRecorder,
  audioChannels: 1 // or leftChannel:true
});

*It will reduce WAV size to half!

This feature is useful only in Chrome and Microsoft Edge (WAV-recorders).

How to set video width/height?

var options = {
   type: 'video',
   video: {
      width: 320,
      height: 240
   },
   canvas: {
      width: 320,
      height: 240
   }
};

pauseRecording

recordRTC.pauseRecording();

resumeRecording

recordRTC.resumeRecording();

getDataURL

recordRTC.getDataURL(function(dataURL) {
   mediaElement.src = dataURL;
});

getBlob

blob = recordRTC.getBlob();

toURL

window.open( recordRTC.toURL() );

save

recordRTC.save();

bufferSize

Here is how to customize Buffer-Size for audio recording?

// From the spec: This value controls how frequently the audioprocess event is 
// dispatched and how many sample-frames need to be processed each call. 
// Lower values for buffer size will result in a lower (better) latency. 
// Higher values will be necessary to avoid audio breakup and glitches
// bug: how to minimize wav size?
// workaround? obviously ffmpeg!
// The size of the buffer (in sample-frames) which needs to 
// be processed each time onprocessaudio is called. 

// Legal values are (256, 512, 1024, 2048, 4096, 8192, 16384). 

var options = {
   bufferSize: 16384
};
var recordRTC = RecordRTC(audioStream, options);

Following values are allowed for buffer-size:

// Legal values are (256, 512, 1024, 2048, 4096, 8192, 16384)

You can write like this:

var options = {
   'buffer-size': 16384
};

sampleRate

Here is jow to customize Sample-Rate for audio recording?

// The sample rate (in sample-frames per second) at which the 
// AudioContext handles audio. It is assumed that all AudioNodes 
// in the context run at this rate. In making this assumption, 
// sample-rate converters or "varispeed" processors are not supported 
// in real-time processing.
// The sampleRate parameter describes the sample-rate of the 
// linear PCM audio data in the buffer in sample-frames per second. 

// An implementation must support sample-rates in at least 
// the range 22050 to 96000.

var options = {
   sampleRate: 96000
};
var recordRTC = RecordRTC(audioStream, options);

Values for sample-rate must be greater than or equal to 22050 and less than or equal to 96000.

You can write like this:

var options = {
   'sample-rate': 16384
};

autoWriteToDisk

Using autoWriteToDisk; you can suggest RecordRTC to auto-write to indexed-db as soon as you call stopRecording method.

var recordRTC = RecordRTC(MediaStream, {
    autoWriteToDisk: true
});

autoWriteToDisk is helpful for single stream recording and writing to disk; however for MRecordRTC; writeToDisk is preferred one.

writeToDisk

You can write recorded blob to disk using writeToDisk method:

recordRTC.stopRecording();
recordRTC.writeToDisk();

getFromDisk

You can get recorded blob from disk using getFromDisk method:

// get all blobs from disk
RecordRTC.getFromDisk('all', function(dataURL, type) {
   type == 'audio'
   type == 'video'
   type == 'gif'
});

// or get just single blob
RecordRTC.getFromDisk('audio', function(dataURL) {
   // only audio blob is returned from disk!
});

For MRecordRTC; you can use word MRecordRTC instead of RecordRTC!

Another possible situation!

var recordRTC = RecordRTC(mediaStream);
recordRTC.startRecording();
recordRTC.stopRecording(function(audioURL) {
   mediaElement.src = audioURL;
});

// "recordRTC" instance object to invoke "getFromDisk" method!
recordRTC.getFromDisk(function(dataURL) {
   // audio blob is automaticlaly returned from disk!
});

In the above example; you can see that recordRTC instance object is used instead of global RecordRTC object.

Clarifications

Is WinXP supported?

No WinXP SP2 based "Chrome" support. However, RecordRTC works on WinXP Service Pack 3.

Is Chrome on Android supported?

RecordRTC uses WebAudio API for stereo-audio recording. AFAIK, WebAudio is not supported on android chrome releases, yet.

Firefox merely supports audio-recording on Android devices.

Stereo or Mono?

Audio recording fails for mono audio. So, try stereo audio only.

MediaRecorder API (in Firefox) seems using mono-audio-recording instead.

Possible issues/failures:

This section applies only to StereoAudioRecorder:

Do you know "RecordRTC" fails recording audio because following conditions fails:

1. Sample rate and channel configuration must be the same for input and output sides on Windows i.e. audio input/output devices mismatch
2. Only the Default microphone device can be used for capturing.
3. The requesting scheme is none of the following: http, https, chrome, extension's, or file (only works with --allow-file-access-from-files)
4. The browser cannot create/initialize the metadata database for the API under the profile directory

If you see this error message: Uncaught Error: SecurityError: DOM Exception 18; it means that you're using HTTP; whilst your webpage is loading worker file (i.e. audio-recorder.js) from HTTPS. Both files's (i.e. RecordRTC.js and audio-recorder.js) scheme MUST be same!

Web Audio APIs requirements

1. If you're on Windows, you have to be running WinXP SP3, Windows Vista or better (will not work on Windows XP SP2 or earlier).
2. On Windows, audio input hardware must be set to the same sample rate as audio output hardware.
3. On Mac and Windows, the audio input device must be at least stereo (i.e. a mono/single-channel USB microphone WILL NOT work).

Why stereo?

If you explorer chromium code; you'll see that some APIs can only be successfully called for WAV files with stereo audio.

Stereo audio is only supported for WAV files.

RecordRTC is unable to record "mono" audio on chrome; however it seems that we can covert channels from "stereo" to "mono" using WebAudio API, though. MediaRecorder API's encoder only support 48k/16k mono audio channel (on Firefox Nightly).

Credits

1. Recorderjs for audio recording
2. whammy for video recording
3. jsGif for gif recording

Spec & Reference

1. Web Audio API
2. MediaRecorder
3. Canvas2D
4. MediaStream Recording
5. Media Capture and Streams

Contribute in RecordRTC.org domain

The domain www.RecordRTC.org is open-sourced here:

• https://github.com/muaz-khan/RecordRTC/tree/gh-pages

License

RecordRTC.js is released under MIT licence . Copyright (c) Muaz Khan.