browser compatibility

[lmick002]

Is it possible to use this package in a browser environment?

[VictorQueiroz]

@lmick002 I don't think so. Maybe if you compile libshout itself using EMScripten, but I doubt it'd work out-of-the-box.

[starapple2]

I'm not clear about usage. If it's not in a browser how do you use this package?

[VictorQueiroz]

@starapple2 You should use it in a Node.js environment. It's a shortcut to the original xiph/Icecast-libshout that's written in C.

libshout

Libshout is a library for communicating with and sending data to an
icecast server. It handles the socket connection, the timing of the
data, and prevents bad data from getting to the icecast server.

With just a few lines of code, a programmer can easily turn any application
into a streaming source for an icecast server. Libshout also allows
developers who want a specific feature set (database access, request taking)
to concentrate on that feature set, instead of worrying about how server
communication works.

Please refer to the api reference and example code to start learning how to
use libshout in your own code.

Libshout is licensed under the LGPL. Please see the COPYING file for details.

If you have any questions or comments, please visit us at
http://www.icecast.org or email us at team@icecast.org.

[starapple2]

@VictorQueiroz, thanks for your response. So I can include it in a project and connect to it from the browser via websockets, is that it? I am trying to get clarification on what I think @lmick002 was asking.

[VictorQueiroz]

@starapple2 I think the server uses TCP sockets. If I am correct, the best final solution would be to have a web socket server, that will be a proxy between the front-end and the TCP server, as it will do the translation for you.

For example:

website <> web socket server <> tcp server

[starapple2]

@VictorQueiroz, aha. Beyond the simplicity I envisaged. As is, what would be a use case? A Node.js dedktop app?

[VictorQueiroz]

@starapple2 In my case, I used it for creating a Node.js coimmand-line tool that would transmit the microphone audio to an Icecast server, so I could listen it from distance.

[starapple2]

@VictorQueiroz, in my case I want to send the mic from the browser window to Icecast.

[VictorQueiroz]

@starapple2 Then create the web socket server using a Node.js server and use the library to transmit whatever is sent to the websocket from the browser.

You can also use a WebWorker to avoid blocking the UI thread.

[starapple2]

@VictorQueiroz, I have created a server and have been looking for the correct libshout package to integrate and receive the connection from the client JavaScript. Each library I've seen appears to want a file rather than a blob that the script generates.

[VictorQueiroz]

@starapple2 Steps I suggest are:

1. Create a web socket server using a Node.js server (install libshout)
2. Send the encoded data (anything your Icecast server supports) of the microphone to the server (I suggest the OPUS codec. You can use this to encode the data to the OPUS format)
3. In the web socket server, whatever comes in as an ArrayBuffer, sends to the icecast server using libshout

It's either this, or add WebSocket support for the icecast server, which is not the best choice IMHO.

[starapple2]

@VictorQueiroz I've just added libshout to the project and will see what's what. I am looking at the Readme and seeing a const to a directory path to a file so I imagine that would need to be amended for what is sent from the encoder.

[VictorQueiroz]

@starapple2 I think you can send the file by chunks however you want and the Icecast server clients will be able to play it as soon as there's something that can be decoded.

This library behaves exactly as the original libshout library. It even has the same methods, so you can just read the libshout documentation for more information about usage of the library.

If any method is missing, let me know, and I can add it quickly. I plan to add asynchronous functionalities to it later.

[starapple2]

@VictorQueiroz , I added opus-codec to the project. Is it then redundant based on what you said just now?

This is what I had originally written for some library shout:

const shout = require('shout');

const stream = shout.createStream({
  host: 'my-server.com',
  port: 8000,
  user: 'username',
  password: 'password',
  mount: '/stream',
  name: 'My Stream',
  genre: 'Talk, Reggae',
});

This part I need to be clear about:

// Read audio data from source and write it to the stream
 const audioData = readAudioDataFromFile();
 stream.write(audioData);

// Close the stream
 stream.end();

[VictorQueiroz]

@starapple2 Seems about right, but read this from my opus library to be able to use it properly. For now, you can try on the UI thread first for simplicity, but the audio decoding is supposed to be done in a WebWorker to avoid blocking the UI thread. Otherwise, you might have cropped milliseconds.