# AudioStreaming

**Repository Path**: wanggjghost/AudioStreaming

## Basic Information

- **Project Name**: AudioStreaming
- **Description**: An example of how to streaming audio over network using Qt
- **Primary Language**: C++
- **License**: MIT
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2019-07-31
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Audio Streaming library for Qt 5

Cross platform audio streaming library for local networks and VOIP for both local networks and internet.

### Features:

* Fully multi threaded.
* Audio can be compressed using Opus codec.
* Audio can be buffered to reduce the occurrence of buffer underrun in the client side.
* Can work on Broadcast or Walkie Talkie modes.
* Tested up with Qt 5.5.1 and later.
* Low latency modes are supported.
* Smart buffer available.

### Needed:

Qt 5.x

C++11 compiler.

You can choose to compress audio with Opus codec or not. If choose to compress you'll need to compile [Opus](https://www.opus-codec.org/) to your targets(Windows, Linux, macOS or Android), and need to download [r8brain](https://github.com/avaneev/r8brain-free-src). To compile Opus for android see the opus-android README.

If you want to static link [OpenSsl](https://www.openssl.org/),(independently if generating a shared or static library) just compile OpenSsl to fit your needs.

If not compressing audio and not directly dealing with OpenSsl the only requirement is Qt5 and a C++11 compiler.

### Compilation Info:

#### QMake:

To compile the library you need to choose if you want to compress audio using Opus codec or not, and if you want to static link OpenSsl or not.

If you want to use Opus edit AudioStreamingLibSettings.pri setting `OPUS_ENABLED = TRUE`, with the appropriated version of Opus compiled just set the include folder of r8brain and Opus, and the Opus libraries location in AudioStreamingLibSettings.pri. If you don't want to use Opus leave `OPUS_ENABLED` blank or with any other value than `TRUE`.

The same applies to OpenSsl, if you want to static link OpenSsl set `OPENSSL_ENABLED = TRUE`.(Please not that even if OpenSsl won't be statically linked it may be available through dynamic linking at runtime.)
With the appropriated version of OpenSsl compiled just set the include folder and the libraries location of OpenSsl in AudioStreamingLibSettings.pri.

With the appropriate settings set (using Opus or not, static linking OpenSsl or not) build the buildlib.pro, if it's already built and you are changing settings remember to re-run qmake and rebuild.

#### Qbs:

Compile with `Qbs` is simpler than with `QMake`, you'll still need to choose compress audio, statically links `OpenSsl`, and set this on the.`qbs` file as a property. 

Different from `QMake`, there is no two build steps, build each application causes the library to be built together.

Please note that Android builds are not currently compatible with `Qbs`.

### Using the library:

#### QMake:

Using the library with `QMake` is easy, just add `include(path/to/AudioStreamingLib.pri)` to your project file.

#### Qbs:

Add to your `.qbs` file: `references: ["..path_to../AudioStreamingLibCore/AudioStreamingLibCore/AudioStreamingLibCore.qbs"]`

#### Coding:

To get started coding add `#include <AudioStreamingLibCore>` to your header/source file.

This project comes with demos of usage, just build and test them.

For more information of how to use the APIs, please see the [Wiki](https://github.com/antonypro/AudioStreaming/wiki).

### Using the server WebServer:

This repository also includes a small server for who wants a server managing the connections, on local networks or the internet!

It's just a Server that manages users with ID and password on a SQLite database thus do not require a SQL server. When one user connects to other, all traffic is routed through server, the original P2P approach doesn't work in a real scenario and was abandoned.

### Smart buffer:

The smart buffer is activated when the time to buffer is 0. It works on `AudioOutput` class and analyze audio input rate and hardware performance to determine the lowest latency to prevent dropouts.

### Third party libraries:

[Eigen](http://eigen.tuxfamily.org/)

[Opus](http://opus-codec.org/)

[r8brain](https://github.com/avaneev/r8brain-free-src)

[OpenSsl](https://www.openssl.org/)

[Kiss FFT](http://kissfft.sourceforge.net/)