Skip to content

Home

Jump to bottom
Jon McGuire edited this page Feb 1, 2024 · 17 revisions

The eyecandy library has several basic parts:

  • Audio capture and audio data (raw and processed)
  • Multi-threaded conversion of audio data to OpenGL textures
  • Audio texture management
  • Helper classes for OpenGL window and shader management

This wiki explains the technical aspects of the library and how to write programs using the library. It is oriented to developers who are already familiar with modern .NET development.

If you're just looking for a visualizer, you should look at my monkey-hi-hat application instead. A Windows installer is available in the Releases section, and usage is detailed in the wiki.

There users and developers alike will also find help setting up your Windows or Linux system for audio loopback in the Quick-Start sections, as well as help understanding audio textures and porting or creating visualization shaders, which are also important topics for writing your own eyecandy-based application.

Although the original plan was to use a Raspberry Pi 4B, GPU performance is worse than expected. Also since they're currently stuck on the very old OpenGL ES 3.1 (with some 3.2 capabilities) it can't support some of the features I want in the monkey-hi-hat program. Consequently, as of the v2 release, the default GL setting has changed to BaseWindow.ForceOpenGLES3dot2 = false. This means the OpenTK default of full-API OpenGL 3.3 will be used by this library unless you override it. The monkey-hi-hat project is targeting GL 4.5, which is the last version supported by Linux MESA drivers (and 4.6 is supposedly the last version of OpenGL ever, while the Khronos Foundation disappears down the rabbit-hole of the "new" Vulkan API).

Major Dependencies

OpenAL and OpenAL-Soft is only required on Linux. Windows can use OpenAL-based audio capture (with a third-party audio loopback driver) but as of eyecandy v3, the default Windows audio capture methodology is to use the internal WASAPI loopback device. (It also provides higher quality data with lower latency.)

Clone this wiki locally