A slightly more advanced example scene
To illustrate some of the more complex features of EAR and the Blender add-on, a slightly more advanced example scene is provided that focuses on the animation aspects of the add-on. As is seen in the simple example scene, sound sources and listeners are defined as Blender Empty objects, but these can also be animated. For that matter soundtracks of the sounds that they make and encounter in their movements can automatically be generated. The scene consists of a listener moving through a hallway, passing a room in which a pianist is playing. The listener continues his path into another room, in which another person enters as well. The acoustical phenomena that are present in this example are much more subtle than those encountered in the simple example. A more detailed documentation for all of the features and settings is being worked on. A .zip archive that contains all necessary files is available for download on github.
Result
Input
The pianist in the first room is playing a prelude by Bach.
Sound behaves very differently for the low frequencies and the high frequencies. For that reason impulse responses are generated for three distinct frequency ranges. If a single sound file is assigned to a sound source it is automatically split into these three frequency ranges by an equalizer algorithm. However, for this example three separate sound files are explicitly specified for the respective frequency ranges that consist of the bass notes, the high notes and the notes in between.
Walkthrough
Storyboards
Isolated convolutions
For reference, the convoluted sound sources are also provided separately. This gives a better impression of the acoustical effects. The convolution sound gives a clear impression of how the volume of the successive spaces alters the reverberation of the footsteps. The second shows how the balance between direct and indirect sound of the piano music alters its timbre. The third illustrates how the perceived distance of the footsteps is much larger, because of the larger indirect component. As these are individually normalized, playing them simultaneously yields a slightly different result than playing the final result.
