Typical distance-driven Spread curve designs use large Spread when the source is near and small Spread when it is far from the listener. With this approach, the source is enveloping when it is near and point-like when it is far. Additionally, when sources are near, their incident direction is subject to abrupt change and they may sound unnatural. Large Spread values mitigate these artifacts.
Similar artifacts occur when a source is located far above or below the listener and is panned onto a planar channel configuration. Although the distance is large and the Spread curve will evaluate to a small value, a source will abruptly pan from one side to the other as it crosses the listener's path in 2D. Alternatively, spreading the sound of an object from all surrounding speakers as it passes above or below the listener helps convey a sense of height, despite the lack of actual speakers in the third dimension. This is what happens when Height Spread is active.
Consider an airplane as an example, first without Height Spread. Since it flies high in the sky, its distance-based Spread will likely be close to 0%. The plane's sound will be panned as a point source and reproduced on the front left and right speaker pair. As it passes above the listener's head, it will abruptly switch to the back left and right speaker pair. This will sound unnatural.
In contrast, when Height Spread is activated, and only if panning occurs on a planar configuration, a minimal Spread value is computed based on elevation. When the elevation angle is +/-90 degrees, that is, directly above or below the listener, the contribution of the height-driven Spread is 100%. It is 0% when the elevation is 0 degrees, and it is interpolated smoothly in between. The effective Spread value is the maximum value between the height-driven Spread and the distance-driven or Spatial Audio-driven Spread. In the example above, the plane would be given a large Spread value thanks to Height Spread as it passes right above the listener, and distance-based spread would likely take over as it drops towards the horizon.
If the speaker configuration was 7.1.4, height would not contribute to the Spread calculation, because when the plane flies over the listener, it is panned normally across the four speakers on the ceiling, as would any point source. On the other hand, Height Spread contributes if the source is under the listener, because there are no speakers below.
Height Spread is not applied in an Audio Object context, but a Boolean property indicating that it is enabled is carried along with the object. It may eventually be applied at the point of panning the object on an actual channel configuration, if that configuration is planar.
In the case of ambisonic configurations, Height Spread is not applied either. This is because ambisonics is a speaker-agnostic format, and the headphone or speaker setup on which it will be later decoded is unknown at the time of panning. If that setup turned out to be a planar speaker configuration, it would not benefit from Height Spread. However, ambisonics is a representation that is intrinsically smooth, and this smoothness should suffice to mitigate the artifacts that can be caused by abrupt direction changes.
Height Spread is enabled by default. It is possible to opt-out of Height Spread by deselecting the corresponding check box in the Attenuation Editor. This may be preferable in some cases, such as with footstep sounds.
Proceed to the next 3D positioning example: Panning on a 3D Configuration.