audio_falloff_set_model

To add more versatility to the audio engine, GameMaker permits you to select the falloff model that suits your game. This model will be used for all the audio functions in the game or app, and so you should make sure that the model you choose is the correct one, as each one will affect how the listener perceives the sounds you play through emitters or with the function audio_play_sound_at().

The default falloff model is audio_falloff_none, meaning there is no falloff when using emitters or positioned audio unless you change the falloff model.

Falloff Models

When playing audio through audio_play_sound_at() or setting the falloff for an emitter, there are three arguments that you will need to set, and each one is appropriate to a specific model and will affect the way the final sound is heard by the player depending on the distance that the listener is from the source. The three arguments are:

Reference distance: this is the distance from the listener under which the volume for the sound playing would normally drop by half before being influenced by roll-off factor or the specified maximum distance.
Maximum distance: this sets the distance where there will no longer be any attenuation of the source sound. This can be the point at which the sound is no longer heard or the point at which the sound volume no longer decreases below the minimum threshold defined by the model chosen.
Falloff factor: The falloff factor is used in distance attenuation based on the inverse distance model and sets the final minimum threshold for a sound with falloff.

The falloff models that are affected by these arguments are represented in GameMaker by the following constants (the table shows the exact calculations used too):

Audio Falloff Constant
Constant	Gain Calculation
audio_falloff_exponent_distance	gain = (listener_distance / reference_distance) ^ (-falloff_factor)
audio_falloff_exponent_distance_clamped	distance = clamp(listener_distance, reference_distance, maximum_distance) gain = (distance / reference_distance) ^ (-falloff_factor)
audio_falloff_exponent_distance_scaled	distance = clamp(listener_distance, reference_distance, maximum_distance) gain = ((distance / reference_distance) ^ (-falloff_factor)) * (((maximum_distance - distance) / (maximum_distance - reference_distance)) ^ (distance / maximum_distance))
audio_falloff_inverse_distance	gain = reference_distance / (reference_distance + falloff_factor * (listener_distance - reference_distance))
audio_falloff_inverse_distance_clamped	distance = clamp(listener_distance, reference_distance, maximum_distance) gain = reference_distance / (reference_distance + falloff_factor * (distance - reference_distance))
audio_falloff_inverse_distance_scaled	distance = clamp(listener_distance, reference_distance, maximum_distance) gain = (reference_distance / (reference_distance + falloff_factor * (distance - reference_distance))) * (((maximum_distance - distance) / (maximum_distance - reference_distance)) ^ (distance / maximum_distance))
audio_falloff_linear_distance	distance = min(distance, maximum_distance) gain = (1 - falloff_factor * (distance - reference_distance) / (maximum_distance - reference_distance))
audio_falloff_linear_distance_clamped	distance = clamp(listener_distance, reference_distance, maximum_distance) gain = (1 - falloff_factor * (distance - reference_distance) / (maximum_distance - reference_distance))
audio_falloff_none	gain = 1

The "_scaled" models are scaled in such a way that sounds are guaranteed to fall off entirely by the maximum distance.

The following graphs are visual representations of how some of the above constants work and affect the sound being played:

Distance Model Examples

HTML5 Limitations

In HTML5 the _clamped and _scaled variants are unsupported, as Web Audio doesn't support them. If you set the falloff model to one of these constants on HTML5, GameMaker will internally use the nearest available model as follows:

exponent_distance_clamped / exponent_distance_scaled maps to exponent_distance
inverse_distance_clamped / inverse_distance_scaled maps to inverse_distance
linear_distance_clamped maps to linear_distance

This change only happens internally, the falloff model value remains the one that you passed to the function.

Syntax:

audio_falloff_set_model(model);

Argument	Type	Description
model	Audio Falloff Constant	The constant used to set the falloff model.

Returns:

N/A

Example:

audio_falloff_set_model(audio_falloff_exponent_distance_clamped);
audio_play_sound_at(snd_Waterfall, x, y, 0, 100, 300, 1, true, 1);