Unit CastleSoundBase

How does distance affect spatial sounds, used for TSoundEngine.DistanceModel.

Values

• dmInverse: Sound is fully audible (with TCastleSound.Volume) at the distance TCastleSound.ReferenceDistance, and then it drops down with distance following a realistic curve. Sound is not attenuated more when it is further than TCastleSound.MaxDistance.

  This is the default distance model, that is realistic and is easy to use.

  The description above is not 100% precise, because different sound backends have different details how is this realized. For OpenAL, this means using default AL_INVERSE_DISTANCE_CLAMPED (the exact equation is on https://www.openal.org/documentation/openal-1.1-specification.pdf ). For FMOD, this means using default FMOD_3D_INVERSEROLLOFF (see https://www.fmod.com/resources/documentation-api?version=2.01&page=core-api-common.html#fmod_3d_inverserolloff , https://www.fmod.com/resources/documentation-api?version=2.01&page=white-papers-3d-sounds.html#inverse , "for every doubling of this mindistance, the sound volume will halve" – so it is inverse squared).

• dmLinear: Sound is fully audible (with TCastleSound.Volume) at the distance TCastleSound.ReferenceDistance, and then it drops down linearly to 0 at TCastleSound.MaxDistance. Sound is no more audible at TCastleSound.MaxDistance.

  It is easy to control, but remember to set TCastleSound.MaxDistance to sensibly large value.

  This matches the X3D sound model most.

Source: audio/castlesoundbase.pas (line 30).

How to load a sound buffer.

Values

• slComplete: Load entire sound file at once. The advantage is that once the sound buffer is loaded, there's zero overhead at runtime for playing it, and loading the sound buffer multiple times uses the cache properly. The disadvantage is that loading time may be long, for longer files.
• slStreaming: Decompress the sound (like OggVorbis) during playback. It allows for much quicker sound loading (almost instant, if you use streaming for everything) but means that sounds will be loaded (in parts) during playback. In general case, we advise to use it for longer sounds (like music tracks).

Source: audio/castlesoundbase.pas (line 75).

Frequency (sample rate) of the loaded sound file. This is a floating point type, because some sound backends (like FMOD or WebAudio) allow for non-integer frequencies. Used for TCastleSound.Frequency.

Source: audio/castlesoundbase.pas (line 94).