tsc_soundfx

Uses the score player as a sound effects manager.

Synopsis

tsc_soundfx

Description

Demonstrate use of ScoreContext as a simple sound effects manager. This gives you dynamic voice allocation, and a simple MIDI-like interface.

To run this program:

1) Install the sample library from the recent 3DO tools disk as documented.

If you don't have the sample library, edit this program to use your own samples. Search for .aiff.

2) Run the program, follow printed instructions.

The idea behind this program is that one can use the virtual MIDI routines in the music library as a simple sound effects manager. You can assign samples to channels, and then pay those samples using StartScoreNote(). This will take advantage of the voice allocation, automatic mixer connections, and other management code already used for playing scores. You do not have to play sequences or load MIDI files to do this.

You can load your samples based on a PIMap text file using the LoadPIMap() routine. This is documented in the manual under score playing. You may also load the samples yourself and put them in the PIMap which will give you more control. This example program shows you how to do that. Samples are assigned to Program Numbers at this step.

Assigning program numbers to channels is accomplished by calling ChangeScoreProgram(). Once you have assigned a program to a channel, you can play the sound by calling:

StartScoreNote (scon, Channel, Note, Velocity);

Channel selects the sound. Note determines the pitch and may also be used to select parts of a multisample, e.g. a drum kit. Velocity controls loudness. The voice allocation occurs during this call.

Implementation

This example was revised to take advantage of a change made for music.lib V22 to permit playing multiple voices per MIDI node, which greatly facilitates sound effects playback.

Associated Files

$samples

Location

examples/Audio

Notes

1) Prior to music.lib V22, if you wanted to get multiple voices on a given channel, you had to use different note values. The allocator would only assign one voice to any given note number, regardless of MaxVoices. music.lib V22 supports having more than one voice per note number. Earlier versions of this example demonstrated a way around this by using a different note for each voice on a channel. Since this trick is no longer necessary and is somewhat less than straightforward, this example has been modified to take advantage of the V22 feature of multiple voices per notes.

2) Prior to music.lib V24, you needed to be careful when assigning high priorities. Higher priority instruments never got stolen for lower priority instruments in that scheme. Thus you could clog up the DSPP if you have a high priority instrument with MaxVoices>1 and are also trying to play lower priority instruments. Starting with music.lib V24, higher priority instruments can be stolen by lower priority instruments if the higher priority instruments are stopped. As in the original voice stealing algorithm, voices do not get stolen from higher priority instruments if they are still playing at all (started or released).

3) Samples that are attached to an InsTemplate are deleted when the Template is deleted. You can detach the sample first by calling DeleteAttachment() if you don't want this to happen.

4) You can share the ScoreContext with code that plays a score as long as you don't overlap the channels or program numbers. You might restrict scores to channels 0-7 and programs 0-19. You could then use channels 8-15 and programs 20-39 for sound effects. The advantage of this is that they share the same voice allocation and thus resources will be shared between them. If desired, you could instead use separate independent ScoreContexts.