historical/toontown-classic.git/panda/include/fmodAudioSound.h

236 lines
7.8 KiB
C
Raw Normal View History

2024-01-16 11:20:27 -06:00
/**
* PANDA 3D SOFTWARE
* Copyright (c) Carnegie Mellon University. All rights reserved.
*
* All use of this software is subject to the terms of the revised BSD
* license. You should have received a copy of this license along
* with this source code in a file named "LICENSE."
*
* @file fmodAudioSound.h
* @author cort
* @date 2003-01-22
* Prior system by: cary
* @author Stan Rosenbaum "Staque" - Spring 2006
*
* [FIRST READ FmodAudioManager for an Introduction if you haven't
* already].
*
* Hello, all future Panda audio code people! This is my errata
* documentation to Help any future programmer maintain FMOD and PANDA.
*
* Well, if you reading this you probably want to know how PANDA deals
* with sounds directly using FMOD-EX. Well I am going to tell you.
*
* The first thing, you as the programmer have to understand,
* especially if you never have done sound programming before, is how
* the FMOD-EX API works.
*
* With FMOD-EX the guys at Firelight, adopted a model of managing
* sounds with FMOD similar to how a Sound Designer creates sound in a
* sound studio using SOUNDS and CHANNELS. Although this may seem
* strange at first, if you are not familiar with sound programming,
* there is a very good metaphor you are probably already familiar with
* to explain how FMOD-EX works.
*
* Think of you standard GUI API. Usually a GUI API is made up of two
* things: Windows and Widgets. These correspond to CHANNELS and
* SOUNDS, where a Channel is a Window and a Sound is Widget. Sounds
* are played within channels, and channels don't exist unless they
* have something to display.
*
* Now why am I explaining all of this? When PANDA was created they set
* up the basic audio classes to handle only the idea of a SOUND. The
* idea of a Channel really wasn't prevalent as in more modern Audio
* APIs. With this rewrite of PANDA to use the FMOD-EX API, the PANDA
* FmodAudioSound Class, now has to handle two different parts of the
* FMOD-EX API in order to play a sound.
*
* SOUND: The object the handles the audio data in form of WAV, AIF,
* OGG, MID, IT, MP3, etc... And CHANNEL: The object that actually
* plays the sound and manipulates it in real time.
*
* Ultimately this isn't a problem expect for a couple situations when
* you go to play a sound, which I will explain in more detail in that
* part of the code. All that you have to know right now is that
* Channels in FMOD do not exist unless they are playing a sound. And
* in the PANDA FmodAudioSound API class there is only ONE dedicated
* channel per sound. Otherwise there is really nothing to worry
* about.
*/
#ifndef __FMOD_AUDIO_SOUND_H__
#define __FMOD_AUDIO_SOUND_H__
#include "pandabase.h"
#include "audioSound.h"
#include "reMutex.h"
#include "fmodAudioManager.h"
#include <fmod.hpp>
#include <fmod_errors.h>
class VirtualFile;
class EXPCL_FMOD_AUDIO FmodAudioSound : public AudioSound {
public:
FmodAudioSound(AudioManager *manager, VirtualFile *file, bool positional);
~FmodAudioSound();
// For best compatibility, set the loop_count, start_time, volume, and
// balance, prior to calling play(). You may set them while they're
// playing, but it's implementation specific whether you get the results.
void play();
void stop();
// loop: false = play once; true = play forever. inits to false.
void set_loop(bool loop=true);
bool get_loop() const;
// loop_count: 0 = forever; 1 = play once; n = play n times. inits to 1.
void set_loop_count(unsigned long loop_count=1);
unsigned long get_loop_count() const;
// 0 = beginning; length() = end. inits to 0.0.
void set_time(PN_stdfloat start_time=0.0);
PN_stdfloat get_time() const;
// 0 = minimum; 1.0 = maximum. inits to 1.0.
void set_volume(PN_stdfloat volume=1.0);
PN_stdfloat get_volume() const;
// -1.0 is hard left 0.0 is centered 1.0 is hard right inits to 0.0.
void set_balance(PN_stdfloat balance_right=0.0);
PN_stdfloat get_balance() const;
// play_rate is any positive float value. inits to 1.0.
void set_play_rate(PN_stdfloat play_rate=1.0f);
PN_stdfloat get_play_rate() const;
const std::string &get_name() const;
// return: playing time in seconds.
PN_stdfloat length() const;
// Controls the position of this sound's emitter. pos is a pointer to an
// xyz triplet of the emitter's position. vel is a pointer to an xyz
// triplet of the emitter's velocity.
void set_3d_attributes(PN_stdfloat px, PN_stdfloat py, PN_stdfloat pz, PN_stdfloat vx, PN_stdfloat vy, PN_stdfloat vz);
void get_3d_attributes(PN_stdfloat *px, PN_stdfloat *py, PN_stdfloat *pz, PN_stdfloat *vx, PN_stdfloat *vy, PN_stdfloat *vz);
void set_3d_min_distance(PN_stdfloat dist);
PN_stdfloat get_3d_min_distance() const;
void set_3d_max_distance(PN_stdfloat dist);
PN_stdfloat get_3d_max_distance() const;
AudioSound::SoundStatus status() const;
virtual PN_stdfloat get_speaker_mix(int speaker);
virtual void set_speaker_mix(PN_stdfloat frontleft, PN_stdfloat frontright, PN_stdfloat center, PN_stdfloat sub, PN_stdfloat backleft, PN_stdfloat backright, PN_stdfloat sideleft, PN_stdfloat sideright);
void set_active(bool active=true);
bool get_active() const;
void finished();
void set_finished_event(const std::string& event);
const std::string& get_finished_event() const;
private:
PT(FmodAudioManager) _manager;
FMOD::Sound *_sound;
FMOD::Channel *_channel;
Filename _file_name;
float _volume;
float _balance;
float _playrate;
int _priority;
float _mix[AudioManager::SPK_COUNT];
float _sampleFrequency;
mutable float _length; //in seconds.
FMOD_SPEAKERMODE _speakermode;
FMOD_VECTOR _location;
FMOD_VECTOR _velocity;
PN_stdfloat _min_dist;
PN_stdfloat _max_dist;
void start_playing();
void set_volume_on_channel();
void set_balance_on_channel();
void set_play_rate_on_channel();
void set_speaker_mix_on_channel();
void set_3d_attributes_on_channel();
// void add_dsp_on_channel();
void set_speaker_mix_or_balance_on_channel();
virtual int get_priority();
virtual void set_priority(int priority);
bool _active;
bool _paused;
PN_stdfloat _start_time;
std::string _finished_event;
// This reference-counting pointer is set to this while the sound is
// playing, and cleared when we get an indication that the sound has
// stopped. This prevents a sound from destructing while it is playing. We
// use a PT instead of managing the reference counts by hand to help guard
// against accidental reference count leaks or other mismanagement.
PT(FmodAudioSound) _self_ref;
static FMOD_RESULT F_CALLBACK
sound_end_callback(FMOD_CHANNEL * channel,
FMOD_CHANNEL_CALLBACKTYPE type,
void *commanddata1,
void *commanddata2);
static FMOD_RESULT F_CALLBACK
open_callback(const char *name, int unicode, unsigned int *file_size,
void **handle, void **user_data);
static FMOD_RESULT F_CALLBACK
close_callback(void *handle, void *user_data);
static FMOD_RESULT F_CALLBACK
read_callback(void *handle, void *buffer, unsigned int size_bytes,
unsigned int *bytes_read, void *user_data);
static FMOD_RESULT F_CALLBACK
seek_callback(void *handle, unsigned int pos, void *user_data);
// These are needed for Panda's Pointer System. DO NOT ERASE!
public:
static TypeHandle get_class_type() {
return _type_handle;
}
static void init_type() {
AudioSound::init_type();
register_type(_type_handle, "FmodAudioSound", AudioSound::get_class_type());
}
virtual TypeHandle get_type() const {
return get_class_type();
}
virtual TypeHandle force_init_type() {
init_type();
return get_class_type();
}
private:
static TypeHandle _type_handle;
// DONE
};
#include "fmodAudioSound.I"
#endif /* __FMOD_AUDIO_SOUND_H__ */