-
Notifications
You must be signed in to change notification settings - Fork 536
Programmer's Guide
OpenAL is a cross-platform three-dimensional audio API. The API’s primary purpose is to allow an application to position audio sources in a three-dimensional space around a listener, producing reasonable spatialization of the sources for the audio system (headphones, 2.1 speaker output, 5.1 speaker output, etc.) Through extensions, Creative Labs has also enhanced OpenAL with EAX and other capabilities. OpenAL is appropriate for many audio applications, but was designed to be most appropriate for gaming audio.
This reference guide is most appropriate for a programmer. Experience with C or C++ is not required to learn the concepts in OpenAL, but will make understanding the OpenAL source as well as sample code easier. Since there are several sample applications included with the OpenAL SDKs as well as with the source distribution, it is recommended that interested programmers take advantage of those resources.
The two most important resources for additional information on OpenAL are the websites at www.openal.org and http://developer.creative.com. The main OpenAL site hosts the specification, the open source implementations, and sample code. The Creative developer site has a section dedicated to OpenAL with SDKs showing how to use OpenAL as well as various extensions.
Use of OpenAL revolves around the use of three fundamental objects – Buffers, Sources, and a Listener. A buffer can be filled with audio data, and can then be attached to a source. The source can then be positioned and played. How the source is heard is determined by its position and orientation relative to the Listener object (there is only one Listener). Creating a number of sources and buffers and a single listener and then updating the positions and orientations of the sources and listener dynamically can present a convincing 3D audio world.
Here is a diagram showing the fundamental OpenAL objects and their relationships to the context and device objects:
When initializing OpenAL, at least one device has to be opened. Within that device, at least one context will be created. Within that context, one listener object is implied, and a multitude of source objects can be created. Each source can have one or more buffers objects attached to it. Buffer objects are not part of a specific context – they are shared among all contexts on one device.
The function call to open a device, alcOpenDevice, takes a
string as input. The string should contain either the name of a valid OpenAL
rendering device, or NULL
to request the default device.
On PC Systems, a number of different OpenAL renderering devices may co-exist. For example a “native” renderer specific to the user’s high-end soundcard, and a host-based software fallback renderer. On platforms where multiple renderers can be present, an OpenAL application may require the ability to identify the different devices available, in order to give the end-user a choice of device. OpenAL’s Enumeration extension makes this possible.
The Enumeration extension allows the programmer to retrieve a string listing the
names of available devices. It can also provide the name of the default device.
Use alcGetString with the device property set to NULL
, and
the enum property set to ALC_DEVICE_SPECIFIER
to get the list of available
devices. To get the default device name, pass in NULL
and
ALC_DEFAULT_DEVICE_SPECIFIER
.
The Enumeration extension also works with capture devices – the equivalent
values are ALC_CAPTURE_DEVICE_SPECIFIER
and
ALC_CAPTURE_DEFAULT_DEVICE_SPECIFIER
.
The programmer can find out more about the capabilities of each device by querying to see which extensions it supports using alcIsExtensionPresent and alIsExtensionPresent.
As described above, the first step to initializing OpenAL is to open a device. Once that is successfully done, then a context is opened on that device. Now the fundamental OpenAL objects can be managed – the listener, various sources, and various buffers.
To generate a set of buffers for use, use alGetError to reset the error state, call alGenBuffers to generate the number of buffers desired, and then use alGetError again to detect if an error was generated.
Fill the buffers with PCM data using alBufferData.
To generate a set of sources for use, use alGetError to reset the error state, call alGenSources to generate the number of sources desired, and then use alGetError again to detect if an error was generated.
Buffers are attached to sources using alSourcei.
Once a buffer has been attached to a source, the source can play the buffer using alSourcePlay.
Source and Listener properties can be updated dynamically using property set and get calls such as alGetListenerfv, alListener3f, alSourcei, and alGetSource3f.
Example:
// Initialization
Device = alcOpenDevice(NULL); // select the "preferred device"
if (Device) {
Context=alcCreateContext(Device,NULL);
alcMakeContextCurrent(Context);
}
// Check for EAX 2.0 support
g_bEAX = alIsExtensionPresent("EAX2.0");
// Generate Buffers
alGetError(); // clear error code
alGenBuffers(NUM_BUFFERS, g_Buffers);
if ((error = alGetError()) != AL_NO_ERROR)
{
DisplayALError("alGenBuffers :", error);
return;
}
/ Load test.wav
loadWAVFile("test.wav",&format,&data,&size,&freq,&loop);
if ((error = alGetError()) != AL_NO_ERROR)
{
DisplayALError("alutLoadWAVFile test.wav : ", error);
alDeleteBuffers(NUM_BUFFERS, g_Buffers);
return;
}
// Copy test.wav data into AL Buffer 0
alBufferData(g_Buffers[0],format,data,size,freq);
if ((error = alGetError()) != AL_NO_ERROR)
{
DisplayALError("alBufferData buffer 0 : ", error);
alDeleteBuffers(NUM_BUFFERS, g_Buffers);
return;
}
// Unload test.wav
unloadWAV(format,data,size,freq);
if ((error = alGetError()) != AL_NO_ERROR)
{
DisplayALError("alutUnloadWAV : ", error);
alDeleteBuffers(NUM_BUFFERS, g_Buffers);
return;
}
// Generate Sources
alGenSources(1,source);
if ((error = alGetError()) != AL_NO_ERROR)
{
DisplayALError("alGenSources 1 : ", error);
return;
}
// Attach buffer 0 to source
alSourcei(source[0], AL_BUFFER, g_Buffers[0]);
if ((error = alGetError()) != AL_NO_ERROR)
{
DisplayALError("alSourcei AL_BUFFER 0 : ", error);
}
// Exit
Context=alcGetCurrentContext();
Device=alcGetContextsDevice(Context);
alcMakeContextCurrent(NULL);
alcDestroyContext(Context);
alcCloseDevice(Device);
For every context, there is automatically one Listener object. The
alListener[f, 3f, fv, i]
and alGetListener[f, 3f, fv, i]
families of
functions can be used to set or retrieve the following listener properties:
Property | Data Type | Description |
---|---|---|
AL_GAIN |
f , fv
|
Master gain. Value should be positive |
AL_POSITION |
fv , 3f , iv , 3i
|
X, Y, Z position |
AL_VELOCITY |
fv , 3f , iv , 3i
|
Velocity vector |
AL_ORIENTATION |
fv , iv
|
Orientation expressed as “at” and “up” vectors |
Example:
ALfloat listenerPos[]={0.0,0.0,0.0};
ALfloat listenerVel[]={0.0,0.0,0.0};
ALfloat listenerOri[]={0.0,0.0,-1.0, 0.0,1.0,0.0};
// Position ...
alListenerfv(AL_POSITION,listenerPos);
if ((error = alGetError()) != AL_NO_ERROR)
{
DisplayALError("alListenerfv POSITION : ", error);
return;
}
// Velocity ...
alListenerfv(AL_VELOCITY,listenerVel);
if ((error = alGetError()) != AL_NO_ERROR)
{
DisplayALError("alListenerfv VELOCITY : ", error);
return;
}
// Orientation ...
alListenerfv(AL_ORIENTATION,listenerOri);
if ((error = alGetError()) != AL_NO_ERROR)
{
DisplayALError("alListenerfv ORIENTATION : ", error);
return;
}
Each buffer generated by alGenBuffers has properties which can
be retrieved. The alGetBuffer[f, i]
functions can be used to retrieve the
following buffer properties:
Property | Data Type | Description |
---|---|---|
AL_FREQUENCY |
i , iv
|
Frequency of buffer in Hz |
AL_BITS |
i , iv
|
Bit depth of buffer |
AL_CHANNELS |
i , iv
|
Number of channels in buffer. > 1 is valid, but buffer won’t be positioned when played |
AL_SIZE |
i , iv
|
Size of buffer in bytes |
AL_DATA |
i , iv
|
Original location where data was copied from generally useless, as was probably freed after buffer creation |
Example:
// Retrieve Buffer Frequency
alBufferi(g_Buffers[0], AL_FREQUENCY, iFreq);
Each source generated by alGenSources has properties which can
be set or retrieved. The alSource[f, 3f, fv, i]
and
alGetSource[f, 3f, fv, i]
families of functions can be used to set or retrieve
the following source properties:
Property | Data Type | Description |
---|---|---|
AL_PITCH |
f , fv
|
Pitch multiplier. Always positive |
AL_GAIN |
f , fv
|
Source gain. Value should be positive |
AL_MAX_DISTANCE |
f , fv , i , iv
|
Used with the Inverse Clamped Distance Model to set the distance where there will no longer be any attenuation of the source |
AL_ROLLOFF_FACTOR |
f , fv , i , iv
|
The rolloff rate for the source. Default is 1.0 |
AL_REFERENCE_DISTANCE |
f , fv , i , iv
|
The distance under which the volume for the source would normally drop by half (before being influenced by rolloff factor or AL_MAX_DISTANCE ) |
AL_MIN_GAIN |
f , fv
|
The minimum gain for this source |
AL_MAX_GAIN |
f , fv
|
The maximum gain for this source |
AL_CONE_OUTER_GAIN |
f , fv
|
The gain when outside the oriented cone |
AL_CONE_INNER_ANGLE |
f , fv , i , iv
|
The gain when inside the oriented cone |
AL_CONE_OUTER_ANGLE |
f , fv , i , iv
|
Outer angle of the sound cone, in degrees. Default is 360 |
AL_POSITION |
fv , 3f
|
X, Y, Z position |
AL_VELOCITY |
fv , 3f
|
Velocity vector |
AL_DIRECTION |
fv , 3f , iv , 3i
|
Direction vector |
AL_SOURCE_RELATIVE |
i , iv
|
Determines if the positions are relative to the listener. Default is AL_FALSE
|
AL_SOURCE_TYPE |
i , iv
|
The source type – AL_UNDETERMINED , AL_STATIC , or AL_STREAMING
|
AL_LOOPING |
i , iv
|
Turns looping on (AL_TRUE ) or off (AL_FALSE ) |
AL_BUFFER |
i , iv
|
The ID of the attached buffer |
AL_SOURCE_STATE |
i , iv
|
The state of the source (AL_STOPPED , AL_PLAYING , …) |
AL_BUFFERS_QUEUED * |
i , iv
|
The number of buffers queued on this source |
AL_BUFFERS_PROCESSED |
i , iv
|
The number of buffers in the queue that have been processed |
AL_SEC_OFFSET |
f , fv , i , iv
|
The playback position, expressed in seconds |
AL_SAMPLE_OFFSET |
f , fv , i , iv
|
The playback position, expressed in samples |
AL_BYTE_OFFSET |
f , fv , i , iv
|
The playback position, expressed in bytes |
* Read Only (alGetSourcei)
Example:
alGetError(); // clear error state
alSourcef(source[0],AL_PITCH,1.0f);
if ((error = alGetError()) != AL_NO_ERROR)
DisplayALError("alSourcef 0 AL_PITCH : \n", error);
alGetError(); // clear error state
alSourcef(source[0],AL_GAIN,1.0f);
if ((error = alGetError()) != AL_NO_ERROR)
DisplayALError("alSourcef 0 AL_GAIN : \n", error);
alGetError(); // clear error state
alSourcefv(source[0],AL_POSITION,source0Pos);
if ((error = alGetError()) != AL_NO_ERROR)
DisplayALError("alSourcefv 0 AL_POSITION : \n", error);
alGetError(); // clear error state
alSourcefv(source[0],AL_VELOCITY,source0Vel);
if ((error = alGetError()) != AL_NO_ERROR)
DisplayALError("alSourcefv 0 AL_VELOCITY : \n", error);
alGetError(); // clear error state
alSourcei(source[0],AL_LOOPING,AL_FALSE);
if ((error = alGetError()) != AL_NO_ERROR)
DisplayALError("alSourcei 0 AL_LOOPING true: \n", error);
To continuously stream audio from a source without interruption, buffer queuing is required. To use buffer queuing, the buffers and sources are generated in the normal way, but alSourcei is not used to attach the buffers to the source. Instead, the functions alSourceQueueBuffers and alSourceUnqueueBuffers are used. The program can attach a buffer or a set of buffers to a source using alSourceQueueBuffers, and then call alSourcePlay on that source. While the source is playing, alSourceUnqueueBuffers can be called to remove buffers which have already played. Those buffers can then be filled with new data or discarded. New or refilled buffers can then be attached to the playing source using alSourceQueueBuffers. As long as there is always a new buffer to play in the queue, the source will continue to play.
Although some 1.0 implementations of OpenAL may not enforce the following restrictions on queuing, it is recommended to observe the following additional rules, which do universally apply to 1.1 implementations:
-
A source that will be used for streaming should not have its first buffer attached using alSourcei – always use alSourceQueueBuffers to attach buffers to streaming sources. Any source can have all buffers detached from it using
alSourcei(..., AL_BUFFER, 0)
, and can then be used for either streaming or non-streaming buffers depending on how data is then attached to the source (with alSourcei or with alSourceQueueBuffers). -
All buffers attached to a source using alSourceQueueBuffers should have the same audio format.
The Doppler effect depends on the velocities of source and listener relative to the medium, and the propagation speed of sound in that medium. The application might want to emphasize or deemphasize the Doppler effect as physically accurate calculation might not give the desired results. The amount of frequency shift (pitch change) is proportional to the speed of listener and source along their line of sight.
The Doppler effect as implemented by OpenAL is described by the formula below. Effects of the medium (air, water) moving with respect to listener and source are ignored.
- SS:
AL_SPEED_OF_SOUND
= speed of sound (default value 343.3) - DF:
AL_DOPPLER_FACTOR
= Doppler factor (default 1.0) - vls: Listener velocity scalar (scalar, projected on source-to-listener vector)
- vss: Source velocity scalar (scalar, projected on source-to-listener vector)
- f: Frequency of sample
- f': Effective Doppler shifted frequency
- SL: Source tt listener vector
- SV: Source velocity vector
- LV: Listener velocity vector
Graphic representation of vls and vss:
3D Mathematical representation of vls and vss:
There are two API calls global to the current context that provide control of
the speed of sound and Doppler factor. AL_DOPPLER_FACTOR
is a simple scaling
of source and listener velocities to exaggerate or deemphasize the Doppler
(pitch) shift resulting from the calculation.
A negative value will result in an AL_INVALID_VALUE
error, the command is then
ignored. The default value is 1. The current setting can be queried using
alGetFloat{v}
and AL_DOPPLER_FACTOR
.
AL_SPEED_OF_SOUND
allows the application to change the reference (propagation)
speed used in the Doppler calculation. The source and listener velocities
should be expressed in the same units as the speed of sound.