DirectSound Playback Objects
[The feature associated with this page, DirectSound, is a legacy feature. It has been superseded by WASAPI and Audio Graphs. Media Casting have been optimized for Windows 10 and Windows 11. Microsoft strongly recommends that new code use Media Casting instead of DirectSound, when possible. Microsoft suggests that existing code that uses the legacy APIs be rewritten to use the new APIs if possible.]
The following objects are used in playing sounds:
| Object | Number | Purpose | Main interfaces |
|---|---|---|---|
| Device | One in each application | Manages the device and creates sound buffers. | IDirectSound8 |
| Secondary buffer | One for each sound | Manages a static or streaming sound and plays it into the primary buffer. | IDirectSoundBuffer8, IDirectSound3DBuffer8, IDirectSoundNotify8 |
| Primary buffer | One in each application | Mixes and plays sounds from secondary buffers, and controls global 3D parameters. | IDirectSoundBuffer, IDirectSound3DListener8 |
| Effect | Zero or more for each secondary buffer | Transforms the sound in a secondary buffer. | Interface for the particular effect, such as IDirectSoundFXChorus8 |
The first step in implementing DirectSound in an application is to create a device object, which represents the rendering device. This object is then used to create buffers.
Note
DirectSound is based on the Component Object Model (COM). However, you do not have to initialize COM explicitly, by calling CoInitialize or CoInitializeEx, unless you are using effect DMOs.
Secondary buffers are created and managed by the application. DirectSound automatically creates and manages the primary buffer, and an application can play sounds without obtaining an interface to this object. However, in order to obtain the IDirectSound3DListener8 interface, the application must explicitly create a primary buffer object.
When sounds in secondary buffers are played, DirectSound mixes them in the primary buffer and sends them to the rendering device. Only the available processing time limits the number of buffers that DirectSound can mix.
Note
Under the Windows driver model (WDM), mixing is done by the kernel mixer, and the primary buffer does not actually contain any data. For more information, see DirectSound Driver Models.
A short sound can be loaded into a secondary buffer in its entirety and played at any time by a simple method call. Longer sounds have to be streamed. An application can ascertain when it is time to stream more data into the buffer either by polling the position of the play cursor or by requesting notification when the play cursor reaches certain points.
Secondary buffers can contain effects such as chorus and echo. They can also have 3D control capabilities. Global 3D parameters are controlled by the IDirectSound3DListener8 interface on the primary buffer.