Share via

DSBUFFERDESC Structure

  • Article

[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 It has been superseded by WASAPI and Audio Graphs.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 DSBUFFERDESC structure describes the characteristics of a new buffer object. It is used by the IDirectSound8::CreateSoundBuffer method and by the DirectSoundFullDuplexCreate8 function.

An earlier version of this structure, DSBUFFERDESC1, is maintained in Dsound.h for compatibility with DirectX 7 and earlier.

Syntax

typedef struct DSBUFFERDESC {
    DWORD dwSize;
    DWORD dwFlags;
    DWORD dwBufferBytes;
    DWORD dwReserved;
    LPWAVEFORMATEX lpwfxFormat;
    GUID guid3DAlgorithm;
} DSBUFFERDESC;

Members

  • dwSize
    Size of the structure, in bytes. This member must be initialized before the structure is used.
  • dwFlags
    Flags specifying the capabilities of the buffer. See the dwFlags member of the DSBCAPS structure for a detailed listing of valid flags.
  • dwBufferBytes
    Size of the new buffer, in bytes. This value must be 0 when creating a buffer with the DSBCAPS_PRIMARYBUFFER flag. For secondary buffers, the minimum and maximum sizes allowed are specified by DSBSIZE_MIN and DSBSIZE_MAX, defined in Dsound.h.
  • dwReserved
    Reserved. Must be 0.
  • lpwfxFormat
    Address of a WAVEFORMATEX or WAVEFORMATEXTENSIBLE structure specifying the waveform format for the buffer. This value must be NULL for primary buffers.
  • guid3DAlgorithm
    Unique identifier of the two-speaker virtualization algorithm to be used by DirectSound3D hardware emulation. If DSBCAPS_CTRL3D is not set in dwFlags, this member must be GUID_NULL (DS3DALG_DEFAULT). The following algorithm identifiers are defined.
    Value Description Availability
    DS3DALG_DEFAULT DirectSound uses the default algorithm. In most cases this is DS3DALG_NO_VIRTUALIZATION. On WDM drivers, if the user has selected a surround sound speaker configuration in Control Panel, the sound is panned among the available directional speakers. Applies to software mixing only. Available on WDM or Vxd Drivers.
    DS3DALG_NO_VIRTUALIZATION 3D output is mapped onto normal left and right stereo panning. At 90 degrees to the left, the sound is coming out of only the left speaker; at 90 degrees to the right, sound is coming out of only the right speaker. The vertical axis is ignored except for scaling of volume due to distance. Doppler shift and volume scaling are still applied, but the 3D filtering is not performed on this buffer. This is the most efficient software implementation, but provides no virtual 3D audio effect. When the DS3DALG_NO_VIRTUALIZATION algorithm is specified, HRTF processing will not be done. Because DS3DALG_NO_VIRTUALIZATION uses only normal stereo panning, a buffer created with this algorithm may be accelerated by a 2D hardware voice if no free 3D hardware voices are available. Applies to software mixing only. Available on WDM or Vxd Drivers.
    DS3DALG_HRTF_FULL The 3D API is processed with the high quality 3D audio algorithm. This algorithm gives the highest quality 3D audio effect, but uses more CPU cycles. See Remarks. Applies to software mixing only. Available on Microsoft Windows 98 Second Edition and later operating systems when using WDM drivers.
    DS3DALG_HRTF_LIGHT The 3D API is processed with the efficient 3D audio algorithm. This algorithm gives a good 3D audio effect, but uses fewer CPU cycles than DS3DALG_HRTF_FULL. Applies to software mixing only. Available on Windows 98 Second Edition and later operating systems when using WDM drivers.

Remarks

When creating a primary buffer, applications must set the dwBufferBytes member to zero. DirectSound will determine the best buffer size for the particular sound device in use. To determine the size of a created primary buffer, call IDirectSoundBuffer8::GetCaps.

The DSBCAPS_CTRLDEFAULT flag is no longer supported. This flag was defined as DSBCAPS_CTRLPAN | DSBCAPS_CTRLVOLUME | DSBCAPS_CTRLFREQUENCY. By specifying only the flags you need, you cut down on unnecessary resource usage.

On VXD drivers, a sound buffer created with DSBCAPS_CTRLPOSITIONNOTIFY is always a software buffer, because the VXD driver model does not support notifications. With WDM drivers, a notification-enabled buffer can be in hardware, if hardware resources are available.

The DSBCAPS_LOCHARDWARE and DSBCAPS_LOCSOFTWARE flags are optional and mutually exclusive. DSBCAPS_LOCHARDWARE forces the buffer to reside in hardware, meaning that it will be mixed by the sound card. DSBCAPS_LOCSOFTWARE forces the buffer to reside in software, where it is mixed by the CPU. These flags are also defined for the dwFlags member DSBCAPS.

The 3D algorithms represent selection of the software emulation layer only: that is, the software algorithm that is used when no hardware is present for acceleration. In order to maximize hardware utilization, DS3DALG_NO_VIRTUALIZATION is treated as a special case. If no free 3D hardware voices are available, the buffer is then treated as a 2D buffer, but with 3D control. Specifically, when a sound buffer is created with DS3DALG_NO_VIRTUALIZATION, or when the buffer is played if the buffer was created with DSBPLAY_LOCDEFER, the following procedure is followed:

  1. If a free hardware 3D voice is available, that 3D hardware voice is used.

  2. If no free hardware 3D voices are available and a 2D hardware voice is available, that 2D hardware voice will be used. This is possible because the DS3DALG_NO_VIRTUALIZATION algorithm is a simple stereo pan algorithm

  3. If no free 2D or 3D hardware voices are available, the voice will be played in software using the DS3DALG_NO_VIRTUALIZATION algorithm.

If a speaker configuration other than DSSPEAKER_HEADPHONE or DSSPEAKER_STEREO is in effect, the processing will be done as if for a two-speaker configuration.

If a buffer is created using one of the HRTF algorithms, and the HRTF algorithm is not available on the system (for example, a non-WDM system), a success code, DS_NO_VIRTUALIZATION, is returned. The sound buffer created will use DS3DALG_NO_VIRTUALIZATION instead. For this reason, applications should use the SUCCEEDED or FAILED macros rather than checking explicitly for DS_OK when calling CreateSoundBuffer.

Requirements

Header: Declared in DSound.h.

See Also

DirectSound Buffers
Dynamic Voice Management
IDirectSound8::CreateSoundBuffer