![[Go to CFHT Home Page]](/images/cfht-icon.gif){kind=icon}
Man Pages 
Back to Software Index 
Manpage Top Level
The audio interface described below is an uncommitted interface and may be replaced in the future.
An audio device is used to play and/or record a stream of audio data. Since a specific audio device may not support all of the functionality described below, refer to the device-specific manual pages for a complete description of each hardware device. An application can use the AUDIO_GETDEV ioctl(2) to determine the current audio hardware associated with /dev/audio.
Digital audio data represents a quantized approximation of an analog audio signal waveform. In the simplest case, these quantized numbers represent the amplitude of the input waveform at particular sampling intervals. In order to achieve the best approximation of an input signal, the highest possible sampling frequency and precision should be used. However, increased accuracy comes at a cost of increased data storage requirements. For instance, one minute of monaural audio recorded in *m-law format at 8 KHz requires nearly 0.5 megabytes of storage, while the standard Compact Disc audio format (stereo 16-bit linear PCM data sampled at 44.1 KHz) requires approximately 10 megabytes per minute.
Audio data may be represented in several different formats. An audio device’s current audio data format can be determined by using the AUDIO_GETINFO ioctl described below.
An audio data format is characterized in the audio driver by four parameters: Sample Rate, Encoding, Precision, and Channels. Refer to the device-specific manual pages for a list of the audio formats that each device supports. In addition to the formats that the audio device supports directly, other formats provide higher data compression. Applications may convert audio data to and from these formats when recording or playing.
Sample rate is a number
that represents the sampling frequency (in samples per second) of the
audio data.
An encoding parameter specifies the audio data representation. *m-law encoding (pronounced mew-law) corresponds to CCITT G.711 , and is the standard for voice data used by telephone companies in the United States, Canada, and Japan. A-law encoding is also part of G.711 , and is the standard encoding for telephony elsewhere in the world. A-law and *m-law audio data are sampled at a rate of 8000 samples per second with 12-bit precision, with the data compressed to 8-bit samples. The resulting audio data quality is equivalent to that of standard analog telephone service.
Linear Pulse Code Modulation (PCM ) is an uncompressed audio format in which sample values are directly proportional to audio signal voltages. Each sample is a 2’s complement number that represents a positive or negative amplitude.
Precision indicates the number of bits used to store each audio sample. For instance, *m-law and A-law data are stored with 8-bit precision. PCM data may be stored at various precisions, though 16-bit PCM is most common.
Multiple channels of audio may be interleaved at sample boundaries. A sample frame consists of a single sample from each active channel. For example, a sample frame of stereo 16-bit PCM data consists of 2 16-bit samples, corresponding to the left and right channel data.
The device /dev/audio is a device driver that dispatches audio requests to the appropriate underlying audio device driver. The audio driver is implemented as a STREAMS driver. In order to record audio input, applications open(2) the /dev/audio device and read data from it using the read(2) system call. Similarly, sound data is queued to the audio output port by using the write(2) system call. Device configuration is performed using the ioctl(2) interface.
As some systems may contain more than one audio device, application writers are encouraged to query the AUDIODEV environment variable. If this variable is present in the environment, its value should identify the path name of the default audio device.
The audio device is treated as an exclusive resource - only one process can open the device at a time. However, two processes may simultaneously access the device: if one opens it read-only, then another may open it write-only.
When a process cannot open /dev/audio because the requested access mode is busy:
- if either the O_NDELAY or O_NONBLOCK flag are set in the open() oflag argument, then -1 is immediately returned, with errno set to EBUSY .
- if neither the O_NDELAY nor the O_NONBLOCK flag are set, then open() hangs until the device is available or a signal is delivered to the process, in which case a -1 is returned with errno set to EINTR . This allows a process to block in the open call, while waiting for the audio device to become available.
Upon the initial open() of the audio device, the driver will reset the data format of the device to the default state of 8-bit, 8Khz, mono *m-law data. If the device is already open and a different audio format has been set, this will not be possible. Audio applications should explicitly set the encoding characteristics to match the audio data requirements, rather than depend on the default configuration.
Since the audio device grants exclusive read or write access to a single process at a time, long-lived audio applications may choose to close the device when they enter an idle state and reopen it when required. The play.waiting and record.waiting flags in the audio information structure (see below) provide an indication that another process has requested access to the device. For instance, a background audio output process may choose to relinquish the audio device whenever another process requests write access.
The read() system call copies data from the system buffers to the application. Ordinarily, read() blocks until the user buffer is filled. The I_NREAD ioctl (see streamio(7I) ) may be used to determine the amount of data that may be read without blocking. The device may alternatively be set to a non-blocking mode, in which case read() completes immediately, but may return fewer bytes than requested. Refer to the read(2) manual page for a complete description of this behavior.
When the audio device is opened with read access, the device driver immediately starts buffering audio input data. Since this consumes system resources, processes that do not record audio data should open the device write-only (O_WRONLY ).
The transfer of input data to STREAMS buffers may be paused (or resumed) by using the AUDIO_SETINFO ioctl to set (or clear) the record.pause flag in the audio information structure (see below). All unread input data in the STREAMS queue may be discarded by using the I_FLUSH STREAMS ioctl (see streamio(7I) ). When changing record parameters, the input stream should be paused and flushed before the change, and resumed afterward. Otherwise, subsequent reads may return samples in the old format followed by samples in the new format. This is particularly important when new parameters result in a changed sample size.
Input data can accumulate in STREAMS buffers very quickly. At a minimum, it will accumulate at 8000 bytes per second for 8-bit, 8 KHz, mono, *m-law data. If the device is configured for 16-bit linear or higher sample rates, it will accumulate even faster. If the application that consumes the data cannot keep up with this data rate, the STREAMS queue may become full. When this occurs, the record.error flag is set in the audio information structure and input sampling ceases until there is room in the input queue for additional data. In such cases, the input data stream contains a discontinuity. For this reason, audio recording applications should open the audio device when they are prepared to begin reading data, rather than at the start of extensive initialization.
The write() system call copies data from an applications buffer to the STREAMS output queue. Ordinarily, write() blocks until the entire user buffer is transferred. The device may alternatively be set to a non-blocking mode, in which case write() completes immediately, but may have transferred fewer bytes than requested (see write(2) ).
Although write() returns when the data is successfully queued, the actual completion of audio output may take considerably longer. The AUDIO_DRAIN ioctl may be issued to allow an application to block until all of the queued output data has been played. Alternatively, a process may request asynchronous notification of output completion by writing a zero-length buffer (end-of-file record) to the output stream. When such a buffer has been processed, the play.eof flag in the audio information structure (see below) is incremented.
The final close(2) of the file descriptor hangs until audio output has drained. If a signal interrupts the close(), or if the process exits without closing the device, any remaining data queued for audio output is flushed and the device is closed immediately.
The conversion of output data may be paused (or resumed) by using the AUDIO_SETINFO ioctl to set (or clear) the play.pause flag in the audio information structure. Queued output data may be discarded by using the I_FLUSH STREAMS ioctl.
Output data will be played from the STREAMS buffers at a rate of at least 8000 bytes per second for *m-law or A-law data (faster for 16-bit linear data or higher sampling rates). If the output queue becomes empty, the play.error flag is set in the audio information structure and output is stopped until additional data is written. If an application attempts to write a number of bytes that is not a multiple of the current sample frame size, an error will be generated and the device will need to be closed before any future writes will succeed.
The I_SETSIG STREAMS ioctl enables asynchronous notification, through the SIGPOLL signal, of input and output ready conditions. The O_NONBLOCK flag may be set using the F_SETFL fcntl(2) to enable non-blocking read() and write() requests. This is normally sufficient for applications to maintain an audio stream in the background.
It is sometimes convenient to have an application, such as a volume control panel, modify certain characteristics of the audio device while it is being used by an unrelated process. The /dev/audioctl pseudo-device is provided for this purpose. Any number of processes may open /dev/audioctl simultaneously. However, read() and write() system calls are ignored by /dev/audioctl. The AUDIO_GETINFO and AUDIO_SETINFO ioctl commands may be issued to /dev/audioctl to determine the status or alter the behavior of /dev/audio. Note: In general, the audio control device name is constructed by appending the letters "ctl" to the path name of the audio device.
Applications that open the audio control pseudo-device may request asynchronous notification of changes in the state of the audio device by setting the S_MSG flag in an I_SETSIG STREAMS ioctl. Such processes receive a SIGPOLL signal when any of the following events occur:
- An AUDIO_SETINFO ioctl has altered the device state.
- An input overflow or output underflow has occurred.
- An end-of-file record (zero-length buffer) has been processed on output.
- An open() or close() of /dev/audio has altered the device state.
- An external event (such as speakerbox volume control) has altered the device state.
The state of the audio device may be polled or modified using the AUDIO_GETINFO and AUDIO_SETINFO ioctl commands. These commands operate on the audio_info structure as defined, in <sys/audioio.h>, as follows:
/* This structure contains state information for audio device
IO streams */
struct audio_prinfo {
/* The following values describe the audio data encoding */
uint_t sample_rate; /* samples per second */
uint_t channels; /* number of interleaved channels */
uint_t precision; /* number of bits per sample */
uint_t encoding; /* data encoding method */
/* The following values control audio device configuration */
uint_t gain; /* volume level */
uint_t port; /* selected I/O port */
uint_t buffer_size; /* I/O buffer size */
/* The following values describe the current device state */
uint_t samples; /* number of samples converted */
uint_t eof; /* End Of File counter (play only) */
uchar_t pause; /* non-zero if paused, zero to resume */
uchar_t error; /* non-zero if overflow/underflow */
uchar_t waiting; /* non-zero if a process wants access */
uchar_t balance; /* stereo channel balance */
/* The following values are read-only device state flags */
uchar_t open; /* non-zero if open access granted */
uchar_t active; /* non-zero if I/O active */
uint_t avail_ports; /* available I/O ports */
} audio_prinfo_t;
/* This structure is used in AUDIO_GETINFO and AUDIO_SETINFO ioctl
commands */
typedef struct audio_info {
audio_prinfo_t record; /* input status information */
audio_prinfo_t play; /* output status information */
uint_t monitor_gain; /* input to output mix */
uchar_t output_muted; /* non-zero if output muted */
} audio_info_t;
/* Audio encoding types */
#define AUDIO_ENCODING_ULAW (1) /* u-law encoding */
#define AUDIO_ENCODING_ALAW (2) /* A-law encoding */
#define AUDIO_ENCODING_LINEAR (3) /* Linear PCM encoding */
/* These ranges apply to record, play, and monitor gain values */
#define AUDIO_MIN_GAIN (0) /* minimum gain value */
#define AUDIO_MAX_GAIN (255) /* maximum gain value */
/* These values apply to the balance field to adjust channel gain values
*/
#define AUDIO_LEFT_BALANCE (0) /* left channel only */
#define AUDIO_MID_BALANCE (32) /* equal left/right balance */
#define AUDIO_RIGHT_BALANCE (64) /* right channel only */
/* Define some convenient audio port names (for port and avail_ports) */
/* output ports (several might be enabled at once) */
#define AUDIO_SPEAKER (0x01) /* output to built-in speaker */
#define AUDIO_HEADPHONE (0x02) /* output to headphone jack */
#define AUDIO_LINE_OUT (0x04) /* output to line out */
/* input ports (usually only one may be enabled at a time) */
#define AUDIO_MICROPHONE (0x01) /* input from microphone */
#define AUDIO_LINE_IN (0x02) /* input from line in */
#define MAX_AUDIO_DEV_LEN (16)
/* Parameter for the AUDIO_GETDEV ioctl */
typedef struct audio_device {
char name[MAX_AUDIO_DEV_LEN];
char version[MAX_AUDIO_DEV_LEN];
char config[MAX_AUDIO_DEV_LEN];
} audio_device_t;
The play.gain and record.gain fields specify the output and input volume levels. A value of AUDIO_MAX_GAIN indicates maximum volume. Audio output may also be temporarily muted by setting a non-zero value in the output_muted field. Clearing this field restores audio output to the normal state. Most audio devices allow input data to be monitored by mixing audio input onto the output channel. The monitor_gain field controls the level of this feedback path.
The play.port field controls the output path for the audio device. It can be set to either AUDIO_SPEAKER (built-in speaker), AUDIO_HEADPHONE (headphone jack), or AUDIO_LINE_OUT (line-out port). For some devices, it may be set to a combination of these ports. The play.avail_ports field returns the set of output ports that are currently accessible. The input ports can be either AUDIO_MICROPHONE or AUDIO_LINE_IN . The record.avail_ports field returns the set of input ports that are currently accessible.
The play.balance and record.balance fields are used to control the volume between the left and right channels when manipulating stereo data. When the value is set between AUDIO_LEFT_BALANCE and AUDIO_MID_BALANCE , the right channel volume will be reduced in proportion to the balance value. Conversely, when balance is set between AUDIO_MID_BALANCE and AUDIO_RIGHT_BALANCE , the left channel will be proportionally reduced.
The play.pause and record.pause flags may be used to pause and resume the transfer of data between the audio device and the STREAMS buffers. The play.error and record.error flags indicate that data underflow or overflow has occurred. The play.active and record.active flags indicate that data transfer is currently active in the corresponding direction.
The play.open and record.open flags indicate that the device is currently open with the corresponding access permission. The play.waiting and record.waiting flags provide an indication that a process may be waiting to access the device. These flags are set automatically when a process blocks on open(), though they may also be set using the AUDIO_SETINFO ioctl command. They are cleared only when a process relinquishes access by closing the device.
The play.samples and record.samples fields are initialized, at open(), to zero and increment each time a data sample is copied to or from the associated STREAMS queue. Some audio drivers may be limited to counting buffers of samples, instead of single samples for the samples accounting. For this reason, applications should not assume that the samples fields contain a perfectly accurate count. The play.eof field increments whenever a zero-length output buffer is synchronously processed. Applications may use this field to detect the completion of particular segments of audio output.
The record.buffer_size field controls the amount of input data that is buffered in the device driver during record operations. Applications that have particular requirements for low latency should set the value appropriately. Note however that smaller input buffer sizes may result in higher system overhead. The value of this field is specified in bytes and drivers will constrain it to be a multiple of the current sample frame size. Some drivers may place other requirements on the value of this field. Refer to the audio device-specific manual page for more details. If an application changes the format of the audio device and does not modify the record.buffer_size field, the device driver may use a default value to compensate for the new data rate. Therefore, if an application wishes to modify this field, it should modify it during or after the format change itself, not before. The record.buffer_size field may be modified only on the /dev/audio device by processes that have it opened for reading. The play.buffer_size field is currently not supported.
The audio data format is indicated by the sample_rate, channels, precision, and encoding fields. The values of these fields correspond to the descriptions in the AUDIO section above. Refer to the audio device-specific manual pages for a list of supported data format combinations.
The data format fields may be modified only on the /dev/audio device. The audio hardware will often constrain the input and output data formats to be identical. If this is the case, then the data format may not be changed if multiple processes have opened the audio device.
If the parameter changes requested by an AUDIO_SETINFO ioctl cannot all be accommodated, ioctl() will return with errno set to EINVAL and no changes will be made to the device state.
All of the streamio(7I) ioctl commands may be issued for the /dev/audio device. Because the /dev/audioctl device has its own STREAMS queues, most of these commands neither modify nor report the state of /dev/audio if issued for the /dev/audioctl device. The I_SETSIG ioctl may be issued for /dev/audioctl to enable the notification of audio status changes, as described above.
The audio device additionally supports the following ioctl commands:
audio_info_t info; AUDIO_INITINFO(&info); info.play.port = AUDIO_HEADPHONE; err = ioctl(audio_fd, AUDIO_SETINFO, &info);
An open() will fail if:
An ioctl() will fail if:
The physcial audio device names are system dependent and are rarely used by programmers. The programmer should use the generic device names listed below.
Due to a feature of the STREAMS implementation, programs that are terminated or exit without closing the audio device may hang for a short period while audio output drains. In general, programs that produce audio output should catch the SIGINT signal and flush the output stream before exiting.
On LX machines running Solaris 2.3, catting a demo audio file to the audio device /dev/audio does not work. Use the audioplay command on LX machines instead of cat.
Future workstation audio
resources will be managed by an audio foundation library. For the time
being, we encourage you to write your programs in a modular fashion, isolating
the audio device-specific functions, so that they may be easily ported to
such an environment.
The AUDIO_GETDEV ioctl is provided for the future implementation of an audio device capability database. In general, applications may use the play.avail_ports and record.avail_ports fields of the audio_info structure to determine the audio device capabilities.