javax.sound.sampled
This package specifies interfaces for capture, mixing, and playback of digital (sampled) audio.

javax.sound.midi
This package provides interfaces for MIDI synthesis, sequencing, and event transport.

Lines

A line is an element of the digital audio "pipeline," such as an audio input or output port, a mixer, or an audio data path into or out of a mixer. The audio data flowing through a line can be mono or multichannel (for example, stereo). Each type of line will be described shortly, but first some of their functional relationships will be illustrated, showing the flow of audio through the "pipeline." The following diagram shows different types of lines in a simple audio-output system:

In this example, an application has asked a mixer for one or more available clips and source data lines. A clip is a mixer input into which you can load audio data prior to playback; a source data line is a mixer input that accepts a real-time stream of audio data. The application preloads audio data from a sound file into the clips, and then pushes other audio data into the source data lines. The mixer reads data from these lines, each of which may have its own reverberation, gain, and pan controls, and uses the reverb settings to mix the "dry" audio signals with the reverberated ("wet") mix. The mixer delivers its final output to one or more output ports, such as a speaker, a headphone jack, and a line-out jack.

Although the various lines are depicted as separate rectangles in the diagram, they are all "owned" by the mixer, and can be considered integral parts of the mixer. The reverb, gain, and pan rectangles represent processing controls (rather than lines) that can be applied by the mixer to data flowing through the lines. (Note that this is just one example of a possible audio system that is supported by the API. Not all audio configurations will have all the features illustrated. An individual source data line might not support panning, a mixer might not implement reverb, and so on.)

A simple audio-input system might be similar:

Here, data flows in to the mixer from one or more input ports, commonly the microphone or the line-in jack. Gain and pan are applied, and the mixer delivers the captured data to an application via the mixer's target data line. A target data line is a mixer output, containing the mixture of the streamed input sounds. The simplest mixer has just one target data line, but some mixers can deliver captured data to multiple target data lines simultaneously.

[D]

The base interface, Line, describes the minimal functionality common to all lines:

The Mixer interface represents a hardware or software device that has one or more input lines and one or more output lines. This definition means that a mixer need not actually mix data; it might have only a single input. The Mixer API is intended to encompass a variety of devices, but the typical case supports mixing.

The Mixer interface provides methods for obtaining a mixer's lines. These can include target data lines, from which an application can read captured audio data, source data lines, to which an application can write audio data for playback (rendering), and clips, in which an application can preload sound data for playback. The mixer can lock these resources. For example, if the mixer has only one target data line and it is already in use, an attempt by an application to obtain a target data line will cause an exception to be thrown.

You can query a mixer for lines of different types, by passing the appropriate type of Line.Info. You can also ask the mixer how many lines of a particular type it supports.

A mixer maintains textual information about its specific device type in an inner class called Mixer.Info. This information include the product's name, version, and vendor, along with a textual description.

Notice that the generic Line interface does not provide a means to start and stop playback or recording. For that you need a data line. The DataLine interface supplies the following additional media-related features beyond those of a Line:

An application can obtain a data line from a mixer. If the data line cannot be allocated because of resource constraints (for example, if the mixer supports only one target data line and it is already in use), an exception is thrown.

A TargetDataLine receives audio data from a mixer. Commonly, the mixer has captured audio data from a port such as a microphone; it might process or mix this captured audio before placing the data in the target data line's buffer. The TargetDataLine interface provides methods for reading the data from the target data line's buffer and for determining how much data is currently available for reading. If an application attempts to read more data than is available, the read method blocks until the requested amount of data is available. This applies even if the amount of data requested is greater than the line's buffer size. The read method returns if the line is closed, paused, or flushed.

Applications recording audio should read data from the target data line fast enough to avoid overflow of the buffer, which results in discontinuities in the captured data. If the buffer does overflow, the oldest queued data is discarded and replaced by new data.

A SourceDataLine receives audio data for playback. It provides methods for writing data to the source data line's buffer for playback, and for determining how much data the line is prepared to receive without blocking. If an application attempts to write more data than is available, the read method blocks until the requested amount of data can be written. This applies even if the amount of data requested is greater than the line's buffer size. The write method also returns if the line is closed, paused, or flushed.

Applications playing audio should write data to the source data line fast enough to avoid underflow (emptying) of the buffer, which may result in discontinuities in audio playback. If audio playback stops due to underflow, a STOP event is generated. A START event is generated when presentation resumes.

A Clip is a data line into which audio data can be loaded prior to playback. Because the data is pre-loaded rather than streamed, the clip’s duration is known before playback, and you can choose any starting position in the media. Clips can be looped, meaning that upon playback, all the data between two specified loop points will repeat a specified number of times, or indefinitely.

Controls

Data lines and ports often have a set of controls that affect the audio signal passing through the line. The way in which the signal is affected depends on the type of control. The Java Sound API defines the following subclasses of Control:

• GainControl
  
  This control allows the signal's volume to be boosted or cut a specified number of decibels. A fade can be applied so that the volume increases or decreases gradually instead of instantly. The gain control includes a mute function to temporarily silence the line without altering its gain setting. A GainControl object can be queried for its resolution and for the minimum and maximum gain values it permits. The resolution is expressed as the number of increments over which the range of possible values is distributed.
  
  The types of gain control defined by the API include an overall gain control as well as gain controls for auxiliary send and return. The send and return typically control a line's contribution to a mixed and reverberated signal.
• PanControl
  
  A pan control affects the sound's right-left positioning. For a stereo audio signal, the pan control affects the balance between the volumes of the left and right channels. An application can set the pan position and can query the control for its resolution. The resolution is expressed as the number of incremental values from "hard left" to "hard right."
• ReverbControl
  
  This control gives access to any number of reverberation settings. Each setting might emulate a different time of room (small club, concert hall, cavern, etc.). The parameters used in constructing each setting consist of the delay time and intensity of early reflections, the delay time and intensity of late reflections, and an overall decay time.
• SampleRateControl
  
  This control affects the rate of playback. Altering the audio sample rate changes the sound's pitch. As with GainControl, the change can be made gradually instead of immediately, and the control can be queried for its resolution and its minimum and maximum possible values.

Programmatically, you obtain a particular control object from a line through a reference to the control’s class. You can also obtain an array of all the controls for that line.

AudioSystem

Here are some of the resources an application can obtain from the AudioSystem:

• Mixers
  
  A mixer can be obtained by passing a specific Mixer.Info object. You can learn what mixers are available by invoking the getMixerInfo method, which returns an array of Mixer.Info objects.
• Lines
  
  An application can get a target data line or source data line directly from the AudioSystem, without dealing explicitly with mixers.

• Format conversions
  
  

  An application can use format conversions to translate audio data from one format to another. (See the discussion of AudioFormat above.) Format conversions are often used to compress and decompress audio data. An application can query the AudioSystem to learn what translations are supported. It can then pass the AudioSystem a stream of audio data and get back a translated stream in a particular format.

• Files and streams
  
  

  An audio stream is an input stream with a specified audio data format (AudioFormat) and data length. The AudioInputStream class represents such a stream, from which you can read bytes. Some audio input streams permit you to remember positions in the stream and skip around in it. The AudioSystem class provides methods for translating between audio files and audio streams. The AudioSystem can also report the file format of a sound file and can write files in the different formats. A file format is represented by the AudioFileFormat class. An AudioFileFormat includes an AudioFormat, the file's length, and its type (WAV, AIFF, AU, etc.).

System configuration (SPI classes)

Service provider interfaces for the sampled audio system are defined in the javax.sound.sampled.spi package. Service providers can extend the classes defined here so that their own audio devices, sound file parsers and writers, and format converters can be installed and made available by a Java Sound implementation.

Interfaces describing MIDI event transport, synthesis, and sequencing are defined in the javax.sound.midi package. The major concepts used in the package are described in the sections below.

Functional Overview

The diagram below illustrates the functional relationships between the major components in a typical Java Sound MIDI configuration. (Java Sound permits a variety of devices to be installed and interconnected. The system shown here is just one possible scenario.) The flow of data between components is indicated by arrows. The data can be in a standard file format, or (as indicated by the key in the lower right corner of the diagram), it can be audio, raw MIDI bytes, or Java Sound's MidiEvent objects.

In this example, the application prepares a musical performance by loading a musical score that is stored as a Standard MIDI File on a disk (lower left corner of the diagram). Standard MIDI files contain tracks, each of which is a list of time-tagged MIDI events. This MIDI file is read into a Sequence object, whose data structure reflects the file. A Sequence contains a set of Track objects, each of which contains a set of MidiEvent objects. The Sequence is then "performed" by a Sequencer. A Sequencer performs its music by sending MidiEvents to some other device, such as an internal or external synthesizer.

As illustrated, MidiEvents must be translated into raw (non-time-tagged) MIDI before being sent through a MIDI output port to an external synthesizer. This conversion is accomplished by a a MIDI-output device called a StreamGenerator. Similarly, raw MIDI data coming into the computer from an external MIDI source is translated into MidiEvents by a StreamParser.

The internal synthesizer (the rectangle marked "Synthesizer" in the diagram) accepts MidiEvents directly from the Sequencer or StreamParser. It parses each event and usually dispatches a corresponding command (such as noteOn) to one of its MidiChannel objects, according to the MIDI channel number specified in the event. (The MIDI Specification calls for 16 MIDI channels, so a Synthesizer typically has 16 MidiChannel objects.)

The MidiChannel uses the note information in these messages to synthesize music. For example, a noteOn message specifies the note's pitch and "velocity" (volume). However, the note information is insufficient; the synthesizer also requires precise instructions on how to create the audio signal for each note. These instructions are represented by an Instrument. Each Instrument typically emulates a different real-world musical instrument or sound effect. The Instruments might come as presets with the synthesizer, or they might be loaded from soundbank files. In the synthesizer, the Instruments are arranged by bank number (the rows in the diagram) and program number (the columns). An Instrument can make use of stored digital audio, included as Sample objects in the soundbank. For example, to play the sound of a trumpet playing a 5-second-long note, the synthesizer might loop (cycle) through a half-second snippet of a recording of a trumpet.

Now that the components have been introduced from a functional perspective, we will take a brief look at the API from a programmatic perspective.

MidiEvent

A MidiEvent object specifies the type, data length, and status byte of the raw MIDI message for which it serves as a wrapper. In addition, it provides a tick value that is used by devices involved in MIDI timing, such as sequencers.

There are three categories of events, each represented by a MidiEvent subclass:

• ShortEvents are the most common and have at most two data bytes following the status byte.
• SysexEvents contain system-exclusive MIDI messages. They may have many bytes, and generally contain manufacturer-specific instructions.
• MetaEvents occur in MIDI files, but not in raw MIDI data streams. Meta events contain data, such as lyrics or tempo settings, that might be useful to sequencers but are usually meaningless for synthesizers.

MidiDevice

The base interface for devices is MidiDevice. All devices provide methods for listing the set of MIDI modes that they support, and for querying and setting the current mode. (The mode is a combination of MIDI's Omni mode and Mono/Poly mode.) Devices can be opened and closed, and they provide descriptions of themselves through a MidiDevice.Info object.

The following diagram illustrates the MidiDevice interface hierarchy. Also depicted are two classes, connected by dashed lines to the MidiDevice interfaces they implement.

Devices are generally either transmitters or receivers of MidiEvents. The Transmitter subinterface of MidiDevice includes methods for setting and querying the receivers to which the transmitter is sending MidiEvents. From the perspective of a transmitter, these receivers fall into two categories: MIDI Out and MIDI Thru. The transmitter sends events that it generates itself to its MIDI Out receivers. If the transmitter is itself also a receiver, it passes along events that it has received from elsewhere to its MIDI Thru receivers. The Receiver subinterface of MidiDevice consists of a single method for receiving MidiEvents. Typically, this method is invoked by a Transmitter.

Java Sound includes concrete classes for converting between MidiEvent objects and the raw byte stream used in MIDI wire protocol. A StreamGenerator is a Receiver that accepts MidiEvent objects from a Transmitter and writes out a raw MIDI byte stream. Similarly, a StreamParser is a Transmitter that accepts a raw MIDI byte stream and writes the corresponding MidiEvent objects to its Receiver.

Synthesis

A Synthesizer is a type of MidiDevice that generates sound. The Synthesizer interface extends both Receiver and Transmitter. It provides methods for manipulating soundbanks and instruments. In addition, it provides access to a set of MIDI channels through which sound is actually produced. A Synthesizer receives MidiEvents and invokes corresponding MidiChannel messages.

MidiChannels have methods representing the common MIDI voice messages such as "note on" and "control change." They also permit queries of the channel's current state.

Sequencing

Like Synthesizer, the Sequencer interface extends both Transmitter and Receiver (and therefore MidiDevice). Sequencer adds methods for basic MIDI sequencing operations. A sequencer can load and play back a sequence, query and set the tempo, and control the master and slave sync modes. An application can register to be notified when the sequencer processes MetaEvents and controller events. (A controller event occurs when a MIDI controller, such as a pitch-bend wheel or a data slider, changes its value. These events are not MidiEvents, but are created when the Sequencer encounters certain ShortEvents in the Sequence.)

The Sequence object represents a MIDI sequence as one or more tracks and associated timing information. A track contains a list of time-stamped MIDI events. Sequences can be read from MIDI files, or created from scratch and edited by adding Tracks to the Sequence (or removing them). Similarly, MidiEvents can be added to or removed from the Tracks.

It is not necessary to load a MIDI file into a Sequence object before playing the file. The setSequence(java.io.InputStream) method of Sequencer lets you read a MIDI file directly into a Sequencer, without creating a Sequence object first.

MidiSystem

MidiSystem acts as an application's entry point to the MIDI music system. It provides information about, and access to, the set of installed devices, including transmitters, receivers, synthesizers, and sequencers.

The MidiSystem class provides methods for reading MIDI files to create Sequence objects, and for writing Sequences to MIDI files. A MIDI Type 0 file contains only one track, while a Type 1 file may contain any number. MidiSystem also provides methods to create Soundbank objects by parsing soundbank files.

System configuration (SPI interfaces)

Configuration of the MIDI system is handled in the javax.sound.midi.spi package. The abstract classes in this package allow service providers to supply and install their own MIDI devices, MIDI file readers and writers, and soundbank file readers.