A Little About Digital Audio and Linear PCM

A Little About Digital Audio and Linear PCM

Most Core Audio services use and manipulate audio in linear pulse-code-modulated (linear PCM) format, the most common uncompressed digital audio data format. Digital audio recording creates PCM data by measuring an analog (real world) audio signal’s magnitude at regular intervals (the sampling rate) and converting each sample to a numerical value. Standard compact disc (CD) audio uses a sampling rate of 44.1 kHz, with a 16-bit integer describing each sample—constituting the resolution or bit depth.

• A sample is single numerical value for a single channel.

• A frame is a collection of time-coincident samples. For instance, a stereo sound file has two samples per frame, one for the left channel and one for the right channel.

• A packet is a collection of one or more contiguous frames. In linear PCM audio, a packet is always a single frame. In compressed formats, it is typically more. A packet defines the smallest meaningful set of frames for a given audio data format.

In linear PCM audio, a sample value varies linearly with the amplitude of the original signal that it represents. For example, the 16-bit integer samples in standard CD audio allow 65,536 possible values between silence and maximum level. The difference in amplitude from one digital value to the next is always the same.

Core Audio data structures, declared in the CoreAudioTypes.h header file, can describe linear PCM at any sample rate and bit depth. “Audio Data Formats” goes into more detail on this topic.

In Mac OS X, Core Audio expects audio data to be in native-endian, 32-bit floating-point, linear PCM format. You can use Audio Converter Services to translate audio data between different linear PCM variants. You also use these converters to translate between linear PCM and compressed audio formats such as MP3 and Apple Lossless. Core Audio in Mac OS X supplies codecs to translate most common digital audio formats (though it does not supply an encoder for converting to MP3).

iOS uses integer and fixed-point audio data. The result is faster calculations and less battery drain when processing audio. iOS provides a Converter audio unit and includes the interfaces from Audio Converter Services. For details on the so-called canonical audio data formats for iOS and Mac OS X, see “Canonical Audio Data Formats.”

In iOS and Mac OS X, Core Audio supports most common file formats for storing and playing audio data, as described in “iPhone Audio File Formats” and