Overview

portaudio wraps PortAudio — the cross-platform audio I/O library used by Audacity, PyAudio, and many DAW applications. It provides low-latency access to microphone input and speaker output with full control over sample rate, channels, and buffer size.

Quick Start — Record & Play

uses portaudio;

Initialize;
PrintDevices;  // list all audio devices

// Record 5 seconds from default microphone (16kHz, mono, INT16)
var pcm := Record(5.0);
WriteLn('Recorded ', Length(pcm), ' bytes');

// Play it back
Play(pcm, PA_RATE_16000, 1, PA_INT16);

Terminate;

Record to WAV File

Initialize;

// Record 10 seconds → WAV file
RecordToFile('/recordings/interview.wav', 10.0, PA_RATE_44100, 2);

// Or record raw PCM and save manually
var pcm := RecordEx(5.0, PA_DEFAULT_INPUT, PA_RATE_16000, 1, PA_INT16);
SaveWAV('/out/voice.wav', pcm, PA_RATE_16000, 1, 16);

Terminate;

Play WAV File

Initialize;

// Simple playback — blocks until done
PlayFile('/audio/notification.wav');

// Play on a specific output device
var devs := ListDevices;
PlayFileEx('/audio/music.wav', devs[2].Index);

Terminate;

Load & Process WAV

var rate, ch, bits: Integer;
var pcm := LoadWAV('/audio/sample.wav', rate, ch, bits);
WriteLn('Rate: ', rate, 'Hz  Channels: ', ch, '  Bits: ', bits);
WriteLn('Duration: ', Length(pcm) / (rate * ch * (bits div 8)):0:2, 's');

// Process PCM bytes here...

SaveWAV('/audio/processed.wav', pcm, rate, ch, bits);

Streaming API (low-latency)

Initialize;

// Open an input stream — 256 frames per buffer for low latency
var stream := OpenInputStream(PA_DEFAULT_INPUT, PA_RATE_16000, 1, PA_INT16, 256);
StartStream(stream);

var chunks: array of string;
var startTime := StreamTime(stream);

// Read in real-time for 3 seconds
while StreamTime(stream) - startTime < 3.0 do begin
  var chunk := ReadStream(stream, 256);
  chunks.Add(chunk);  // accumulate chunks
end;

StopStream(stream);
CloseStream(stream);

// Join chunks and save
var allPCM := String.Join('', chunks);
SaveWAV('/out/realtime.wav', allPCM, PA_RATE_16000, 1, 16);

Terminate;

Device Enumeration

Initialize;
var devs := ListDevices;
for I := 0 to Length(devs) - 1 do begin
  var d := devs[I];
  WriteLn('[', d.Index, '] ', d.Name);
  WriteLn('  In: ', d.MaxInputChannels, ' ch  Out: ', d.MaxOutputChannels, ' ch');
  WriteLn('  Default rate: ', d.DefaultSampleRate:0:0, ' Hz');
  if d.IsDefaultInput  then WriteLn('  -> Default Input');
  if d.IsDefaultOutput then WriteLn('  -> Default Output');
end;
Terminate;

Sample Rates & Formats

// Common sample rates
PA_RATE_8000   // telephone, speech
PA_RATE_16000  // speech recognition (Whisper, etc.)
PA_RATE_44100  // CD quality
PA_RATE_48000  // professional audio

// Sample formats
PA_INT16   // 16-bit PCM — most compatible
PA_FLOAT32 // 32-bit float — best for DSP
PA_INT32   // 32-bit integer