audioresample

Sample-rate conversion of audio signals

Since R2023b

collapse all in page

    Syntax

    audioOut = audioresample(audioIn,InputRate=inputFs,OutputRate=outputFs)
    audioOut = audioresample(audioIn,InputRate=inputFs,OutputRate=outputFs,Quality=q)
    audioOut = audioresample(audioIn,ResamplerSource="custom",Resampler=customResampler)
    [audioOut,resampler] = audioresample(___)

    Description

    audioOut = audioresample(audioIn,InputRate=inputFs,OutputRate=outputFs) filters and resamples the audio signal to the desired sample rate.

    example

    audioOut = audioresample(audioIn,InputRate=inputFs,OutputRate=outputFs,Quality=q) specifies the resampling quality.

    audioOut = audioresample(audioIn,ResamplerSource="custom",Resampler=customResampler) resamples the audio signal with the provided custom resampler.

    example

    [audioOut,resampler] = audioresample(___) also returns the resampler used by the function.

    example

    Examples

    collapse all

    Open Live Script

    Read in an audio signal containing speech. See the sample rate is 22,050 Hz.

    [x,fs] = audioread("speech_dft.wav");
fs
    fs = 
22050

    Listen to the audio.

    sound(x,fs)

    Use audioresample to resample the audio to have a sample rate of 8000 Hz.

    outputFs = 8000;
y = audioresample(x,InputRate=fs,OutputRate=outputFs);

    Listen to the resampled audio.

    sound(y,outputFs)
    Open Live Script

    Read in an audio signal sampled at 96 kHz and convert it to single precision.

    [x,fs] = audioread("RandomOscThree-24-96-stereo-13secs.aif");
x = single(x);

    Use designAudioResampler to design a multistage resampler for single-precision signals. The resampler converts audio from 96 kHz to 44.1 kHz.

    outputFs = 44100;
resampler = designAudioResampler(InputRate=fs,OutputRate=outputFs, ...
    DesignMethod="multistage",DataType="single");

    Resample the audio with the custom resampler.

    y = audioresample(x,ResamplerSource="custom",Resampler=resampler);
    Open Live Script

    Create a list of audio files to resample. Each of the files has a sample rate of 44.1 kHz.

    audioFiles = ["Ambiance-16-44p1-mono-12secs.wav", ...
              "Counting-16-44p1-mono-15secs.wav", ...
              "TrainWhistle-16-44p1-mono-9secs.wav"];

    Design a resampler to resample the audio signals to 48 kHz. This saves computation by not redesigning the resampler for each signal.

    inputFs = 44100;
outputFs = 48000;
resampler = designAudioResampler(InputRate=inputFs,OutputRate=outputFs);

    In a loop, read in the audio file, resample the signal, and store it in a cell array.

    resampledSignals = cell(size(audioFiles));
for ii = 1:numel(audioFiles)
    [x,fs] = audioread(audioFiles(ii));
    resampledSignals{ii} = audioresample(x,ResamplerSource="custom",Resampler=resampler);
    reset(resampler)
end
    Open Live Script

    Read in an audio signal with a 96 kHz sample rate.

    [x,fs] = audioread("RandomOscThree-24-96-stereo-13secs.aif");

    Use audioresample to resample the audio to 44.1 kHz. Specify an additional output argument to get the resampler object that audioresample used.

    [y,resampler] = audioresample(x,InputRate=fs,OutputRate=44100);
resampler
    resampler = 
  dsp.FIRRateConverter with properties:

   Main
    InterpolationFactor: 147
       DecimationFactor: 320
        NumeratorSource: 'Property'
              Numerator: [4.2508e-08 4.0303e-08 3.8039e-08 3.5716e-08 3.3333e-08 3.0891e-08 2.8388e-08 2.5825e-08 2.3202e-08 2.0517e-08 1.7772e-08 1.4965e-08 1.2096e-08 9.1655e-09 6.1727e-09 3.1176e-09 0 -3.1804e-09 -6.4237e-09 … ] (1×14113 double)

  Show all properties
    Open Live Script

    Design a resampler to resample audio from 44.1 kHz to 16 kHz.

    inputFs = 44100;
outputFs = 16000;
resampler = designAudioResampler(InputRate=inputFs,OutputRate=outputFs);

    Create dsp.AudioFileReader and audioDeviceWriter objects to read in the audio and write the resampled signal to your audio device to listen to it.

    samplesPerFrame = 1024;
reader = dsp.AudioFileReader( ...
    Filename="Counting-16-44p1-mono-15secs.wav", ...
    SamplesPerFrame=samplesPerFrame);
deviceWriter = audioDeviceWriter(SampleRate=outputFs,SupportVariableSizeInput=true);

    In a streaming loop, read in an audio frame, resample the signal, and play the resampled signal on your device.

    while ~isDone(reader)
    audioIn = reader();
    audioOut = audioresample(audioIn,ResamplerSource="custom",Resampler=resampler);
    deviceWriter(audioOut);
end

    Input Arguments

    collapse all

    Audio input, specified as a column vector or matrix. If the input is a matrix, audioresample treats the columns as independent channels.

    Data Types: single | double

    Sample rate of the input signal in Hz, specified as one of the following values.

    • 8000

    • 16000

    • 32000

    • 11025

    • 22050

    • 44100

    • 88200

    • 176400

    • 352800

    • 705600

    • 12000

    • 24000

    • 48000

    • 96000

    • 192000

    • 384000

    • 768000

    The audioresample function does not use this value if you specify customResampler.

    Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64

    Sample rate of the output signal in Hz, specified as one of the following values.

    • 8000

    • 16000

    • 32000

    • 11025

    • 22050

    • 44100

    • 88200

    • 176400

    • 352800

    • 705600

    • 12000

    • 24000

    • 48000

    • 96000

    • 192000

    • 384000

    • 768000

    The audioresample function does not use this value if you specify customResampler.

    Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64

    Custom resampler to be used by audioresample, specified as one of the following objects.

    You can design the custom resampler using designAudioResampler.

    Resampling quality, specified as "medium", "high", or "ultra". Higher quality resampling requires more computation.

    • "medium" — audioresample attenuates spectral aliases and spectral images by 96 dB and preserves about 90% of the bandwidth of interest.

    • "high" — audioresample attenuates spectral aliases and spectral images by 120 dB and preserves about 94% of the bandwidth of interest.

    • "ultra" — audioresample attenuates spectral aliases and spectral images by 144 dB and preserves about 98% of the bandwidth of interest.

    The audioresample function does not use this value if you specify customResampler.

    Data Types: char | string

    Output Arguments

    collapse all

    Resampled audio signal, returned as a column vector or matrix with the same number of channels as the input audioIn. The length of the input signal and the output sample rate determine the length of the output signal.

    Resampler object used by audioresample, returned as one of the following.

    The returned object is the same as customResampler if you specify it.

    Version History

    Introduced in R2023b

    See Also

    | |

    Topics