Create a datastore that maintains parity between the audio of the underlying datastores. Create two separate audio datastores, and then create a combined datastore representing the two underlying datastores.

Create a datastore ads1 that points to the audio files included with Audio Toolbox.

folder = fullfile(matlabroot,'toolbox','audio','samples');
ads1 = audioDatastore(folder);

Create a second datastore ads2 by adding noise to the audio in the ads1.

ads2 = transform(ads1,@(x) x + 0.01*randn(size(x)) );

Create a combined datastore from ads1 and ads2.

adsCombined = combine(ads1,ads2);

Read the first pair of audio files from the combined datastore. Each read operation on this combined datastore returns a pair of audio signals in a 1-by-2 cell array and a pair of info structs in a 1-by-2 cell array.

[dataOut,infoOut] = read(adsCombined)

dataOut=1×2 cell array
    {539648x1 double}    {539648x1 double}

infoOut=1×2 cell array
    {1x1 struct}    {1x1 struct}

Plot the spectrograms of the first channels from both audio signals.

figure(1)
spectrogram(dataOut{1},hamming(512),256,512,infoOut{1}.SampleRate,'yaxis')
title('Original Data')

figure(2)
idx = size(dataOut,2)/2+1;
spectrogram(dataOut{2},hamming(512),256,512,infoOut{2}.SampleRate,'yaxis')
title('Noised Data')