How do I select one audio channel from a multichannel array in librosa?

Jun 17 '21 Comments: 1 Answers: 0

0

I am new to librosa (and python in general) and I am getting started with audio analysis. My starting wav files contain multichannel arrays. I am wondering if I can use librosa to choose a single, specific channel. As of now, I convert my files to mono as shown here:

…

Open Full Question

Hi all!
I am new to librosa (and python in general) and I am getting started with audio analysis. My starting wav files contain multichannel arrays. I am wondering if I can use librosa to choose a single, specific channel. As of now, I convert my files to mono as shown here:

import librosa

y, sr = librosa.load(file, sr=None, mono=True)

However, I suspect that mono=True averages the channels instead of selecting individually. Please correct me if I am wrong! I know that SciPy allows individual channel selection:

from scipy.io import wavfile

sample_rate, data = wavfile.read(file)
#[:,1] slices the channel array
spectrum, freqs, t, im = plt.specgram(data[:,1], Fs=sample_rate))

How can I slice and select an individual channel like that in librosa?

Many thanks! Much help is greatly appreciated!