Loading .wav Files in Python

This repository brings some tips on loading .wav files in Python.

Tip #1 -- Choose the library

In this case, you can see different ways to open wav files.

• Access the loading_wav_files.ipynb jupyter notebook.

In this example, you can see how to use 4 different libraries.

Libraries:

• SoundFile [https://pysoundfile.readthedocs.io/en/latest/]
• Scipy [https://docs.scipy.org/doc/scipy/reference/generated/scipy.io.wavfile.read.html]
• Librosa [https://librosa.org/doc/latest/generated/librosa.load.html#librosa.load]
• Parselmouth [https://parselmouth.readthedocs.io/en/stable/index.html]

Libraries notes:

• Soundfile, Librosa and Parselmouth libraries have the input signal in the range [-1, 1]. On the other hand, Scipy library has the input signal in the range according to the digitalization level (e.g., 16 bits). Thus, for Scipy, we can define a normalization method (see the aforementioned jupyter notebook).
• For the Librosa library, if you do not specify the sampling rate, it will consider sr = 22050 Hz.

Tip #2

In this case, you are interested in asking the user to put the file name.

• Access the loading1.py file.

In this example, you can see four different ways to open the input file.

Tip #3

In this case, you are interested in loading a sequence of files with a specific prefix word.

• Access the loading2.py file and/or the loading2.ipynb jupyter notebook.

In this example, you can choose whether you or the user put the file name and the amount of files.

Tip #4

In this case, you are interested in loading any .wav file from a folder.

• Access the loading3.py file and/or the loading3.ipynb jupyter notebook.

Repository Notes

All .wav files in this repository are just examples of usage. You can apply to your own files.

Acknowledgments

I would like to thank the developers of all the aforementioned libraries. It is important to note that all these libraries have a lot of other interesting functions. Feel free to access and explore them.