spectrogram of an .m4a file

To solve my main problem, I have recorded some .m4a audio files (sample). I want to first get an spectrogram like this:

Generated by Audacity.

From here I can import the file:

from pydub import AudioSegment
seg = AudioSegment.from_file("voice.m4a")
print("Information:")
print("Channels:", seg.channels)
print("Bits per sample:", seg.sample_width * 8)
print("Sampling frequency:", seg.frame_rate)
print("Length:", seg.duration_seconds, "seconds")

and I know these spectrogram ploting functions:

scipy.signal.spectrogram
matplotlib.pyplot.specgram

for which there are a lot of examples here. But I don't know the steps between. I would appreciate if you could help me know how should I get the above plot. It doesn't have to be PyBud. Any other library (or even other language as far as it is Free Software!) is also fine. Thansk for your support in advance.

P.S. Next step would be to analyse the ambient noise in real-time like the one here. So any help in that direction would also be highly appreciated.

Solution

Here is an example straight from the documentation (see page 10):

# Example for plotting a spectrogram using this function
import audiosegment
import matplotlib.pyplot as plt
#...
seg = audiosegment.from_file("voice.m4a")
freqs, times, amplitudes = seg.spectrogram(window_length_s=0.03, overlap=0.5)
amplitudes = 10 * np.log10(amplitudes + 1e-9)
# Plot
plt.pcolormesh(times, freqs, amplitudes)
plt.xlabel("Time in Seconds")
plt.ylabel("Frequency in Hz")
plt.show()

You need to have [AudioSegment library] installed (e.g. pip install AudioSegment)
FFmpeg library needs to also be installed. On Windows one can use Chocolatey: choco install ffmpeg