Spectrogram generation in SciPy - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a spectrogram?

A spectrogram is a visual representation of how the frequency content of a signal changes over time. It shows time on one axis, frequency on another, and the intensity or power of frequencies as colors or brightness.

beginner

Which function in scipy.signal is commonly used to generate a spectrogram?

The function scipy.signal.spectrogram is used to compute the spectrogram of a signal. It returns the frequencies, times, and the spectrogram data (power or intensity).

intermediate

What are the main inputs to scipy.signal.spectrogram?

The main inputs are the signal array and the sampling frequency (fs). Optional inputs include window type, segment length (nperseg), overlap (noverlap), and FFT length (nfft).

intermediate

How does changing the segment length (nperseg) affect the spectrogram?

A larger segment length gives better frequency resolution but worse time resolution. A smaller segment length gives better time resolution but poorer frequency resolution. It's a trade-off.

beginner

What does the output matrix of scipy.signal.spectrogram represent?

The output matrix contains the power spectral density values for each frequency and time segment. Each element shows how strong a frequency is at a specific time.

Which axis in a spectrogram represents time?

A None of the above

B Horizontal axis

C Color intensity

D Vertical axis

In a spectrogram, time is shown on the horizontal axis, frequency on the vertical axis, and intensity by color or brightness.

What does the fs parameter in scipy.signal.spectrogram specify?

A Signal amplitude

B Window type

C Sampling frequency

D Segment length

The fs parameter specifies the sampling frequency of the input signal, which is needed to correctly calculate frequencies.

Increasing nperseg in spectrogram calculation will:

A Improve frequency resolution

B Improve time resolution

C Decrease frequency resolution

D Have no effect

A larger segment length (nperseg) improves frequency resolution but reduces time resolution.

What does the output 'Sxx' from scipy.signal.spectrogram represent?

A Time values

B Signal amplitude

C Frequency values

D Power spectral density

'Sxx' is the matrix of power spectral density values for each frequency and time segment.

Which of these is NOT a parameter of scipy.signal.spectrogram?

A learning_rate

B noverlap

C nfft

D window

'learning_rate' is not a parameter of spectrogram; it is unrelated to signal processing.

Explain how a spectrogram helps us understand a signal's frequency changes over time.

Think about how music or speech changes pitch and loudness over time.

Describe the trade-off between time and frequency resolution when choosing segment length in spectrogram generation.

Consider how zooming in on time affects frequency detail and vice versa.