Rectangular window limitations in Signal Processing - Mini Project: Build & Apply

Rectangular Window Limitations in Signal Processing

📖 Scenario: Imagine you are analyzing a short sound clip to find its main frequencies. You use a rectangular window to cut the clip into a segment before analyzing it. This project will help you see how the rectangular window affects the frequency analysis.

🎯 Goal: You will create a simple signal, apply a rectangular window, compute its frequency spectrum using the Fast Fourier Transform (FFT), and observe the limitations of the rectangular window by plotting the results.

📋 What You'll Learn

Create a simple sine wave signal with a known frequency

Create a rectangular window of the same length as the signal

Apply the rectangular window to the signal

Compute the FFT of the windowed signal

Plot the magnitude spectrum to observe the rectangular window effects

💡 Why This Matters

🌍 Real World

Windowing is used in audio processing, communications, and radar to analyze signals in small segments.

💼 Career

Understanding window effects helps data scientists and engineers improve signal analysis accuracy in real-world applications.

1

Create a sine wave signal

Create a numpy array called t with 500 points linearly spaced between 0 and 1 second. Then create a sine wave signal called signal with frequency 5 Hz using t.

Signal Processing

import numpy as np
# Create time array t with 500 points from 0 to 1
# Create sine wave signal with frequency 5 Hz
# Your code here

Hint

Use np.linspace(0, 1, 500) to create t. Use np.sin(2 * np.pi * 5 * t) for the sine wave.

2

Create a rectangular window

Create a numpy array called rect_window of length 500 filled with ones to represent the rectangular window.

Signal Processing

import numpy as np

t = np.linspace(0, 1, 500)
signal = np.sin(2 * np.pi * 5 * t)
# Create rectangular window of length 500
# Your code here

Hint

Use np.ones(500) to create the rectangular window.

3

Apply the rectangular window and compute FFT

Multiply signal by rect_window and store in windowed_signal. Then compute the FFT of windowed_signal using np.fft.fft and store it in fft_result.

Signal Processing

import numpy as np

t = np.linspace(0, 1, 500)
signal = np.sin(2 * np.pi * 5 * t)

rect_window = np.ones(500)
# Multiply signal by rectangular window
# Compute FFT of windowed signal
# Your code here

Hint

Multiply arrays element-wise with *. Use np.fft.fft() to get the frequency spectrum.

4

Plot the magnitude spectrum

Import matplotlib.pyplot as plt. Compute the magnitude spectrum by taking the absolute value of fft_result and store it in magnitude. Plot the first 50 points of magnitude using plt.plot(). Label the x-axis as 'Frequency bins' and y-axis as 'Magnitude'. Finally, call plt.show() to display the plot.

Signal Processing

import numpy as np
import matplotlib.pyplot as plt

t = np.linspace(0, 1, 500)
signal = np.sin(2 * np.pi * 5 * t)

rect_window = np.ones(500)

windowed_signal = signal * rect_window
fft_result = np.fft.fft(windowed_signal)
# Compute magnitude spectrum
# Plot first 50 points
# Label axes
# Show plot
# Your code here

Hint

Use np.abs() for magnitude. Use plt.plot(), plt.xlabel(), plt.ylabel(), and plt.show().

Practice

1. What is a main limitation of using a rectangular window in signal processing?

easy

A. It causes frequency spreading and high side lobes

B. It completely removes noise from the signal

C. It increases the signal amplitude

D. It shortens the signal duration

Solution

1. Step 1: Understand the rectangular window effect

   The rectangular window is a simple window that cuts the signal abruptly, causing spectral leakage.

2. Step 2: Identify the impact on frequency analysis

   This abrupt cut causes frequency spreading and high side lobes, making it hard to distinguish close frequencies.

3. Final Answer:

   It causes frequency spreading and high side lobes -> Option A

4. Quick Check:

   Rectangular window = frequency spreading and side lobes [OK]

Hint: Rectangular window causes side lobes and spreading [OK]

Common Mistakes:

• Thinking it removes noise completely
• Believing it increases signal amplitude
• Assuming it shortens signal duration

2. Which of the following is the correct way to apply a rectangular window to a signal array sig in Python using NumPy?

easy

A. windowed = sig + np.zeros(len(sig))

B. windowed = sig * np.ones(len(sig))

C. windowed = sig / np.ones(len(sig))

D. windowed = sig - np.ones(len(sig))

Solution

1. Step 1: Recall rectangular window definition

   A rectangular window is an array of ones with the same length as the signal.

2. Step 2: Apply window by element-wise multiplication

   Multiplying the signal by an array of ones keeps the signal unchanged but applies the window.

3. Final Answer:

   windowed = sig * np.ones(len(sig)) -> Option B

4. Quick Check:

   Rectangular window = multiply by ones array [OK]

Hint: Multiply signal by ones array for rectangular window [OK]

Common Mistakes:

• Adding or subtracting ones instead of multiplying
• Dividing by ones which is unnecessary
• Using zeros which nullifies the signal

3. Given the following Python code, what will be the approximate side lobe level in dB when using a rectangular window on a signal?

import numpy as np
N = 64
rect_win = np.ones(N)
# Side lobe level for rectangular window is about -13 dB

medium

A. -3 dB

B. -60 dB

C. -40 dB

D. -13 dB

Solution

1. Step 1: Recall typical side lobe level for rectangular window

   The rectangular window has a main lobe and side lobes around -13 dB relative to the main lobe.

2. Step 2: Match the known value to options

   Among the options, -13 dB matches the known side lobe level for rectangular windows.

3. Final Answer:

   -13 dB -> Option D

4. Quick Check:

   Rectangular window side lobes ≈ -13 dB [OK]

Hint: Rectangular window side lobes near -13 dB [OK]

Common Mistakes:

• Confusing side lobe level with main lobe width
• Choosing much lower side lobe levels like -40 dB
• Assuming side lobes are negligible

4. You have this code snippet to apply a rectangular window but it causes unexpected frequency leakage:

import numpy as np
signal = np.sin(2 * np.pi * 5 * np.arange(100) / 100)
window = np.ones(50)
windowed_signal = signal * window

What is the main error?

medium

A. Window length does not match signal length

B. Window contains zeros instead of ones

C. Signal frequency is too high for window

D. Multiplication should be addition

Solution

1. Step 1: Check window and signal length

   The signal length is 100, but the window length is only 50, causing a length mismatch.

2. Step 2: Understand effect of length mismatch

   Multiplying arrays of different lengths causes broadcasting or unintended results, leading to frequency leakage.

3. Final Answer:

   Window length does not match signal length -> Option A

4. Quick Check:

   Window length must equal signal length [OK]

Hint: Window length must match signal length [OK]

Common Mistakes:

• Ignoring length mismatch
• Thinking zeros in window cause leakage
• Confusing multiplication with addition

5. You want to analyze two close frequencies in a signal but the rectangular window causes them to blur together. Which approach best reduces this limitation?

hard

A. Apply rectangular window twice in a row

B. Increase the signal amplitude before applying rectangular window

C. Use a window with lower side lobes like Hamming or Hann

D. Shorten the signal length to reduce frequency resolution

Solution

1. Step 1: Identify the problem with rectangular window

   Rectangular window has high side lobes causing frequency spreading, blurring close frequencies.

2. Step 2: Choose a better window to reduce side lobes

   Windows like Hamming or Hann have lower side lobes, improving frequency resolution and reducing blur.

3. Final Answer:

   Use a window with lower side lobes like Hamming or Hann -> Option C

4. Quick Check:

   Lower side lobes = clearer frequency separation [OK]

Hint: Choose low side lobe windows to separate close frequencies [OK]

Common Mistakes:

• Increasing amplitude does not fix side lobes
• Shortening signal reduces resolution, worsening problem
• Applying rectangular window twice has no benefit