Audio equalization in Signal Processing

Introduction

Audio equalization helps adjust sound to make it clearer or more pleasant. It changes the balance of different sound frequencies.

When you want to make music sound better on headphones or speakers.

To reduce background noise or unwanted sounds in a recording.

When mixing audio tracks to balance instruments and vocals.

To improve speech clarity in podcasts or calls.

When customizing sound for different environments, like a car or room.

Syntax

Signal Processing

equalized_audio = apply_equalizer(audio_signal, frequency_bands, gain_values)

audio_signal is the input sound data.

frequency_bands are the ranges of sound frequencies to adjust.

Examples

This raises bass frequencies by 5 dB, lowers mid frequencies by 3 dB, and slightly raises treble frequencies by 2 dB.

Signal Processing

equalized_audio = apply_equalizer(my_audio, [100, 1000, 5000], [5, -3, 2])

This applies no change, keeping the original sound.

Signal Processing

equalized_audio = apply_equalizer(my_audio, [60, 250, 4000], [0, 0, 0])

Sample Program

This program creates a simple audio signal with bass, mid, and treble tones. It then applies an equalizer that boosts bass and treble frequencies and reduces mid frequencies. Finally, it plots the original and equalized signals to see the effect.

Signal Processing

import numpy as np
import matplotlib.pyplot as plt
from scipy.signal import butter, lfilter

def butter_bandpass(lowcut, highcut, fs, order=4):
    nyq = 0.5 * fs
    low = lowcut / nyq
    high = highcut / nyq
    b, a = butter(order, [low, high], btype='band')
    return b, a

def apply_bandpass_filter(data, lowcut, highcut, fs, order=4):
    b, a = butter_bandpass(lowcut, highcut, fs, order=order)
    y = lfilter(b, a, data)
    return y

def apply_equalizer(audio, fs, bands, gains):
    equalized = np.zeros_like(audio)
    for (lowcut, highcut), gain in zip(bands, gains):
        filtered = apply_bandpass_filter(audio, lowcut, highcut, fs)
        equalized += filtered * (10 ** (gain / 20))
    return equalized

# Create a sample audio signal: mix of 3 sine waves
fs = 44100  # sample rate
T = 1.0     # seconds
t = np.linspace(0, T, int(fs*T), endpoint=False)

# Frequencies: bass=100Hz, mid=1000Hz, treble=5000Hz
audio = 0.5*np.sin(2*np.pi*100*t) + 0.3*np.sin(2*np.pi*1000*t) + 0.2*np.sin(2*np.pi*5000*t)

# Define bands and gains (in dB)
bands = [(80, 120), (900, 1100), (4800, 5200)]
gains = [6, -3, 4]  # boost bass and treble, reduce mid

# Apply equalizer
equalized_audio = apply_equalizer(audio, fs, bands, gains)

# Plot original and equalized signals (first 0.01 seconds)
plt.figure(figsize=(10,4))
plt.plot(t[:441], audio[:441], label='Original')
plt.plot(t[:441], equalized_audio[:441], label='Equalized')
plt.legend()
plt.title('Audio Signal Before and After Equalization')
plt.xlabel('Time [seconds]')
plt.ylabel('Amplitude')
plt.show()

Important Notes

Equalization changes volume of specific frequency ranges, not the whole sound.

Boosting frequencies too much can cause distortion or noise.

Use plots or listening tests to check if equalization sounds good.

Summary

Audio equalization adjusts sound by changing frequency levels.

It helps improve clarity, reduce noise, or customize sound.

Simple filters can boost or cut bass, mid, and treble frequencies.

Practice

1. What is the main purpose of audio equalization in signal processing?

easy

A. To convert audio signals into text

B. To increase the volume of the audio signal only

C. To adjust the balance of frequency components in an audio signal

D. To remove all background noise completely

Solution

1. Step 1: Understand audio equalization

   Audio equalization changes the levels of different frequency parts of sound to improve or customize it.

2. Step 2: Identify the main function

   It adjusts bass, mid, and treble frequencies, not just volume or noise removal.

3. Final Answer:

   To adjust the balance of frequency components in an audio signal -> Option C

4. Quick Check:

   Audio equalization = frequency adjustment [OK]

Hint: Think: equalization means balancing frequencies [OK]

Common Mistakes:

• Confusing equalization with volume control
• Assuming it converts audio to text
• Believing it removes all noise

2. Which of the following is the correct way to represent a simple bass boost filter gain in code?

easy

A. gain = 1.5 # Boost bass frequencies by 50%

B. gain = -1.5 # Boost bass frequencies by 50%

C. gain = 0 # No change to bass frequencies

D. gain = 'boost' # String value for gain

Solution

1. Step 1: Understand gain values in filters

   Positive gain values increase (boost) frequency levels; negative values reduce (cut) them.

2. Step 2: Identify correct gain for bass boost

   Gain of 1.5 means 50% increase, which is a boost; negative or zero values do not boost.

3. Final Answer:

   gain = 1.5 # Boost bass frequencies by 50% -> Option A

4. Quick Check:

   Positive gain boosts frequencies [OK]

Hint: Positive gain means boost, negative means cut [OK]

Common Mistakes:

• Using negative gain to boost frequencies
• Setting gain to zero expecting boost
• Using non-numeric gain values

3. Given this Python code snippet for equalizing audio frequencies, what will be the output?

frequencies = {'bass': 100, 'mid': 500, 'treble': 10000}
gain = {'bass': 1.2, 'mid': 1.0, 'treble': 0.8}
equalized = {k: frequencies[k] * gain[k] for k in frequencies}
print(equalized)

medium

A. {'bass': 120.0, 'mid': 500.0, 'treble': 8000.0}

B. {'bass': 100, 'mid': 500, 'treble': 10000}

C. {'bass': 1.2, 'mid': 1.0, 'treble': 0.8}

D. SyntaxError due to dictionary comprehension

Solution

1. Step 1: Multiply each frequency by its gain

   bass: 100 * 1.2 = 120.0, mid: 500 * 1.0 = 500.0, treble: 10000 * 0.8 = 8000.0

2. Step 2: Construct the equalized dictionary

   Result is {'bass': 120.0, 'mid': 500.0, 'treble': 8000.0}

3. Final Answer:

   {'bass': 120.0, 'mid': 500.0, 'treble': 8000.0} -> Option A

4. Quick Check:

   frequency * gain = equalized value [OK]

Hint: Multiply each frequency by its gain [OK]

Common Mistakes:

• Confusing keys and values in comprehension
• Expecting original frequencies without gain
• Thinking dictionary comprehension causes syntax error

4. Identify the error in this code snippet for applying a treble cut filter:

frequencies = {'bass': 100, 'mid': 500, 'treble': 10000}
gain = {'bass': 1.0, 'mid': 1.0, 'treble': 0.5}
for freq in frequencies:
    frequencies[freq] = frequencies[freq] * gain['freq']
print(frequencies)

medium

A. Multiplying frequencies by gain is incorrect

B. The print statement is missing parentheses

C. The for loop should iterate over gain, not frequencies

D. Using 'freq' as a string key instead of variable freq in gain dictionary

Solution

1. Step 1: Check dictionary key usage in loop

   The code uses gain['freq'], which looks for key 'freq' literally, not the variable freq.

2. Step 2: Correct key access

   It should be gain[freq] without quotes to use the variable's value as key.

3. Final Answer:

   Using 'freq' as a string key instead of variable freq in gain dictionary -> Option D

4. Quick Check:

   Variable keys need no quotes [OK]

Hint: Use variable keys without quotes in dictionaries [OK]

Common Mistakes:

• Using string keys instead of variables
• Thinking multiplication is wrong
• Confusing loop iteration variable

5. You have an audio signal with frequency bands: bass=80Hz, mid=1000Hz, treble=12000Hz. You want to reduce bass by 20%, keep mid unchanged, and boost treble by 30%. Which gain dictionary correctly applies this equalization?

hard

A. {'bass': 1.2, 'mid': 1.0, 'treble': 0.7}

B. {'bass': 0.8, 'mid': 1.0, 'treble': 1.3}

C. {'bass': 0.2, 'mid': 1.3, 'treble': 1.0}

D. {'bass': 0.8, 'mid': 0.7, 'treble': 1.3}

Solution

1. Step 1: Calculate gain values for each band

   Reduce bass by 20% means multiply by 0.8; mid unchanged means 1.0; boost treble by 30% means multiply by 1.3.

2. Step 2: Match gains to options

   {'bass': 0.8, 'mid': 1.0, 'treble': 1.3} matches these values exactly; others do not.

3. Final Answer:

   {'bass': 0.8, 'mid': 1.0, 'treble': 1.3} -> Option B

4. Quick Check:

   Gain = 1 ± percentage change [OK]

Hint: Gain = 1 ± percentage change (0.8 means 20% cut) [OK]

Common Mistakes:

• Mixing boost and cut values
• Applying wrong percentages
• Changing mid frequency unintentionally