Examples

This page lists all available example scripts demonstrating various features of the spectrograms library.

All examples are located in the python/examples/ directory of the repository.

Audio Processing Examples

Basic Spectrograms

basic_linear.py

Compute a simple linear power spectrogram from a sine wave. Demonstrates the most basic usage of the library.

python python/examples/basic_linear.py

mel_spectrogram.py

Compute mel spectrograms with different amplitude scales (power, magnitude, dB). Shows how to use mel filterbanks for perceptual frequency scaling.

python python/examples/mel_spectrogram.py

Window Functions

compare_windows.py

Compare different window functions (Hanning, Hamming, Blackman, Kaiser). Visualizes the trade-offs between frequency resolution and spectral leakage.

python python/examples/compare_windows.py

Batch Processing

batch_processing.py

Efficient batch processing using the planner API. Demonstrates 1.5-3x speedup by reusing FFT plans across multiple signals.

python python/examples/batch_processing.py

Streaming

streaming.py

Frame-by-frame processing for real-time applications. Shows how to process audio incrementally with minimal latency.

python python/examples/streaming.py

Audio Features

mfcc_example.py

Compute Mel-Frequency Cepstral Coefficients (MFCCs). Common features for speech recognition and audio classification.

python python/examples/mfcc_example.py

chromagram_example.py

Compute chromagrams (pitch class profiles). Useful for music analysis and chord recognition.

python python/examples/chromagram_example.py

Image Processing Examples

2D FFT Basics

fft2d_basic.py

Basic 2D FFT operations on images. Demonstrates forward/inverse FFT and power spectrum computation.

python python/examples/fft2d_basic.py

Image Filtering

image_blur_fft.py

Apply Gaussian blur using FFT-based convolution. Shows how FFT convolution is faster for large kernels.

python python/examples/image_blur_fft.py

image_edge_detection.py

Edge detection using high-pass filtering in frequency domain. Demonstrates spatial filtering techniques.

python python/examples/image_edge_detection.py

Performance Analysis

fft_performance_analysis.py

Performance comparison against NumPy and SciPy implementations. Measures execution time across different configurations.

python python/examples/fft_performance_analysis.py

notebook.ipynb

Comprehensive Jupyter notebook with interactive benchmarks. Includes visualization of performance results and detailed comparisons.

jupyter lab python/examples/notebook.ipynb

Reference Implementations

numpy_impls.py: Reference NumPy/SciPy implementations used in benchmarks. Useful for understanding the algorithms and comparing approaches.

Running Examples

All examples are self-contained and can be run directly:

# Install the library first
pip install spectrograms

# Run any example
python python/examples/<example_name>.py

Some examples may require additional dependencies for visualization:

pip install matplotlib seaborn jupyter

Example Template

When creating new examples, follow this template:

#!/usr/bin/env python3
"""
Brief Title

Longer description of what this example demonstrates.
Include any prerequisites or special requirements.
"""

import numpy as np
import spectrograms as sg


def main():
    # Your example code here
    print("Example output")


if __name__ == "__main__":
    main()