A professional real-time digital signal processing application for audio analysis and frequency visualization. Built with Java 21 and JavaFX, featuring an intuitive GUI and advanced signal processing capabilities.

• Real-time Audio Capture: Multi-device support with automatic source selection
• FFT Analysis: Fast Fourier Transform using Apache Commons Math (pure Java, cross-platform)
• Frequency Visualization: Live frequency spectrum chart with JavaFX charts featuring:
  • Normalized dB scale (peak frequency = 0 dB)
  • Gridlines for easy frequency/amplitude reading
  • Clean tick labels (e.g., "1k" instead of "1000.0")
• Advanced Signal Processing:
  • Hamming window function for reduced spectral leakage
  • Noise threshold filtering (3x average magnitude)
  • DC component removal
  • 5-sample frequency smoothing for stable readings
  • Configurable FFT size (16, 32, 64, 128, 256, 512, 1024, 2048, 4096, 8192, 16384, 32768, 65536 samples)
• Cross-Platform: Native builds for Windows, macOS (Intel & Apple Silicon), and Linux (x64 & ARM64)

• Modern JavaFX-based UI with native platform look and feel
• Intuitive controls with real-time status updates
• Dynamic audio device selection with automatic sample rate detection
• Responsive chart with auto-scaling axes
• Performance optimized with 50ms update throttling
• Configurable sample rates (8kHz - 48kHz)

Download the latest release for your platform from the Releases page:

• Java 21 or higher
• Audio input device (microphone)

java -jar java-dsp-{version}-{platform}.jar

Or run directly with Maven:

mvn javafx:run

• Java 21 or higher
• Maven 3.6+
• Platform-specific JavaFX SDK (optional, Maven will download automatically)

# Clone repository
git clone https://github.com/gpoole/Java-DSP.git
cd Java-DSP

# Build for current platform
mvn clean package

# Build for specific platform (CI/CD)
mvn clean package -Djavafx.platform=linux        # Linux x64
mvn clean package -Djavafx.platform=linux-aarch64  # Linux ARM64
mvn clean package -Djavafx.platform=mac          # macOS Intel
mvn clean package -Djavafx.platform=mac-aarch64  # macOS Apple Silicon
mvn clean package -Djavafx.platform=win          # Windows x64

mvn test

Tests are skipped on headless CI environments for JavaFX-dependent components.

mvn clean test jacoco:report
# Report: target/site/jacoco/index.html

1. Select Audio Source: Choose your microphone from the dropdown menu
2. Configure Settings:
   • Select sample rate (recommended: 48000 Hz)
   • Choose FFT size (recommended: 2048 for balanced resolution/performance)
3. Start Capture: Click "Start Capture" to begin real-time analysis
4. View Results: Watch the frequency spectrum update in real-time
5. Stop Capture: Click "Stop" when finished

flowchart LR
    A[Audio Input] --> B[Stereo to Mono]
    B --> C[Hamming Window]
    C --> D[FFT]
    D --> E[Magnitude Calculation]
    E --> F[Noise Filtering]
    F --> G[DC Removal]
    G --> H[Frequency Smoothing]
    H --> I[Normalized dB Scale]
    I --> J[Visualization]

    style A fill:#e1f5fe
    style D fill:#fff3e0
    style J fill:#e8f5e9

• GUIFX.java: Main JavaFX application window with audio capture and control logic
• XYLineChart.java: JavaFX chart component with update throttling, thread safety, gridlines, and normalized dB display
• signal.DSP: Static utility class for FFT, power spectra, and frequency detection
• signal.SignalPipeline: Fluent builder API for chaining DSP operations
• signal.WindowFunction: Enum of standard window functions (Hamming, Hanning, Blackman, Flat-top)
• util.Preconditions: Shared input-validation utility for fail-fast error reporting
• FFT Processing: Apache Commons Math FastFourierTransformer
• Audio API: javax.sound.sampled for cross-platform audio capture

Use the signal-processing classes directly, without the GUI:

import com.gpoole.dsp.signal.*;

// One-shot: detect the dominant frequency of an audio buffer
double freq = DSP.dominantFrequency(samples, 48_000);

// Pipeline: chain window → DC removal → power spectrum
double[] spectrum = SignalPipeline.of(samples, 48_000)
        .window(WindowFunction.HAMMING)
        .removeDC()
        .executeToPowerSpectrumDB(0.776);

// Available windows: RECTANGULAR, HAMMING, HANNING, BLACKMAN, FLAT_TOP
double[] coefficients = WindowFunction.BLACKMAN.getCoefficients(1024);

This project uses GitHub Actions for automated building, testing, and releases:

• Build and Test: Runs on every push/PR across Windows, macOS, and Linux
• CodeQL Analysis: Automated security vulnerability scanning
• Release: Automated releases on version tags (v*), building platform-specific JARs

git tag -a v1.0.0 -m "Release version 1.0.0"
git push origin v1.0.0

This automatically triggers the release workflow, building artifacts for all platforms.

Java-DSP/
├── src/
│   ├── main/java/com/gpoole/dsp/
│   │   ├── GUIFX.java                    # Main JavaFX application
│   │   ├── XYLineChart.java              # JavaFX chart component
│   │   ├── signal/
│   │   │   ├── DSP.java                  # Static DSP helpers (FFT, spectra)
│   │   │   ├── SignalPipeline.java       # Fluent pipeline builder
│   │   │   └── WindowFunction.java       # Window function enum
│   │   └── util/
│   │       └── Preconditions.java        # Input validation
│   └── test/java/com/gpoole/dsp/
│       ├── FFTProcessingTest.java
│       ├── AudioFormatTest.java
│       ├── XYLineChartTest.java
│       ├── signal/
│       │   ├── DSPTest.java
│       │   ├── SignalPipelineTest.java
│       │   └── WindowFunctionTest.java
│       └── util/
│           └── PreconditionsTest.java
├── .github/workflows/
│   ├── build-and-test.yml        # CI pipeline
│   ├── release.yml               # Release automation
│   └── codeql.yml                # Security scanning
├── pom.xml
├── README.md
├── CHANGELOG.md
└── CONTRIBUTING.md

• Ensure your microphone is connected and enabled in system settings
• Check system permissions for microphone access
• Try running with administrator/sudo privileges

• The release JARs include all JavaFX dependencies
• For development, use mvn javafx:run which handles module paths automatically

• These are harmless warnings from JavaFX accessing native libraries
• To suppress, the application runs with --enable-native-access=javafx.graphics

• Increase FFT size for better resolution
• Ensure adequate signal strength (speak/whistle loudly)
• Try different sample rates
• Check for background noise

• Reduce FFT size (e.g., 1024 instead of 4096)
• Lower sample rate
• Close other audio applications

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.

See CHANGELOG.md for version history.

This project is licensed under the MIT License - see the LICENSE file for details.

• Apache Commons Math for robust FFT implementation
• JavaFX for modern cross-platform UI
• Java Sound API for cross-platform audio support