tambapps / fourier-transform-library

A FFT library written in Java

Java Fourier Transform Library

This is a library for computing 1-2 dimensional Fourier Transform. It was written with Java 8, and should be Android-compatible (you can use it in an Android project). This library was written without any compile dependencies.

You can import it with maven.

If you are more familiar with the old implementation (version 1.0), consult the legacy branch

<dependency>
  <groupId>com.tambapps.fft4j</groupId>
  <artifactId>fft4j</artifactId>
  <version>2.0</version>
</dependency>

Fast Fourier Transform

Here is an example of a 1D fast fourier transform. There are several algorithms to perform FFT You can see all of them on the FastFouriers class.

double[] inputRe = getInputDataRe();
double[] inputIm = getInputDataIm();

double[] outputRe = new double[inputRe.length];
double[] outputIm = new double[inputRe.length];

FastFouriers.BASIC.transform(inputRe, inputIm, outputRe, outputIm);

// consult result in outputRe and outputIm

You can also use a Signal which encapsulates a real and imaginary double arrays.

Signal input = inputData();
Signal output = FastFouriers.BASIC.transform(input);

3 different Fast Fourier Algorithms were implemented

Basic

This is a trivial implementation of a Fourier Transform using the basic Fourier Transform formula

Recursive Cooley–Tukey

A recursive implementation of the Cooley–Tukey FFT algorithm. Note that this algorithm requires the input length to be a power of 2.

Iterative Cooley–Tukey

An iterative implementation of the Cooley–Tukey FFT algorithm Note that this algorithm requires the input length to be a power of 2.

2D Fast Fourier Transform

You can apply 2D FFT with a FastFourierTransformer2D. You can change the algorithm used by the transformer to compute fft by setting the AlgorithmChooser.

Signal2d signal2d = new Signal2d(M, N);
fill(signal2d);

FastFourier2d transformer2D = new FastFourier2d();
transformer2D.transform(signal2d);
// do some things with the fourier transform
transformer2D.inverse(signal2d);

// don't forget to shut it down as it uses an executor service
transformer2D.shutdown();

Example use case

You can consult the fft-image-processing repo, an app that transforms images using FFT

fourier-transform-library's Issues

FFT size

Hello! I am conducting an experiment in which I take the frequency contents out of songs and then cluster them. For that, I need the FFT size, but I cannot seem to find it or be able to calculate it.

Could you tell me the size of the FFT from this repository? I am using the Basic Fourier Transform class.

Thank you very much

build process failed in jdk11 and maven3

i get error message

[ERROR] D:\ebook\python_dsp\fourier-transform-library-master\src\main\java\com\tambapps\fft4j\fourier\fft_1d\CooleyTukeyRecursiveFFT.java:11: error: incorrect use of inline tag
[ERROR] * @link from https://rosettacode.org/wiki/Fast_Fourier_transform
[ERROR] ^
[ERROR] D:\ebook\python_dsp\fourier-transform-library-master\src\main\java\com\tambapps\fft4j\fourier\util\Padding.java:5: error: bad use of '>'
[ERROR] * zero-padding => raise resolution
[ERROR] ^
[ERROR]
[ERROR] Command line was: C:\Java\jdk-11.0.11\bin\javadoc.exe @options @packages
[ERROR]
[ERROR] Refer to the generated Javadoc files in 'D:\ebook\python_dsp\fourier-transform-library-master\target\apidocs' dir.

Java 16 SIMD support for faster lib

Hi ,
Java 16 has released Vector API

You can look at the links for details:
https://metebalci.com/blog/what-is-new-in-java-16/
https://openjdk.java.net/jeps/338

Java 16's SIMD API has intrinsic capability.
Much faster then JNI.That directly converts these method calls to ARM NEON or AVX512 etc..

It will be good to fit fourier-transform-library to new Java 16's Vector API