Solved

# FFT Java

Posted on 2011-02-15
1,661 Views
I used this tutorial that uses DFT http://jvalentino2.tripod.com/dft/index.html but the algorithm is too slow. Does anyone knows any similar tutorials explaining step by step of trasnforming audio into frequency domain using FFT? I wanted to use jtransforms library http://sites.google.com/site/piotrwendykier/software/jtransforms but can't find any examples how to use it. What method use, what input does it take? It is not specified. Anybody can help me with that?
0
Question by:chudyksg
• 2

LVL 7

Expert Comment

ID: 34910288
I am no expert on this stuff but since no one else has answered, I'll give it a shot.

Try this (I haven't tested it but it should be close):

First, change "int x[] = new int[n];" to "double x[] = new double[n];"

Then, replace the last box with the following:
``````DoubleFFT_1D f = new DoubleFFT_1D(n/2);
f.realForwardFull(x);
for (int j = 0; j < n; j += 2) {
double amplitude = 2 * x[j]/n;
double frequency =(j/2) * h / T * sample_rate;
System.out.println("frequency = "+frequency+", amp = "+amplitude);
}
``````

Be sure to check these results against your existing algorithm to determine if this method is correct.

Also, since the transform only uses the first half of x, you can save time in the loop populating x by changing "for (int i = 0; i < n*2; i+=2) {" to "for (int i = 0; i < n; i+=2) {".  There's no sense populating values that you're not using.
0

LVL 18

Expert Comment

ID: 34935025
There is a complete package, including Java source code at

It covers 1D, 2D and 3D. You should select what algorithm is the right for your needs, but most common is DFT.  The author claims it as the fastest FFT around and a benchmark is shown, to compare it against other packages.

It was developed in Unix environment, but compiles without problems in Windows or Linux, providing the right Java libraries.

Jose
0

LVL 18

Accepted Solution

JoseParrot earned 500 total points
ID: 34935103

Please note that you should select if you need 1D, 2D or 3D functions.
Let use single precision (faster) if it is enough for your case. If you need the best precision, then use the doubleFFT functions. As noticed by garypfirstech about work with just half of the matrix, due its symmetry, this package has such approach too..

You must first create the instance, by calling FloatFFT_2D (or 1D or 3D).
Must populate the matrix, with random numbers by calling fillMatrix_2D or filling it by yourself.
Use showReal_2D to watch the results and
writeToFileReal_2D to save it.

If you have difficulty with this package, you can also try the fft pacjkage at sourceforge: http://sourceforge.net/projects/jfftpack/ , but the Piotr's one is faster.
0

## Featured Post

### Suggested Solutions

Prime numbers are natural numbers greater than 1 that have only two divisors (the number itself and 1). By “divisible” we mean dividend % divisor = 0 (% indicates MODULAR. It gives the reminder of a division operation). We’ll follow multiple approac…
In this post we will learn how to connect and configure Android Device (Smartphone etc.) with Android Studio. After that we will run a simple Hello World Program.
Viewers learn about how to reduce the potential repetitiveness of coding in main by developing methods to perform specific tasks for their program. Additionally, objects are introduced for the purpose of learning how to call methods in Java. Define …
This tutorial will introduce the viewer to VisualVM for the Java platform application. This video explains an example program and covers the Overview, Monitor, and Heap Dump tabs.