Commercial Software Release (August 2013):
We've invested so much time, energy and capital into the development of these new algorithms and new features that we've decided to make Sonogram a commercial product. The latest and best release of Sonogram v.4 is available in two versions: The free, time-limited trial version, which will close itself automatically after each 5 minutes of use, but which you may re-start as often as you like. This version is for testing purposes only and for non-commercial usage. The personalized full version of Sonogram 4 can be obtained in our web store, and will be compiled from our automated shop system. You'll receive free updates for all successive versions. Your name and personal data will be binary-compiled into your download, and will be shown in the splash-screen and the about-dialog upon installation of the application.

Realtime Spectrumanalyzer (October 2012):

The image below shows the realtime spectrum analyzer. The realtime Sonograph is now part of the official Sonogram implementation and can be started out of the Sonogram main window. The audio recorder has the capability to store the files and reopen the wave files in Sonogram. The image above shows the recorder clock, the realtime fourier transformation and the realtime Spectogram.

Version 4 implements the Wigner-Ville-Distribution (Juni 2011):

This next major release of Sonogram supports the long-awaited Wigner-Ville-Distribution, which permits very deep user zooms into the Time and Frequency domains with an accuracy never before seen. Unlike the classic FFT, we can finally overcome the Quantum Mechanical equivalent of the Nyquist Sampling Theorem, the Time/Frequency Uncertainly Relation (Δt∙Δf≥SR/2). This Algorithm arises from the early days of the quantum mechanics where the classical mechanical world was adapted and bended to the quantum effects. Later the full quantum mechanic theory makes this algorithm needless. So far as we know, we are the first to implement this blindingly accurate algorithm into an audio spectrum analyzer. The core of the PSWVD algorithm (Pseudo-smoothed Wigner-Ville-Distribution) is massively parallelized, allowing it to take full advantage of modern multicore systems, just as Sonogram's FFT core algorithm did. The Beta Version of Release 4 can be started directly via the WebStart mechanism. In order to use the new UltraZOOM PSWVD function, you must first deeply zoom into the time-domain signal with your mouse, or manually select a small Time area. Because the zoom function is so intensive, it works only for very small signal chunks comprised of less than 32768 samples. If the selected area is small enough, the UltraZOOM PSWVD button will be highlighted with a border, enabling you to start the UltraZOOM. Check out the new "Demo6.wav" example, shown in the image below. Key improvements in detail include:

  1. UltraZOOM function with the Viígner-Ville-Distribution, including an analyze window and rendering dialog with the 3D-Perspectogramm and PNG-Image export.

  2. Re-design of the Transformaton algorithms with multicore parallelization for the SPWV and FFT algorithms and a better memory management.

  3. MP3 Support.

  4. Simple Audio Recorder with live spectogram.

  5. Re-design of the entire user interface, including the „Look and Feel“, extended Toolbar, new Dialog designs, a new splash screen and new color themes.

  6. Java 7 ready (though one is advised to use JAVA 6 until the new version becomes stable).

  7. Many major and minor features and bugfixes (Colors, splash-screen, system dialogs, frequency to musical note converter, about-dialog, open remote demos, ...).

  8. Reset function (activate CapsLock while startup to reset).

  1. Aufzählungszeichen Sonogram 4.0 for Mac OSX (must allow third party apps in security settings)

  2. Aufzählungszeichen Sonogram 4.0 for Windows

  3. Aufzählungszeichen Sonogram 4.0 for Linux/Unix

In order to use Sonogram, you must first have JAVA installed on your machine. Observe the difference in appearance between the new Cohen-Class Pseudo-smoothed Wigner-Ville Distribution with the display yielded by the traditional Fast Fourier Transformation (FFT). The two screencaps below depict analysis of the "Demo 4.wav" file from the included Sonogram online demos.

Above↑: Wigner-Ville-Distribution (WVD)                         Below↓: Short-Time-Fourier-Transformation (ST-FFT)

Version 3.2 (Build 4822) - Multiprocessor Support for Sonogram (Mai 2011):
Version 3.2 of Sonogram can make use of two or more user CPU's, because the core of the FFT Transformation Algorithm was parallelized. This means that if you have, for example, a processor with 4 cores on your machine, the Sonogram will be generated 4 times faster than before. The new version may be downloaded below, or launched directly using the Java WebStart mechanism.

Introducing a new Algorithm with better Time-Frequency resolution (October 2010):
The Wigner-Ville transformation (right) is an excellent algorithm class for detail-rich Time/Frequency analysis,   and offers a much sharper,  more brilliant image than that of the classic Fourier Transformation (left).  Further details about my newly-developed Cohen-Class algorithm may be found here.


Howto install Java3D on Windows 7  (May 2010):
There is an issue with the Java3D installation under Windows 7, so here is a short instruction how to install Java3D on Windows 7 properly:

  1. 1.Install Java3D for Windows with the Java3D installer program.

  2. 2.Copy the files from .../Program Files/Java/Java3D/bin to .../Program Files/Java/JRE/bin

  3. 3.Copy the files from .../Program Files/Java/Java3D/lib/ext to .../Program Files/Java/JRE/lib/ext

  4. 4.The JRE Folder could be called „jre6“ or „jdk1.6“ depending on your Java installation.

  5. 5.NOTE: For 64bit version of Windows 7 you should install Java x64 and the AMD64 version of Java3D.

Version 3.0 (Build 4248) Released (January 2010):
This new major release fixes many Java bugs in the GUI and the logarithmic frequency scale, giving the application an overall performance boost. The new features,  in detail, are:

  1. A welcome screen added, featuring basic instructions on what Sonogram is, and how to get a Quickstart on the program.  A Welcome screen shown at application start.

  2. A floating information wndow added in the main panel of Sonogram showing the amplitude, frequency, time and level under the current mouse position.

  3. A new color added.

  4. The logarithm frequency scale was re-written, and became, like the image-smoothing functions, a toolbar button in the Main window.

  5. The single-spectrum view in the Main window was re-written, and now shows a better spectrum image.

  6. The entire GUI was overhauled, and the application's startup procedure received, like the imaging routines, a stronge performance boost. The overall design given a facelift.

The new version is published via JAVA WebStart, and was tested on every popular operating system (Linux/Mac/Windows). JAVA WEbStart has a functioning enabling the user to add the program to the start menu or applications folder. Note that the user must have installed a JAVA runtime (JRE) in at least Version 1.6, with the optional JAVA 3D.

If you need a sharper, more brilliant Time/Frequency resolution, check out my recently-finished implementation of the "Wigner-Ville" Transformation here. If you are pleased with Sonogram, you can support our development by donating online via the button below. Good bug reports and feature requests are most welcome in thev open-source software world, so don't hesitate to visit us at the SourceFore Sonogram Page.

iPhone Version of Sonogram (November 2008):

Yes - I invested the 79€ and am now a paying member of the Apple Developer Connection, and an official iPhone Developer, with a license to sell iPhone software at the online iTunes Store. And yes, this means that the iPhone version of Sonogram's algorithm collection will cost some money. Following their specification:
  1. The name will be changed.  I plan to call it simply the "iPhone Signal Analyzer".

  2. All algorithms run as a realtime analyzer application.

  3. At the least, it should allow possible switching between the following signal analysis algorithms: Waveform,   Spectrum, Spectrogram/Sonogram, LPC Spectrum, LPC Prediction Error, Auto-correlation, Cepstrum, Wavelet Transformation.

  4. One may use the microphone as well as the acceleration sensor for realtime analysis.

  5. It might be feasible to implement the 3D Spectrogram using OpenGL ES.

  6. Develop a pleasing, recognizable icon for the tool, and feature a smart design using Sonogram's now well-known color scheme.

  7. Write a software documentation which includes as much as possible formulas from the algorithms including there derivation.

Version 2.4 (Build 2440) (Released (September 2007):
The new version fixes many Java bugs and brings an extended anti-aliased 3D Perspectogram engine. Along with the Java Runtime (at least in version 1.6) it is strongly recommended the user install the latest Java3D and JMF extensions in order to make use of all the features. Sonogram can run without Java3D and JMF, but the generation of Perspectograms and the playing of sounds will not be possible without these needed extensions. The next Sonograms releases will only be offered via WebStart, and its minor changes will not be explicitly announced on this page. Sonogram can be launched directly via the WebStart button, and can run on any operating system (Windows/Mac OSX/Linux-Unix). Version changes in detail include:

  1. Many, many bugfixes in the core Java code (afecting render, display, windowing, etc.) and all bugs reported in the Sourceforge bug-tracking area are now fixed  (Thanks for all bug reports!).

  2. An improved, anti-aliased Perspectogram engine, featuring a coordinate system, rotation effect, point cloud,  etc.

  3. Most of the dialog, tooltip and GUI element text labels have been revised and re-written.

  4. A feedback message system has been included for bug reports and information exchange.

  5. Re-written pitch-tracking algorithm.

  6. Better design of the entire GUI.

  7. Java 6 ready.

Sonogram goes Open Source (August 2005):

Sonogram was open-sourced after we obtained necessary permissions from the DKFI, where Sonogram was initially developed. The source of the latest version can be found at the Sonogram Sourceforge Webpage. Please help us to improve the software submitting bug reports to the Sonogram open source page in the bugtracking area.

Start Sonogram via Java WebStart (July 2003):
The latest and future builds of Sonogram may be launched directly, online, via JAVA WebStart, and I will use the WebStart platform exclusively in future versions. Downloads from Sourceforge will, in future, be outdated, and I will not announce any minor version changes here on this webpage. New releases and updates will be offered to the user automatically in this fashion via WebStart. Sonogram requires a Java Runtime in version 1.5. It is strongly recommended to install Java3D and JMF in order to use all the newest fancy features of Sonogram.

This is the first open source release of Sonogram. We've put a lot effort into testing and debugging this release, but of course the software may still have some "teething problems". Sonogram itself features a message feedback system whereby bug reports may be sent immediately, natively within the application.



Sonogram is a highly flexible audio spectrum analyzer for the analysis of sound, music and speech signals in the frequency-domain using different new algorithms. It runs on OSX, Windows and Linux.

Version 4 
includes the
new WVD
classic FFT
Click her to start Sonogram