Physical Modelling Synthesis and TAO

What do they have to do with each other?

Well, Tao is a software package built for sound synthesis with physical models. I’ve recently found out about this software after seeing a link for it in the Supercollider mailing list. It is a software project initiated and executed by Mark Pearson and Niall Moody that aims to build a framework for sound synthesis and composition by using and utilising somewhat hardcore physical modelling synthesis techniques so that you won’t have to deal with the complex physics and mathematics for controlling the synthesis process and generating the actual sound. For this purpose, it gives you some physically preprogrammed and highly parametrized mediums that generate sound when excited(like a string, circular membrane etc.), and gives you some parametrized devices to excite them(bows, hammers etc.). You can tweak your parameters and program actions to be executed over a time span and generate some really organic sounds and compositions around them.

The synthesis engine is not built for realtime usage, but there is a reason for that. The calculations carried out are quite complex and the physical models do not compromise the accuracy of modelling of a medium for realtime usage. You need to write some code to build a scene with instruments and you also need to code to compose. Your code is interpreted to C++ code, then compiled as an executable file.

One of the cool features of the main synthesis application is its ability to visualize the physical environment as it is being executed. You can watch the process you’ve described with code in an OpenGL powered window and this makes the instrument much more tangible. This is consistent with the “physical modeling” approach. In physical modelling, your parameters are mostly consistent with their real world counterparts, for example to get a certain spectral response from a string, you don’t define a filter and tweak its freq, res and Q parameters as in subtractive techniques, but you directly control the bow position and that position becomes your parameter. So it makes much more sense to visualize what really is going on during the process.

The modelling accuracy being a priority, the sounds generated by Tao are highly organic and rich. You can literally build a stringed instrument, compose with it, and change its physical properties (morph the instrument!) as it is generating sound, and it will sound natural. Check out the sound examples here:

http://taopm.sourceforge.net/examples/sounds.html

Building instructions for OsX:

It was fairly easy for me to build Tao on my OsX box. You need to have Xcode installed on your system. There are several package requirements for building that can be traced from the readme file of Tao package, but I only happened to need the audiofile library. You don’t need to change any settings, first extract audiofile library package, fire up a terminal, go to your extracted folder and follow the standard routine:

./configure make sudo make install

If all goes well, the library will be compiled and installed on your system. Then extract Tao to a folder, go to that folder with your terminal and repeat the configure-make-install procedure. If all goes well again, your computer will compile the software and install the necessary binaries to your “/usr/local/bin” folder. There is one more step you need, to be able to use Tao.

If you don’t have /usr/local/bin in your search path (which happens to be the default case in Tiger, but not in Leopard), launch your favorite text editor(not word processing software), create a file with contents:

PATH=$PATH:/usr/local/bin

And save this file as .profile in your home folder. It shouldn’t have an extension. Then logoff and relogin to your system and you are ready to go. You may then safely consult to the documentation of Tao for testing your installation, there is a test script and the manual and readme explains what you need to start and keep going.

I’m already planning to compose some music utilizing the great power of Tao so when it is done, we’ll hopefully see it in action.