#karplus-strong

For his Mini-Report, Theodore presented on Physical Modeling.

Theo covered a lot of interesting concepts pertaining to the synthesis and modeling of sound and the algorithms used in the process. Physical Modeling is a method in which a sound wave is generated and modeled to simulate a musical instrument. One of the first methods of achieving that was the Karplus-Strong algorithm: A burst of white noise is sent to a delay line. Afterwards, the sound is simultaneously output and fed into delay line that is L samples long. Lastly, a low pass filter is applied to the signal, which is then simultaneously mixed back into the output and fed back into the delay line. The Yamaha VL1 modeling synthesizer (released in 1994) uses this method to emulate different instruments.

A more modern version of physical modeling is exemplified in Logic’s Sculpture.

Sculpture provides three unique options to shape a sound:

String: This section allows us to choose the material of the exciter, media loss or how much sound is lost from the environment and we can also modulate the tension of the string, thus changing its momentary detuning after its plucking.

Object: Here we can set up to three parameters that affect the string and consequently change its timbre.

Pickup: This section emulates the pickups of an electric guitar and allows us to choose where they are placed along the string. We can also change where the three objects from the previous section affect the string.  The rest of the sections work as expected and allow for a fine modeling of the newly generated sound.

Theo's presentation has definitely given the class a spark of inspiration and has certainly inspired me to look into physical modeling for my future projects. He finished his mini-report with the world’s first “singing computer”:  https://www.youtube.com/watch?v=XIrX8YpiJgA