Loading Events

Taught by:

Join the live stream on 20th April 2020, 6:30pm BST / 10:30am PT, or watch it later.

In music, physical modeling synthesis is a technique aiming to reproduce analog sounds in software. Our guest speaker, Pat Scandalis, along with Julius Smith and Nick Porcaro have worked together since the 90s at CCRMA, Stanford, to create physical models of instruments. It isn't a very common form of synthesis, due to the complexity of doing it well, and the CPU it requires to run.

Physical modeling synthesis is the key technique used in Pat Scandalis' app GeoShred, which he built for iOS and iPad with MoForte and Jordan Rudess. Check out the impressive reel below.



Physical modeling synthesis has come a long way and is becoming a very attractive method to produce expressive sounds. Think of a physical model as the instrument itself: if it is the model of a violin, then it has 4 strings, a body, it can be played with a bow, staccato, etc. Each of these parts play a role in producing the sound. In a physical model, like with the real instrument, it is possible to alter a few of these parameters at the same time. When connected to a MIDI controller with multiple dimensions of control, a physical model instrument will offer a more reactive and richer response than traditional sampling engines, because of the variety of dimensions that can be controlled at once.

A great way to get started experimenting with physical modeling is to use the Karplus-Strong model, which gives the basics for modeling a plucked or percussion instrument. Sam's tutorial, below, gives a quick account of how to implement it in Max (it's nearly 10 years old, but still applies!).


The JUCE library also has a basic implementation of the Karplus-Strong algorithm in its examples, and you can check out the code on GitHub here.


What you'll learn

Course content

Membership plan: | Topic: ...
{{ error }} {{ success }}