It is already known why brass instruments can sound brassy, and it's nothing to do with being made of brass: it turns out to be down to (weak) nonlinearity. However, it is not known what effect the shape of the instrument has on the sound. This talk gives an overview of some recent work to understand weakly nonlinear sound waves in curved waveguides (such as a trombone, for example). Along the way, a _huge_ amount of mathematical algebra gets swept away by using a really neat notation that also suggests how to extend all the usual linear concept to the weakly-nonlinear regime. The majority of this work was by Cambridge PhD student Dr James McTavish.