Engraving--Hammering--Casting was co-composed by Alexandros Kontogeorgakopoulos and Edgar Berdahl and was made possible by a collaboration with the Cardiff School of Art and Design. The composition was performed live in September, 2014 at the International Computer Music Conference (ICMC) in Athens, Greece. Here is an excerpt of the audio:  

 

For a detailed description, please see the paper.

 

The composition consists of three sections. In the first section, the performers interact with model parameterizations designed to evoke perceptions of engraving. With high resonance frequencies and low masses for the resonators, the sound is delicate and responds intimately to the small, precise movements made by the performers.

 

In the second section, the resonators are re-tuned to sound more like pieces of metal or bells. The performers make hammering gestures to play melodic-like passages.

 

Finally, in the third section, the k and R parameters of the contact links are varied rhythmically in time. Through this modulation, the virtual instrument seems to gain the ability to exert forces on the performer. It asserts a rhythmic form on the gestures of the performers, as if it were casting the performers' gestures into a specific form.

 

The score for the composition consists of six staves, which are notated in a special manner but also contain traditional marks from Western music notation such as remarks, dynamics, etc. The first staff describes which sides the first performer should play and at what time. The f note describes the right side, the a note indicates the left side, c indicates the bottom side, and e indicates the top side.

 

Consider the engraving section, for which k is small and R is big, resulting in a kind of frictional interaction. Arrows on the score indicate bowing-like gestures to be performed. For example, subject to this interaction, the hypothetical top staff would specify that the performer should first play a rest for four beats, and then for five beats the performer should slowly push down into the bottom c side. Next, the performer should push to the left into the left a bar, at a position low enough that both the bottom and the left sides will create sound. Similarly, the second staff would indicate that only in the third measure, the second performer should play by gradually pushing into his or her the bottom side.

 

The stiffness (k) and damping (R) interaction parameters are prespecified by the score and not under the control of the performers. The lower four staves of the score specify how k and R interaction parameters vary during the composition. In the excerpt from the engraving section shown, the interaction stiffness remains low for both performers while the interaction damping gradually increases over five bars for both performers. On the right, the figure below shows another example in which the damping remains generally low for both performers. The stiffness for performer one varies periodically to emulate engraving, and after three bars, the stiffness for performer two also begins to vary to emulate casting for performer two. Through the variation of the interaction parameters, the haptic force-feedback device asserts its influence over the performers, in a sense casting their gestures into a form that suits the model's programming.