Page  00000037 STIRRING, SHAKING, AND MIXING: MUSICALIZING EVERYDAY ACTIONS Juraj Kojs Virginia Center for Computer Music University of Virginia koj s@virginia.edu ABSTRACT This paper discusses how physical and cyber everyday objects formed a union in the composition Neither Stirred, Nor Shaken. Cocktail glasses, metal shakers, and electric blenders were defined as musical instruments. Actions such as stirring, shaking, and blending were tracked by sensor technology. The physical models of these everyday objects were created to timbrally augment the sonic qualities of the physical instruments. A real-time performance engages three percussionists who operate the everyday objects and control the cyber ones. 1. INTRODUCTION Everyday objects have rich history in music. Already in the early 20th century, Italian Futurist Luigi Russolo designed, built, and used in music compositions a set of musical instruments called intonarumori (noise intoners), which mechanically and electrically reproduced everyday sounds. After Russolo, the everyday objects became subjects of exploration for the generation of Fluxus musicians and experimentalist composers such J. Cage, H. Partch, L. M. Young, C. Cardew [1]. Engaging everyday objects in electroacoustic music compositions is however uncommon. Pairing physical and physically modeled everyday objects and contextualizing them in a musical performance stimulates a creation of a novel musical forms and expressions. In Garden of the Dragon (2003), the amplified cellophane and plastic corrugated tubes were used to shape a musical ecosystem. A choir of the singing tube physical models was 'nurtured' by the action of the physical performance. The parameters of the sonic signal provided by the performers controlled the models real-time [2]. In Revelations (2005) for circular toys, resonant plates and electronics, the everyday objects such as plastic superballs and glass marbles excited a set of custom-designed resonant plates of while bouncing, rolling, and scraping. Figure 1 displays everyday objects involved in Revelations. The audio signal emanated from the plates was mapped into a number of cyberinstruments by physical modeling synthesis implemented in MAX/MSP [3]. The cyberinstruments complemented the actions of the physical objects and Stefania Serafin Medialogy, Aalborg University Copenhagen, Denmark sts@media.aau.dk participated in creation of augmented analog-digital instruments. Figure 1. Individual and group set up of Revelations for circular toys, resonant plates, and electronics. Music making with everyday objects is also the primary activity in the composition Neither Stirred, Nor Shaken (2007) for cocktail glasses, shakers, blenders, and electronics. As opposed to the previous works, however, this composition uses sensor technologies to mediate contact between the physical actions of the performers and responses of cyberinstruments by means of physical modeling. In the following paragraphs, we examine the compositional structure, design and function of involved cyberinstruments, tracking and mapping consensus between the physical and cyber objects, and performance issues connected to composing with everyday objects and electronics. 2. NEITHER STIRRED, NOR SHAKEN Neither Stirred, Nor Shaken (2007) is orchestrated for everyday objects such as cocktail glasses, shakers, blenders, and electronics. Writing the composition was motivated by research published in [4], in which Trevithick et al. discuss the potential health benefits of stirring over shaking in preparation of martini. It became curious to examine the sonic properties of making the cocktails in a multiplicity of ways. In Neither Stirred, Nor Shaken, the three performers produce sounds, while they stir liquids and ice with metal spoons in highball cocktail glasses, shake their concoctions in metallic shakers and mix them in electric blenders. The sonorities produced by these activities are pitched and percussive. While the percussiveness is directly related to the performance action, the pitch content depends on the resonance frequencies of the particular container. The instruments are rather simple and produce limited, yet attractive, timbres. In general, simplicity is both the most 37

Page  00000038 Section 0'-2'20" appealing and, in the same time, limiting aspect while composing with these everyday objects. To augment the expressive potential of the instruments, the physical objects are coupled with cyberinstruments by physical modeling synthesis such as shakers, engine rattles, and singing glasses. This pairing has resulted in a creation of novel physically augmented textures and timbres. Ecosonicscape and its characteristic formations resulting from the interaction between the everyday objects and computer simulated sonic entities presents a unique, yet familiar, terrain for compositional interplay. In particular, the cyberinstruments enable smooth timbral extension of the physical objects. The continuum between the physical and cyber sounds frames the composition. Neither Stirred, Nor Shaken can be structurally divided in three parts according to the dominating performance mode: stirring, shaking, and blending. The players explore the musical potential of the selected activity in each part while producing a cocktail drink. At seven minutes, the players garnish and taste the three drinks they prepared and compare their quality. Following the tasting, the players order the drinks according to their quality. The principal player collects the ordered data and inserts them into a MAX/MSP assessment patch. The evaluated data influences the selection of one of the three potential solo-electronic conclusions. Each of the final sections highlights the sonorities of physical and physically modeled stirring, shaking, or blending actions. The pitch content of the physical instruments is given, and that of electronics is pre-composed. The content of electronics was derived from the studied frequency spectrum of the physical cocktail glasses. Thus, the computer generated spectrum mimics and registrally extends the acoustic spectrum. The frequencies inherent to the physical spectrum emerge registrally displaced, fluently commuting, and independent from their amplitude in the sonorities of the virtual objects. During the composition, multiple pitches converge to smaller streams and finally to a single flow. This convergence is additionally reflected in the continuous registral movement, which is primarily that of a descent. While the composition's opening presents pitches in the C8 octave, the concluding section displays migration to the C4 area. The overall pitch structure has a shape of a branch or river as shown in the Figure 2. The horizontal axis of the three systems indicates time in ten second increments. Each of the three systems is divided into three parts signifying frequencies of the virtual objects aligned with the performance of the three players. The rhythmic complexity arises from combination of simple patterns produced by three performers and approximately 30 virtual objects per player. As detailed later, the human gestures provide the excitation mechanism for these cyber objects. Fib 9' ii ~.............. - A:,L7 Pw, 2 ---7-- " 4(" ---v..- -----.-- = --------....... ii, i......................... -....... - can - - - --------........ ---------- cz:: Section 2'20" ^''"i----"'*"" E7,^,^.FS va...................................------ ------ - - ---5 ' _' 3:2o':'0m '.............................: - - - ** "::.::.,::.-,::;....i ii.....................................;; - I L ;.; L I...... L..................: n~i: F4................-" "l~l i,:: G^as t p^ 1^-- A77 F47 ---- ---- ----............"....... - -...... --- - ", - ~"-;---~I~-I-------- ------10 -____~IIII~LI~ _~~~L~iii.......... Section 5'-9' F0 U G" "I 8~f Figure 2. Overall "river-like" trajectory of the pitch content. The performers' parts are notated in traditional notation, as the music is based on the exploration of inherent rhythmic and percussive qualities of stirring, shaking, and mixing. 2.1. Set up, sensors, tracking and mapping The set up for each player consists of highball cocktail glasses, a tall metal spoon, metal shaker, electric blender, ice, liquid, sensors, and Make Controller board [5] as shown in Figure 3. The Make Controller board is 38

Page  00000039 connected with a computer running the Max/MSP application via a USB cable, although Ethernet/wireless connections are also possible. There are three sensors, which track some of the parameters of cocktail making in real time: a temperature sensor, an accelerometer, and a piezo sensor. The temperature sensor is embedded in a used plastic CD case in order to avoid moisture from the cocktail glasses as displayed in Figure 4. The cases became a practical glass stand. Figure 4. A four-part process of creating a shell/stand for the temperature sensor. The piezo sensor (a LDTO solid state switch/vibration sensor) comes into play in the final portion of the composition, when the players perform on three electric blenders. The sensors are taped to the mixers' engine exits in order to track the air pressure changes at varying mixing speeds. The cubes and liquid in the shaker create the engine's resistance. The mixer sonorities and tarcked data varies from performer to performer, as they operate with a different number of ice cubes cubes. In Neither Stirred Nor Shaken, the performers control the two-speed-five-mode blenders with the pulse function. Figure 3. Individual set up for each player. The LM34 temperature sensor [5] measures the decrease of temperature of the stirred drink over the period of the composition (ca. 9 minutes). Each player operates with a different number of cubes (5, 10, and 15). Consequently, the temperature decrease and rate differs from player to player. Stirring is a delicate performance mode. Thus, the slowly decreasing temperature alligns with the subtle sounds produced. In the second part, the shaking presents a more intense activity. Each player operates a shaker onto which an ADXL330 accelerometer [6] is fastened as shown in Figure 5. The sensor tracks the movements of the container and reports a series of data immediately used to control the cyber shakers. This part concludes with pouring the drink into a cocktail glass and preparation of the electric mixer. Figure 5. An accelerometer in action. Data obtained from the sensors control the cyber shakers, engine rattles, and singing glasses physical models. The models' parameters have been divided into two categories: indispensable and variable. The indispensable parameters such as frequency were pre-composed as previously detailed. Thus, there is a definite direction in the pitch progression. Other parameters such as a number of ice cubes in the shaker, resonant spread, and the energy were extracted from the data provided by the physical objects. The programmability of the indispensable parameters ensures, that the models will sound. The inconsistent and unexpected behavior of the variable parameters ensures that each consecutive performance will differ slightly. 39

Page  00000040 2.2. Cyberinstruments by physical modeling The physical model of a cocktail glass, an adaptation of wineglass model previously described in [7], largely contributes to the creation the pitched material in Neither Stirred, Nor Shaken. We examined the spectrum of the hit target cocktail glass and identified two main resonances as shown in Figure 6. These resonances constitute the target for the resonator's model. Either hitting or rubbing can excite the modeled singing glass, which were simulated by banded waveguides. The cyberinstruments produce a truthful simulation of physical crystal wineglasses. The rubbed wineglass can be obtained by using the friction model described in [8]. Shakers and rattles are simulated using particle physical models, adapting the algorithm proposed in [9]. Each ice particle was implemented as a modal resonator. Given the physical nature of the particle, mapping from data from the accelerometer to the particle model was rather straightforward. 0.5 o n-9----------------------- 0 0.5 1 1,5 2 2,5 3 3.5 4 Time (s) x 10 601 " 4 0o... i............ environment in a creative process and thus raising awareness about the world around. Secondly, we suggest a continuous augmentation of the physical soundscape into cyberspace. Thirdly, we strive to encourage people with little or no musical training to perform music, which combines familiar objects with technology. 4. REFERENCES [1] Cope, D. New Directions in Music. Propsect Heights: Waveland Press, 2001. [2] S. Serafin and J. Kojs. "Computer models and compositional applications of plastic corrugated tubes." Organised Sound, 10(01):67-73, 2005. [3] M. Puckette, "Max at Seventeen," Computer Music Journal Vol. 26, No.4 Winter 31-43 (2002). [4] C. C. Trevithick, M. M. Chartrand, J. Wahlman, F. Rahman, M. Hirst, and JR Trevithick. "Shaken, not stirred: bioanalytical study of the antioxidant activities of martinis". British Medical Journal, 319(7225):1600 -1602, 1999. [5] [6] [7] wwwjBm ak iitLs.co ( wwwsarkfma corn G. Essl, S. Serafin, P.R. Cook, and J.O. Smith. "Theory of Banded Waveguides," Computer Music Journal, 28(1):37-50, 2004. S20 -20 1! -40 0.5 1.5 requency(Hz) 2,5 x 104 [8] F. Avanzini, S. Serafin and D. Rocchesso. "Interactive Simulation of Rigid Body Interaction With Friction-Induced Sound Generation." IEEE Transactions on Speech and Audio Processing, Vol. 13 (5), 2005. [9] P. R. Cook. "Physically informed sonic modeling (Phism): Synthesis of percussive sounds." Computer Music Journal, 21 (3): 38 -49, 1997. Figure 6. Top: time domain simulation of the cocktail glass used in the simulations. Bottom: spectrum of the wineglass with two strong resonances. The percussive sonorities of the physical stirring, shaking, and blending are augmented by the sounds of the cyberinstruments. All physical models implemented in MAX/MSP and can be downloaded from msg.media.aau.dk. 3. CONCLUSION This paper discussed how everyday objects such as cocktail glasses, shakers, and electric mixers can be engaged in the musical context. The cyberinstruments by physical modeling enabled various timbral extensions of the physical objects and participated in the creation of novel analog-digital instruments. The purpose of compositions such as Neither Stirred, Nor Shaken is threefold. Firstly, we are interested in involving the elements from our 40