The Role of Interactive Systems in Audience s Emotional Response to Contemporary Dance

Transcription

1 The Role of Interactive Systems in Audience s Emotional Response to Contemporary Dance Craig Alfredson University of British Columbia 2329 West Mall, Vancouver BC Canada V6T1Z craig.alfredson@gmail.com ABSTRACT This paper describes an experiment to determine whether an audience s emotional response is affected by making the performance environment interactive with the performers. Using the framework of contemporary dance, a set of custom made pressure sensors are mounted on a performer, the output of which is used to control the theatrical lighting system. A user study is then conducted to determine if an audience has a higher emotional response to a performance piece using such methods over a static performance environment. The results showed a decent improvement in emotional response and audience engagement. Although sample size was small, it could be deemed significant enough to warrant further investigation. Categories and Subject Descriptors J.5 [Arts and Humanities] Performing Arts (e.g., dance, music). General Terms Measurement, Experimentation, Human Factors. Keywords Dance, Theatre, Haptics, Pressure Sensors, Wearable Computing, Emotional Response, Interactive, Choreography, Improvisation. 1. INTRODUCTION The term interactive art seems to be thrown around a lot these days. In fields such as media art, this typically means that the audience can somehow manipulate the art piece to make it somehow individual to them. In contemporary dance, this typically means that the dancer(s) are to some degree in control of their environment so that there is an intrinsic link between the two. The environment in question is the lighting, sound, projection and in some cases scenic elements within the performance space. What I am trying to determine is whether this interaction between dancer and environment enhances the audience s experience in a meaningful way. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. HIT2012, Vancouver, BC, Canada. UBC2012. Note: In this paper I have mostly ignored the area of interactive music with respect to performance. The primary reason that I have done so is because I find that music is more intrinsically linked to performance than it is to environment. This link dates back to the pre-digital era when it was commonplace to have live musicians accompanying live performances, such as classical opera and ballet. In contemporary dance, it is often the case that music and sound fall more into the category of environment, but in those cases the tools and methods described here in terms of visual environment can be equally applied. 2. BACKGROUND 2.1 Choreography versus Improvisation In theatrical dance, the movement of the dancer(s) can be placed on a continuum between fully predetermined, or choreographed, and fully improvised. While work exists that spans the entire continuum, it is quite difficult to create work at either extreme. At the extreme of fully improvised, work can lack cohesiveness and structure. Typically dancers in improvised mediums such as contact improvisation discuss and rehearse basic interaction patterns or at the very least have an underlying theme or motivation for their movement [8]. At the other extreme, fully choreographed performance, while often strived for, particularly in classical forms such as ballet, is practically impossible. There are too many variables involved in live performance not the least of which human error which create changes in each subsequent performance. Although these may be very minute, they always exist and are often attributed as one of the major appeals of live performance. Whether mostly choreographed or mostly improvised, the human character of dance (and live performance as a whole) and the singularity of the relationship between these performers, this audience and their particular environment is what differentiates live performance from other artistic mediums. 2.2 Improvisation and Environment To follow on the idea of the continuum of choreographed versus improvised movement, the environment that is created to house the performers must also respect the point on the continuum that the performance occupies. In a tightly choreographed piece, the designers of the environment can tightly control the environment to suit the needs of the performance. For example, tightly focused lights can be brought up on specific times in the performance to highlight a specific performer or action. As the piece falls further along the continuum towards improvisation, this becomes harder

2 and harder to do. Due to the unpredictable nature of improvisation, the designer of the environment no longer knows the specific place and time to highlight, furthermore, they may not even know whether these places and times will exist. Several methods have been used to overcome these inabilities. The simplest method is to create a generic, static environment that covers all possible variations in the performance. This method, however, limits the ability for the environment to enhance the performance in a meaningful way. A second method is to allow the environment designer to have real time control over the environment. In essence, they become another performer in the performance. A classical example of this would be the follow spot which is the large human-operated light that is typically shone on the lead performers and follows them around the stage. There are a couple of limiting factors in this method. Firstly, the designer is only as good as his tools. Although a follow spot has control over position, size and colour of the light beam, it can t be (interactively) moved to become a silhouette shot from the back. Secondly, as tools become more and more complex, the designer reaches a certain limit in their cognitive and physical ability to control the tools. For instance, if they were given a button to control each of 100 lights pointing at the stage in various fashions, they would be limited by their hands and fingers to controlling a maximum of 10 lights at any given moment. A third method offloads some (or all) of this physical and cognitive load to computers. It is this third method of creating an interactive environment that I will be focusing on in this paper. 2.3 Macro versus Micro Interactivity Macro interactivity is what most people would think of when using this term. Performers make large gestures or perhaps move their whole body to various positions on the stage. In these cases it is obvious to the observer if the environment is being affected by the performer. Micro interactivity is more subtle, uses interactive element to augment a static environment. In these cases, the audience may not be aware that the interaction is happening. Thecla Schiphorst at the Simon Fraser University School of Interactive Art and Design has done much research in the area of wearable computing and touch sensitive garments [7]. In the professional dance community, many contemporary choreographers are exploring the concept of interactive environments. One forerunner in this field is Australian choreographer Gideon Obarzanek and interactive software creator Frieder Weiss [6]. 4. METHODS AND DESIGN In order to create a method for the environment to be affected by the dancers, I built several pressure sensitive pads, which were mounted in various locations on one of the dancers bodies. Pressure information was then transmitted wirelessly to a host computer, which used the information to affect the lighting state in the studio. 4.1 Pressure Sensors In my initial research and prototyping, I used commercially available Force Sensitive Resistors (FSR s) as pressure sensors. Although they gave me the pressure information I wanted, they were difficult to mount to the dancer s skin or clothing. They also did not have a very natural feel and are somewhat limiting in their small size. I began researching alternative methods and came up with two piezo-resistive materials that could be used to build home made pressure sensors. 3. RELATED WORK The idea of analyzing dance in terms of qualities of movement dates back to Rudolf Laban, who developed the concept of Effort/Shape Analysis which later developed into Labanotation, a concept of notating dance movement in a way similar to a musical score[5]. It is still widely used today. The ACR Media Integration and Communication Research Laboritories in Kyoto, Japan created a system called MIDAS [4] which deciphered dance movement and altered the environment to suit the dancer. This system was intended to be interactive with the dancer themselves and did not consider the concept of dancing for an audience. Camurri et al have shown that emotion can be conveyed through dance and deciphered fairly effectively by the audience [1]. Their work mostly deals with linking specific movements to corresponding emotions. Figure 1. FSR's mounted on insoles.

3 The first material is called VeloStat, by 3M. It is a thin plastic material, typically used for static dissipation. The second material is Eeontex by Eeonyx[2], which is a purpose-built piezo-resistive fabric. To build the pressure sensors, I took a piece of the piezo-resistive material and sandwiched it between two pieces of conductive silver-lined nylon, taking care that at no point the two pieces of conductive nylon touched directly. I then sewed a wire lead onto each of the conductive nylon pieces with conductive thread. Finally these three layers were placed between two pieces of nonconductive material to protect the circuit. These could be anything, based on need, but I found that iron-on nylon patches worked quite effectively. The middle three layers created a circuit that varied in resistance depending on the pressure applied. When no pressure was applied, the resistance was very high (in the order of 50KΩ). As pressure was applied, the resistance began to lower. Although I did not do any empirical testing, the reaction seemed to be non-linear, with resistance lowering quickly at first, then slowing until it held steady around 50Ω when pressed firmly with one hand against a hard surface. In my tests, the two piezoresistive materials reacted similarly (with slight variations in the resistance values). Although the Eeontex fabric had a slightly softer feel, I ended up building the majority of the sensors using the VeloStat for pricing and availability reasons. 4.3 Data Interpretation I used a visual programming environment called Isadora[3] to parse the serial data and then utilize it to affect various parameters of the visual environment. Isadora is commonly used in contemporary dance to create interactive environments, typically using video projection. For this study, I utilized the capability of Isadora to transmit lighting control data via the Digital Multiplex protocol (DMX512) to a standard theatrical lighting dimmer, which in turn controlled the intensity of the various lights that were hung in the studio. Figure 3. Isadora screen shot showing FSR values. 5. USER STUDY Figure 2. Home made FSR hooked up to Arduino Fio 4.2 Data Collection In order to collect the pressure data from the sensors, each one was connected to a voltage divider circuit. The divided voltage was then fed into an analogue input of an Arduino Fio board, which, using a 10 bit analogue to digital converter, converted the pressure to a value between 1023 (no pressure) and 0 (full pressure). These values were then transmitted wirelessly via ZigBee protocol to a host computer that read the values as serial (RS232) data. In practice, however, there was always a little bit of pre-load on the sensors, since the weight of the sensor itself and any costuming on top of it caused a drop in resistance. Similarly, at the low end, full pressure was mathematically impossible to produce due to the inherent limitations of the voltage divider circuit. The real time values ranged from approximately 900 down to 50. To test my hypothesis that an interactive environment will augment emotional response to performance, I commissioned 2 professional contemporary dancers to create a short (6 minute) dance piece, which was performed twice in front of an audience. The first time the piece was performed in a static visual environment created with theatrical lighting. The second time the piece was performed, various elements of the visual environment were affected by the data collected from the pressure sensors. For instance, when sensors mounted on one of the dancer s shoulders were pressed, the intensity of amber lights was increased and sensors mounted on the same dancer s rib cage caused the intensity of blue light to increase. Audience members were asked to rate their emotional engagement to the piece after each performance. Among other questions, they were asked to rate on a Likert scale from 1 to 5 how they felt the piece conveyed 6 standard emotions: Love, Joy, Surprise, Anger, Sadness and Fear. I did not give any specific instruction to the dancers in terms of choreographing specific emotional content, nor was any effort made to link specific colours or movements to predetermined emotional content. Instead we were hoping to find a basic, intrinsic link between the movement, the environment and the emotions conveyed. 5.1 Audience/Participants The audience consisted of 12 people, primarily students (11 of 12), with a mostly equal mix male and female. Most participants (11 of 12) were at least vaguely familiar with contemporary dance attending performances at least once a year, and all participants attended live theatre in general at least once a month. Cultural background was split, with 7 participants identifying themselves with a Caucasian (European/ North American or Australian) background and 4 participants identifying themselves with an Asian background (one identified with an other background).

4 6. RESULTS Although the sample size was quite small, there is a noticeable increase towards conveying more emotion in the piece with interactive lighting (see figure 5). 4 of the 6 emotions showed generally higher ratings, with only surprise and fear not showing a clear increase. Figure 4. Number of people who identified 6 basic emotions in each version of the piece. Based on a Likert scale - 1 being did not convey emotion and 5 being conveyed emotion strongly. Interestingly, when the choreographers were asked to fill out a similar Likert scale, they rated love and anger the highest and surprise and fear the lowest, which is somewhat consistent with audience s responses. Additionally, when asked which version of the piece they preferred and which one was more engaging, a significant number chose the one with interactive lighting (see figure 6). 11 out of 12 participants stated that the lighting enhanced the piece either somewhat or significantly. Table 1. Choreographers response to the same questionnaire. Love Joy Surprise Anger Sadness Fear

5 tracking, popular lately due to the release of Microsoft s Kinect. In terms of the hardware, development of more ergonomic and robust sensors using the Eeonyx fabric would be a good place to start. Furthermore, exploring the use of more sensors, and possibly sensors on multiple dancers would be interesting to explore both for its artistic and scientific merits. Figure 5. Number of people selecting their preference of piece. The second piece used interactive lighting elements. 7. CONCLUSION Although we have shown that emotional response can be enhanced by interacting with the environment, survey responses also show that overall enjoyment and engagement of the piece was also enhanced by the interactive environment. With a sample size of only 12 however, the data is not all telling especially with such subjective data. The hardware utilized also appears to be successful. The range of data that can be acquired via the body mounted pressure sensors is immense, and we have only just scratched the surface. 8. FUTURE WORK The most obvious place to continue this technology would be to continue with the user study to gather more data. In addition to this, other forms of interaction could be explored, such as optical 9. REFERENCES [1] Camurri, Antonio, S. Hashimoto, K. Suzuki,and R. Trocca. Kansei Analysis of Dance Performance. In P roc. IEEE SMC'99,IV, pages , [2] Eeonyx. [3] Isadora. [4] Iwadate, Yuichi et al. MIC Interactive Dance System- An Emotional Interaction System. Fourth International Conference on Knowledge-Based Intelligent Engineering Systems, Brighton, UK [5] Laban, R. and Lawrence, F.C.: Effort: economy of human movement, second edition. MacDonald and Evans, 2nd edition, [6] Obarzanek, Gideon. Chunky Move Dance Company. [7] Schiphorst, Techla et al. Threads of Recognition: Using Touch As Input With Directionally Conductive Fabric. CHI, Portland, Oregon, USA [8] Zimmer, Elizabeth. Contact Improvisation Comes of Age. Dance Magazine. June, 2004