~Proceedings ICMCISMCI2014 14-20 September 2014, Athens, Greece ing keyboards as the input device, however, nowadays the facto standard is the use of peripheral devices either similar in shape to the musical instruments (Guitar Hero, Rock Band etc.) or real musical instruments (e.g. Rocksmith). There are also some educational games in this genre which are intended to teach beginners how to play musical instruments (e.g. Joytunes, 1). Although at first glance similar, those games and our approach differs in several fundamental ways. In our work games are created in a way that encourages creativity and improvisation, while in rhythm games, players are supposed to mimic predefined actions as close as possible. Another difference is that music rhythm games have themes related to music playing, whereas our games are just generalthemed video games. On the other hand video games have also influenced music. The most notable example is the creation of new music genres, two of which are: game soundtracks, and those based on musical application of game consoles - chiptunes [4]. There were also more subtle influences. First of all, the aforementioned use of generative music in games made the idea of algorithmic composition known and accessible for the general public. Second worth mentioning influence is the usage of sound effects, characteristic synthesis techniques, and motifs from the game music to evoke certain moods associated with the video games. Two interesting systems such as scrabble2midi [5] and "Music for 32 Chess Pieces" [6] use sonification of game states as the main musical element. In these systems the game controls the generation of music, our work also includes the opposite direction, where music controls the game. More closely related to our work is the application of game elements in musical composition and improvisation. One of the earliest examples is the eighteenth century Musikalisches Wurfelspiele [7], in which dice were used for choosing music segments form a given set of possibilities. Another example are the game-theory-based Xenakis' compositions, i.e. Stratigie and Duel [8]. The work most relevant to our system is the so called "game pieces" - which consists of sets of rules used in group improvisations. The compositions by John Zorn (e.g. Cobra, Hockey, Lacrosse, Xu Feng)[9] are the best known game pieces. Two other examples of game pieces by Shiba Tetsu can be found in http://www20.brinkster. com/improarchive/ shtgp.htm. 3. CONTROL AND ANALYSIS STAGE We chose to divide each game into two modules which communicates via the MIDI protocol. Those modules are control/analysis and game. In this section we will discuss the control modules which were created in the Max/MSP environment. We have also used Sonuus i2M audio to MIDI converters for simplifying the controls modules in case of pitch related analysis. 1 http://www.joytunes.com/ 3.1 Goals and Problems There were few goals and problems which we needed to answer during the design and implementation of the control modules. The first of them, and one of the two most important, was choosing a control scheme which allows players to take intended action in games. The most important factor in this scheme, which affects the level of control that players have, is calibration of the analysis parameters. Without correct calibration, there were situations where players were not able to take action or many false readings occurred. Finetuning the parameters after any change in the audio systems, change of microphones, levels, instruments, etc, was crucial to achieve correct performance. The second most important problem was allowing musical freedom for players. Since one of the main elements of this project was improvisation, we wanted to avoid situations similar to the ones in rhythm games, where a player can only repeat a predefined sequence of actions. In the games where players were rewarded for performing certain actions, we identified the risk of encouraging only high intensity playing rather than the desired artistic expression. Two solutions to this problem were investigated. First we consciously designed the games so they did not encourage high intensity playing all the time. Second, we introduced artificial limits to the number of events in given periods of time above which, new events were ignored. 3.2 Control Schemes There are many possibilities for which musical elements should be involved in the control of the games. This section presents and comments on a few of them. " dynamics- The use of dynamics for playing gave players a natural control of the effect of their actions in the game by playing louder or softer. A good example of this was the Pong game, where the speed and direction of the paddle movement was controlled by the intensity difference between two musicians playing. In games where the possible actions were just one-off discrete events (such as in Mustetris) we implemented two possibilities: The first one was to set a threshold above which every loudness measurement produced an action. We found this impractical and limiting from musical point of view. The second option was to define a "counter trick". The counter trick took each value of dynamics measurement added it to the counter. When the counter reached a limit, the action was triggered and the counter is zeroed. This resulted in a simple yet effective way for controlling the games without limiting musical freedom excessively. " density - A similar idea is to use density of playing as a controlling element. We measured it either by counting NOTEON messages (for MIDI instru ments) or analyzing transients (for acoustic instruments). We have found that using density gives similar results to dynamics, but limits the types of music textures possible to use. - 89 -
Top of page Top of page