The Cranky Bear is an interactive sound installation that let participant to tangibly interact with a toy teddy bear that could give different responses under different circumstances it experienced. When the bear put down (lying), the system will play lullaby song and snoring sound. When the bear is lifted up, the system will trigger sound that makes the bear apparently talking or ranting. When the bear is being shaken, the system will trigger shaking sound as well. If it is shaken vigorously, the system will trigger the angry sound of the bear. There is also a voice recognition system that can calm the bear and make him sleeping again. All of these actions create an interesting interactivity out of simple interaction.
All of these actions achieved by measuring movement and orientation from the motion sensor tucked inside the bear that communicates wirelessly with the computer system. The computer system determines actions should be triggered under different kind of specific circumstances.
At initial stage, we would like to create a horror-themed interaction by using the same teddy bear object. Later on, we revised our plan by characterized the bear and creating a unique interaction around the character of the bear itself. All we would like to achieve from this installation is creating a unique user-experience from interacting with toy teddy bear that has a mild character that quite different from ordinary children toys. We believed that if unique character can be displayed by an ordinary teddy bear, the results would be interesting.
Related or Influential Works
We were inspired from several talking toy dolls available on the market. These dolls could play some pre-recorded sounds when being triggered, such as pushed, squeezed, etc. For example, there is talking Barney doll, Bob the Builder doll, These kinds of interactions are very stereotyped with kids-friendly voices/sounds. Therefore, we would like to slightly alter this stereotype in this installation by creating a non kids-friendly sound/voices out of popular kids toy object.
Technical Method
For this installation, we take advantage from sophisticated motion sensor of Nintendo Wii remote. By using custom Max/MSP patch, the three-dimensional axis (X, Y and Z-axis) data from the remote can be read by the computer and used from useful purposes.
From several experiments, we figured out a set of axis ranges of the bear (with the remote tucked inside) that used to define various positions of the bear (i.e., lying position or seating position, etc.) and action being experienced by the bear (i.e., being shaken). Out of these specific axis readings, the patch will trigger other patch that initiate specific sound/voice playback. All of the music and voices used in this installation project is custom-made by one of our group members.
Max patch for detecting lying position of the teddy bear:
Max patch for detecting amount of movement of the teddy bear:
Max patch for triggering sound:
We implemented a voice recognition system as well. It is achieved by planted a wireless lapel microphone in the teddy bear’s right ear. There is other independent Max/MSP patch that recognizes human voices. When the system received a voice input, it will analyse it and recognized pre-determined words, such as “sleep”, “relax”, “quiet” and “calm”. If the system recognizes any of these words, it will trigger other sound.
Max patch of voice recognition:
User Experience/Usability Considerations
Most of the audiences enjoyed playing with the teddy bear. They enjoyed the shaking voice of the bear when it is being shaken and some quite surprised by the rather impolite response of the bear when being shaken vigorously. However, the cuteness of the teddy bear itself is a visual appeal that attracts audience’s attention at the first place. Sometimes the system experienced few latency problems while switching between positions/sounds. There is no transitional sound programmed while switching between sounds, therefore it is very rough when transitioning between each sound. There was a problem in voice recognition system as well. The system does not work properly, thus cannot constantly detect voice input.
One major problem of this installation is fixing the Wii remote inside the bear, so it will stay at the same place even being shaken extensively. We experienced a coincidence that the system cannot trigger specific sound because of the remote has displaced inside the bear.
Resources/References
aka.wiiremote - http://www.iamas.ac.jp/~aka/max/
aka.listen - http://www.iamas.ac.jp/~aka/max/
few examples of media documentation published on the blog
1 comments:
Hello Yosafat Gunawan,
I like your project and feel it very funny. I also want to know how do your doing the programming in MAX/MSP. So, could i get a copy of your project MAX/MSP patcher?
This is my email: bw670@hotmail.com
Thank you very much
Bochen Wei
Post a Comment