SoundStage is an ambisonic surround sound mixer easy enough for a three year old to use. By moving “sound objects” around the table, users are able to pan audio around the room and immerse themselves within a soundscape of their own creation. For example, take a classic like Old MacDonald’s Farm. Depending on where children place a cow on the table, they can make a “moo-moo here, a moo-moo there, here a moo, there a moo, everywhere a moo-moo.”

My goal is to make surround sound fun and accessible to kids.

Background
The ReacTable project was created by a group of student's from Spain for their Master's thesis, and is a music synthesizer that uses a tangible interface where physical objects are being tracked on a table. As someone manipulates those objects they are able to control different parts of the synthesis. It is quite advanced and very expensive and Bjork even uses it on her tour. Thankfully the group makes the underlying framework available opensource. There are lots of multitouch tables being created and gestural computing seems to be very popular on the floor at ITP. However, I was interested in using a tangible object as an interface for kids because of the intuitive nature of the interaction of placing something on a table.

Audience
Children ages 3-8 (parents and adults have a lot of fun too!)

User Scenario
There are two main ways that the SoundStage is used. The first is a more guided experience led by an adult. It is very similar to storytelling in the library or ghost stories around a campfire. The main storyteller is leading the experience, but asking for participation from the children, who interact with the table one at a time or in small groups. So if the story involves a pig, a child will be asked to come find a pig out of the group of sound objects and place it on the table. Because of the spatialization of the sound, you can make object louder or sound further away, by how close they are to the center of the table. Additionally, you pan sounds all the way around the room, and there is height information built into the objects, so that things that fly sound above the listener, and small animals can sound below. Another aspect that adds to the experience is that many objects can trigger different sound effects based on their state. So when Thomas the Train is placed on the table, you hear a toot of his horn and some steam releases, but when you start to move Thomas he slowly begins to Chugga-Chugga, which speeds up as you continue to move him around the table. Adding to the realism of the experience is a built in doppler effect, which pitches up the sound as you approach the center of the table and pitches down as you go away from the center. Thus, it can sound as if Thomas is zooming by you at full speed.

The second way in which the table can be used, is in more open ended, but supervised playtime. No activities, just exploring all the sound objects. This is better done in small groups.

Implementation
I am using a small end table that is 2'x 2' and has lights and a camera underneath the acrylic surface. For the computer vision I am using the opensource framework developed by the ReacTable project, which was created as a tangible interface music synthesizer. Each object is tagged with a fiducial marker and I am then using their coordinate position to place the triggered sound effect into a 3D soundfield. This is all done in Max MSP and I am using the Institute for Computer Music and Sound Technology's ambisonic externals for sound spatialization. So when a user places a horse on the table, they can hear a Neeehhh. Where they place the object is where the noise comes from, so if they place the object on the back right corner of the the table, the sound emanates from the back right corner of the room. As the start to move the object the sound start to pan around the space.

There are 8 speakers arranged in a cube which allows for height localization. So birds and planes can sound as if it is above the user, and small animals can sound as if they are below the user. I am also using distance cues, so that the further away from the center of the table an object is, the more distant the object sounds in the soundfield.

I currently have a working prototype with around 70 "sound objects" which are divided into three categories: animals, transportation, and musical instruments. I did user testing at the New York Hall of Science with preschool age kids as well as a 2nd grade classroom of 22 7 and 8 year olds. I will be showing SoundStage at the Music Technology Open House on Saturday May 14th, 2011 and the ITP Spring 2011 show.

Conclusion
Based on my user testing and watching others play with my project, I learned that sound can be a lot of fun and kids like making lots of noise. When I set out to build this project, it was tied much more strongly to narrative storytelling and through the New York Hall of Science, I took a look at how this could be used as a platform to learn about the science of sound. However, in both of these cases, the situations felt forced. Getting kids to try to answer questions about their understanding of sound, never quite felt comfortable at NYsci. I am sure this had a lot to do with my lack of experience teaching kids this age. Also, the storytelling was almost entirely driven by the sounds kids wanted to hear. The electric guitar was the most popular sound, so the narrative round-robin kept coming back to someone starting a band with an electric guitar. But I learned there is nothing wrong with letting kids just play with sound. Inevitably, they would want to hear as many sounds as they could fit on the table.

I learned to embrace the chaos and let the kids have fun. Rather than teach science learning or make a tool for more immersive storytelling, I found I really wanted to create an environment for a unique sonic experience they couldn't have anywhere else that would inspire kids and hopefully interest them in the world of sound.

Overall, I think I was very successful in making surround sound accessible and fun for kids.