Monster in Progress…

Updated Wed Dec 1, 2010.

Our set up is pretty simple. Still there are some challenges, but we think we have feasible design solutions for them.

The Floor

We were able to quickly put together a prototype consisting of a 4×5 plywood platform with 4 force sensor resistors (fsrs), one in each corner. To get a finer range of data, each sensor is cushioned by a piece of rubber.

Challenges

1. We want to raise the platform off the ground so that a) you don’t need to crouch to get onto it, and b) the instinct is to lie down on it, not stand on it.

   To do this in a reasonably secure way, we’re hoping to mount the platform onto 4 upside-down milk crates that we will reinforce by sandwiching the bottom of each milk crate with 2 pieces of plywood.

   

   The mounting job is a bit tricky because the platform needs to be able to move up and down to register shifts in weight, so it can’t be firmly screwed into the plywood-reinforced milk crates.

2. Just the weight of the platform itself is maxing out the sensitivity of the sensors. As a result, we may need to use springs to keep the platform elevated above the sensors so that *only* the weight of the user will register on the sensors. I found these adjustable springs used to pop car trunks that claim to take up to 50lb in weight:

Here is a sketch of the spring and bolts and sensor on one of the legs of the platform.

The Data, The Code and The Behavior

We spent a bunch of time staring at graphs of the data coming out of the fsrs, both the raw data and calculated data.

We also spent quite a bit of time “choreographing” the kinds of movements we’re interested in and how they translate into a soundscape of monster noises.

I think we’re still undecided as to whether there is only 1 monster under the platform, or several monsters.

Right now, we have a basic set of variables that track the raw data coming out of the fsrs, the shift in weight in each corner, the total shift in weight, the % distribution of weight across all 4 corners.

We also have a handful of rudimentary controls to affect the audio with these variables.

1. Mapping a variable directly to gain. (Total shift in weight maps to the volume of growling.)
2. Adding to gain over time if the user sustains a certain level of activity. (If you shift your weight around quickly, you set off a giggling or laughing sound. OR If you shift your weight around deliberately and evenly, perhaps you can elicit a purring sound.)
3. Buttons in each corner that activate a specific sound. The buttons need to “sleep” for a while between activations so you can’t just keep firing them off.
4. Pain threshold. If there is a spike in shift in weight, the monster growls in pain.

We also have some ideas for sounds the monster makes when you get on (sigh) and get off the board or if you lie on it for too long without moving (yawn). Other sounds we’re thinking of using include: burp, mumbling, stomach gurgling…

Mapping certain variables like the rate or magnitude of shift in weight directly to volume gain results in a pretty expressive system that gives the user the ability to tightly control the quality of sound effects by varying the quality of their movement: acceleration, attack.

I’m also looking into creating resonance effects with environmental sounds (leaves, wind, rain) to create a layer of sound that users don’t quite have direct control over.

This may add a layer of complexity that enriches the sound experience, but it may also result in sound overload.