Playing with Fire


Dimitra Grovestine, 3165616

Melissa Roberts, 3161139

Kiana Romeo, 3159835


This project is a physical experience of a German children’s story, Die Gar Traurige Geschichte mit dem Feuerzeug. The story is about a girl who is left home alone with her two cats, plays with matches, and accidentally sets herself on fire.

The juxtaposition between the project and the story lies in the fact that the user must play with fire in order to experience the story, and the moral of the story is not to play with fire. Through the manipulation of objects related to the children’s tale (a cat, a matchbox, a picture of a little girl, a pair of shoes, a casket, and a shoe-box house) the experiencer navigates through the story without ever reading it.

Though the story is narrated, and that narration is triggered by some objects, the story is not exposed to the user. The narration is in the story’s original language: German, meaning the user hears the story being told, but doesn’t understand what’s being said. The German narration, the images pasted onto the inside of the roof, and the cues on the objects hint towards the story, encouraging user interaction.

concept_housefront     concept_houseback

Circuit & Sensors

Our project had six switches — 3 analog and 3 digital. The switches / sensors are attached to the above-mentioned objects, allowing the user to trigger sound by exploring the objects. Due to the complexity of the circuit (having 6 switches was difficult to manage) led us to solder rather than simply plug the wires into a breadboard.

Most of the sensors triggered narration, or narration with sound effects in the background. The shoes being placed in the casket and the cat trigger only sound effects. The fire sound doesn’t have a trigger, but plays throughout the whole experience.

Object List:

Shoe box

  • Sensor: Tinfoil Switch
  • Sound: Narration 1
  • Type: Digital


    • Sensor: Pressure fabric
    • Sound: Meowing
    • Type: Analog


    • Sensor: Pressure fabric
    • Sound: Narration 2/tapping
    • Type: Analog

Jar with paper doll

    • Sensor: Flame
    • Sound: Narration 3
    • Type: Analog


    • Sensor: Tinfoil switch
    • Sound: Narration 4
    • Type: Digital


    • Sensor: Tinfoil switch
    • Sound: Meowing/crying/water
    • Type: Digital
  • Paper dolls (10) → MELISSA
  • Barbie shoes → KIANA
  • Glass jars → MELISSA
  • Box of matches → DIMITRA
  • Ground pepper → DIMITRA
  • Coffin → MELISSA
  • Shoe-box house → DIMITRA
  • Cat toy → MELISSA
  • Tinfoil → MELISSA
  • Flame sensor → MELISSA



concept_circuit concept_objects


Challenge: background noise / creating “cat crying” and “fire” sounds/ sound recording in one channel

recording_auditionThe sounds we recorded were taken from the story, and included crying, meowing, dripping water/ trickling sounds, dancing feet and fire sounds as well as the most important aspect of the story: the narration. We used a TASCAM and shotgun microphone to record Melissa’s cat meowing, a condenser microphone to record Melissa’s narration, and a TASCAM alone for the remaining sound effects. To edit the sounds together we primarily used Audition and also used Audacity to edit more technical aspects of the audio like balancing channels and changing the pitch of some files.

One challenge was attempting to record with unwanted and generally unavoidable background noise. For example, when recording the water trickling sounds it was hard to simply get the dripping and splashing sounds without the sounds of people in the background or the mechanical hum of the tap itself that somewhat got in the way. This issue presented itself in most of our audio files, but was somewhat solved by splicing the audio and carefully editing out the sections with excessive noise as well as layering audio tracks over one another. With this we were able to effectively reduce the excess background sounds and get clean audio for the project.

recording_firesound1     recording_firesound2

Another challenge was creating a realistic fire sound effect. To create our fire sound, we combined the sound of a water bottle cracking, tinfoil being crinkled, and blowing into the microphone. By adjusting the volume and speed for each of these recordings, we were able to create a sound effect that resembled fire burning and crackling.
recording_audacityWith some of the audio there was the issue of sound mistakenly being recorded in only one channel. This challenge only presented itself in audio recorded with the TASCAM microphone as there is a dial on the device itself that very easily slides to one side with the slightest of touches. This challenge was fixed by simply normalizing the audio in Audacity and combining samples to create stereo tracks that played sounds equally in both the left and right channels.

Digital Sensors
Challenge: tinfoil trouble / sensitive casket switch

digital_coffin_oldswitch     digital_coffin_newswitch

We created digital switches for the coffin, the shoe-box house, and the shoes. For these switches we used tinfoil to create a larger surface area for the wire ends to meet. This worked out fine for the shoe-box and the shoes, however the coffin had some issues. The coffin had to be closed firmly in order for Maxuino to recognize that the switch was closed, and since we had sound triggered by the opening of the coffin (opening of the switch), the audio would sometimes be triggered even when the coffin was closed. To resolve this, we took away the tinfoil. The theory was that the wire didn’t have a tight enough connection with the tinfoil, causing the flimsy switch. Once the wires were bare, the switch became much more secure.

Analog Sensors
Challenge: if-statement noise / flame sensor (never used before) / trigger sound, continue playing

analog_changeThe analog switches caused some difficulty in Maxuino, as the noise from the sensors resulted in a buzzing sound — the sound of audio being triggered to start repeatedly and rapidly — rather than triggering the audio to play once. We scaled the sensor values to larger numbers, and included an if-statement that should have only allowed the audio to trigger when the sensor value was above a certain number. However, this only resulted in the buzzing occurring above a certain number. We thought perhaps the sensor values, even after being scaled, were floats, and the trigger required an integer. The if-statement should have taken care of that, as an integer or a float can be above a certain number, and the if-statement output is “0” or “1”, both integers. In the end, the if-statement also required a “change” object to properly trigger the audio.

analog_flamesensorThe flame sensor presented a challenge simply because it was new to us. We wanted to incorporate real fire in the project to better integrate the story in the user’s experience. We had never used a flame sensor before, and couldn’t test it during class (due to the school’s fire code), so it was unclear how the sensor would work. After testing it out at home, the flame sensor proved to be relatively easy. When no flame is nearby, the sensor output is 0, and the close a flame is to the sensor, the higher the value. The sensor doesn’t require any contact with fire, but rather, it can tell when a flame is in the vicinity.

Another challenge was that the sound triggered by the matchbox and flame sensor would stop as soon as the value didn’t agree with the if-statement anymore. For the matchbox sensor, the user must lift the matchbox, which puts pressure on the pressure-sensing fabric, sending a certain value that triggers the audio. When the user puts the matchbox back down, the audio ceases, even if it hasn’t played to the end. Similarly, the flame sensor only triggers audio when a flame is nearby, and stops before the narration is finished.

analog_sel1  analog_help

It took some research to resolve this issue, but thanks to this blog post, we found that adding “sel 1” between “change” and the trigger is the solution. What happens here is, once the sensor agrees with the if-statement, the if-statement output temporarily goes from “0” to “1”. The temporary “1” output it is transformed into a constant “1”, which turns on the “play/stop” trigger. However, when the sensor no longer agrees with the if-statement, thus returning “0”, it doesn’t affect the trigger, allowing the audio to play until it naturally ends.

Max / Maxuino
Challenge: tedious pin setting / file name issues / organization & clarity

Using Max & the Maxuino patch for the first time cam with it’s own challenges. It was a new program, and difficult to navigate. To begin with, though we had set up the file-path correctly, our sound files wouldn’t play. We double and triple checked the file-path, adding the “library” folder as well as “projects”. We also tried every name combination we could think of: “soundfile.wav” in the patch and “soundfile” in the folder, “soundfile.wav” and “soundfile.wav” in the folder, etc., and still nothing would play. In the end it really was just a naming issue: while the tutorial file had “open train.wav”, our patch wouldn’t play the sounds unless the patch said “open meow” or “open narration4”, without adding “.wav” to the end, and have them named like that in the folder as well.


Another issue was keeping the patch organized, considering we had six switches, each with their own sound file. Because this was a new program, it was important that we maintain some oversight, and know what we’re looking at, what function each little box has. This was difficult to do when we had a criss-cross of paths and a mess of boxes. We ended up organizing everything into sections, and lining those sections up horizontally. First, the three analog switches, then the three digital ones. We also “labelled” each section with the object it was getting input from, so that we could easily tell which switches were functioning and which weren’t.

Lastly, and this is less of a challenge and more of an annoyance: setting the pins. Each time we opened the file we had to set three digital pins and three analog pins. Then, we had to open up the patch six times, once for each pin. Though it just opened up the same patch over and over again, in order to receive input from all the pins, we had to do the following:

p digital INPUT → 2 → open
p digital INPUT → 3 → open
p digital INPUT → 4 → open

p analog INPUT → 13 → open
p analog INPUT → 14 → open
p analog INPUT → 15 → open

maxuino_inputs     maxuino_digital     maxuino_analog
Challenge: poor connections / soldering

Our project consists of a shoe-box doll house, which in itself is a switch, that houses all of our other objects/triggers. We chose to paint the shoe-box like a house in order to reflect the setting of the original story, which is the girl’s house. Images from the story were pasted onto the inside of the roof in order to provide the user a visual story-line and indication of how they can manipulate the objects to bring the story to life. Each of the objects is meant to be touched and moved, and most likely lifted out of the doll house, during the presentation. For this reason, we chose to house the Arduino and circuitry at the bottom of the shoe-box under a platform (or “carpet” flooring).

assembly_breadboard1 assembly_breadboard2 assembly_soldercircuit

Wiring was our main issue during assembly. The objects had to attach to the Arduino with relatively long wires due to the fact that they would be lifted a good distance away from the Arduino. When the objects were moved, we discovered that the connections to the breadboard were loose, and often the wires would become disconnected. To avoid having the project fall apart in this manner, we thought it would be a good idea to solder all of the wires. None of us had ever soldered before, so we used a full class to learn and complete the soldering. Not only does learning how to solder provide greater support and strength to the circuit, but with some colour-coding, our circuit became more organized and easier to understand (though it doesn’t look that way!).


img1     img2

While setting up for the critique, we had to make some adjustments to the project. Because of the use of fire, we had to present the project outside. However, being outside meant no power for speakers, and we had to rely on the laptop for sound instead of hooking up louder speakers.

The environment also affected the actual presentation, as the wind affected how easily the flame sensor read the vicinity of the flame, and how quickly the fire went out when the users attempted to set the paper doll on fire. 

img5 img4 img3

Some suggestions that came up during the critique were to use different types of cables to connect the objects to the Arduino, or perhaps to make the project wireless using Bluetooth. The lack of speakers also came up, and it was suggested that in future we try integrating wireless speakers into the shoe-box doll house, as the roof is about the right size to hold a speaker. Something else we might consider in the future is expanding the project, recreating more stories from Der Struwwelpeter. A particularly fun one to do next would be Die Geschichte vom Daumenlutscher (about a boy who sucks his thumbs and then gets them cut off) or Die Geschichte vom Suppen-Kaspar (about a boy who doesn’t like soup and starves to death).

critique_daumen critique_suppe


Der Struwwelpeter

context_book1     context_book2

The project has a sentimental tinge; it’s from a book that was read to Melissa as a child, and she always found it the most compelling story from that book. Though a grim story, it’s associated only with sweet memories.

The book is called Der Struwwelpeter, and is a collection of children’s stories. Die Gar Traurige Geschichte mit dem Feuerzeug reads like a light, sweet story, due to the rhyming, iambic pentameter, and repetition of the cats. The juxtaposition between tone and content made was why it was chosen to be reworked into an interactive experience.

Leave a Reply