The first round of the project involved a Max patch, an Arduino, and a measurement microphone with an LED attached to it.     As the microphone is moved around the room, it senses changes in sound pressure level and sends them to Max. Max maps and sends this information on to an Arduino, which controls the PWM of an LED that's attached to the microphone. Manipulating the PWM can change either the brightness or hue of the LED. In low lighting then, I'm able to move the microphone around the room and capture these changes with long exposure photography. Initially, I just controlled the brightness of a single color LED depending on the change in amplitude. 
  Here we're looking at how 282Hz from my monitors interacts with my desk and later the parallel walls of my living room.
 Phase cancellations across my living room, mapped with a single, red led.  
  These next pictures were done with a tricolor LED. The green streaks are analogous to the absence or fading red light in the picture above - which is where there's a loss in volume. The red light in these pictures aims to represent the antinodes or where there's a higher amplitude, the green light the node, and the orange color represents the transition between the two points.     
  The photography has it's own challenges and it's been a while for me since I've thought about aperture and lighting and so forth. The nice thing about these photographs is that you can actually see the room as opposed to seeing a computer representation of the walls and where standing waves might occur. So, the best photos seem to be like the previous pictures that aren't in an entirely pitch black room, like this one.    
  Additionally, the current setup is so dependent on how you move the microphone around. A single trace from one wall to the next makes it easier to understand what's actually happening at first, though we're just looking at one dimension. The goal would be make these traces more three-dimensional just as the reflections I'm attempting to map are. Moving the microphone in a consistent way along two axes begins to give a better sense of how sound is moving throughout the room. 
 ROUND 2 !   Finally, I started playing with a larger color spectrum and fixed my coding so the LED faded smoothly across the spectrum. This was challenging at the time since I was controlling the LED with one signal over serial and the RGB LED has three values that need to be controlled. I leaned on this graph enormously and learned a bit about additive color mixing with LEDs. 
 RGB mixing
  The following pictures were taken at a friend's loft, which has a cross-gable style roof and 6 speakers installed in the walls. The acoustics were obviously much different than in my apartment. Since I'm more or less new to the world of acoustics, I'll hold off on making any truly objective scientific comments for the time being. Here are the speaker locations and layout for reference and again, refer to the graph just above for what the colors mean. I really should be including some decibel reference for each hue but basically it's magenta for the lowest amplitudes and red for the highest.     
 More speaker locations.  
 The purple sections here indicating the lowest amplitudes, and the red sections indicating the highest.  
 ROUND 3!    While manipulating hue with amplitude allowed for a lot of resolution in these photographs,  some  have pointed out the obvious error in using color to represent amplitude - as color is itself a representation of frequency, not amplitude.    Also, in my attempt to move the microphone along two axes, I ended up producing shapes that were way too similar to a sine wave and left most people confused as to what they were looking at, or at least assuming that somehow these were actual representations of a sound wave at a particular frequency.  They are, but remember, sound waves travel through air as longitudinal, not transverse waves.     After getting some more feedback on the project and seeing some of the great pictures being done with the   Pixel Stick  by Bitbanger Labs  I started reworking my setup to include a pixel strip and an array of microphones.  At the same time, I began moving away from Max and my preamp and instead built up a few microphone preamplifier / peak detector circuits to drive the Arduino ADC pins directly.  The result is a meter long LED strip which is segmented depending on how many microphones are attached, and illuminates growing outward from where each microphone is located. Essentially, it's a big VU meter with multiple inputs, but more accurate would be to think of it as a persistence of vision Rueben's tube.    I'm still only working with two or three mic inputs at time while I refine the code and improve upon the circuit, but the photographs seem much easier to decipher than the previous round.  
 A new board!  A four channel mic pre, precision rectifier Arduino shield.  
Breakout for Teensy
prev / next