Uncategorized

Visualizing Sounds in a Day

A physical manifestation of sounds

I have always been excited about how sounds affect human beings. Be it the irritation caused by a noisy traffic signal or the soothing effect of orchestral music, sound does have a relation to the emotional self and hence it has kept evolving ever since the existence of mankind. For this reason, I decided to have my take on sounds that I hear during a complete 24 hour cycle and visualize it in a physical form.

 

Process

  1. Gather audio recordings around a 24 hour span and keep a record of the activities I did during the time. Complete 24 hrs of audio was recorded for 3 days. Audio for day specific activities (like eating out, going to a new place etc) was recorded in separate chunks. Audio was recorded using my phone (One Plus 3) using the “Smart Voice Recorder” app.
  2. After collecting over 80 hours of audio samples over a span of 7 days, I loaded them up in Matlab for performing signal processing. I analysed the audio using Fast Fourier Transforms over time domain to get average values of amplitude at any given time. This gave me the average amplitude of noise levels that I am exposed to during a typical day.
  3. Using the values from the above process, I wrote a program for plotting these values visually with respect to a circle. This circular spectrogram represents the amplitude of sound in the time domain of 24 hours.
  4. The circular spectrogram was then imported into illustrator to export it to a laser cutting machine. I used DoIIIT lab’s laser cutter to carve the spectrogram on an acrylic sheet.
  5. I used LED lights to light up the artifact to make it aesthetic. The LEDs are powered by an Arduino and a portable battery pack.

 

Algorithmic Reproduction

This visualization consists of a series of steps which needed some logic to weave together. Averaging out chunks of similar data and joining the various pieces into a signal of 24 hours was the challenging part. Although this can be done programmatically, it requires a detailed listing of the type of activity, environment and time to get the accurate values for noise levels. Also, exporting it in the form that it can be laser cut can be done algorithmically. I feel that this can be converted into a programmable flow given that deviations need to be adjusted as per the activity the noise level corresponds to. Not only does this piece look very attractive, it can be used as a clock or ambient mood light. This also carries a personal attachment to the person it is made for, adding an emotional touch the design.

Leave a Reply

Your email address will not be published. Required fields are marked *