Information Hypnosis – ICM Final

See the entire composition here.

This sketch is a commentary on our mind’s response to the excessive amounts of information it is forced to process. It is a visual emulation of the cloud-gazing experience where one keeps staring at clouds thinking about everything and yet nothing. The user’s experience is contained within a geodesic dome, isolating him from his environment.

// add experience video

The code uses information from an arbitrary API to make the clouds move. In the first iteration I used weather data to control the motion of the cloud, that I later removed because the data was constant and stopped the clouds from actually moving.

P.Comp Final – Project Documentation


An exploration of smell and sound, our project encourages students of Indian classical music to practice singing by filling the room with a delicious smell when they hit the correct notes. “Smell is subjective, it’s ephemeral, and it’s not binary”. Our project explores people’s perception of sound by associating it with smells.



Construction of the Box
For each note pressed, the box begins to listen to the singer’s pitch. Based on how the singer performs, the box projects a scent.


The Scents
We used two distinct scents that are easily recognizable (or familiar) – so the user doesn’t spend time familiarizing himself with the smell. This way they can focus more on the experience. The first is a spicy, earthy scent ant the second, Lemon essential oil.

3 Scents

Project done in collaboration with Dhruv Damle & Jen Kagan

P.Comp Final – Programming

The biggest technical challenge in this project has been learning and executing our project in MaxMSP. Many trial-and-errors and a long learning curve later, we finally got the pitch detection to work on Max.

I used the object ‘retune~’ with its ‘@pitchdetection’ attribute to get the frequency values. Once that was in place, I used if statements to limit the frequency detection to the acceptable frequency ranges of all the 8 notes. I then added ‘on(1)’ and ‘off(0)’ to each note range. Next step – serial communication between Arduino and Max. Max sends out these 0 & 1 values in the form of messages to Arduino based on which two leds (eventually smell dispensers) would blink.

Screen Shot 2015-12-07 at 1.16.05 AM

After this iteration, the next step was to detect the selected note only. I added 8 separate gates for each note. These gates would allow one to switch on and off the process of pitch detection. Messages from Arduino (switching on a button) will travel through these gates and start the process. There is a two way communication : Buttons open gates (arduino to max) and pitch detection triggers the smells (max to arduino).

Screen Shot 2015-12-07 at 12.14.24 AM


Special thank yous to Justin, Jason, T.K, Gabe and Matt for all their help.

P.Comp Final – Testing Smells

Last week, we hacked into the air freshener based on the Scratch and Sniff TV mechanism – and decided to add our custom bottles and scents for it.

One of the key components of this project is how the user reacts to the smells – which is what we will be testing in tomorrow’s class.


Some months ago, I stumbled across Fragrance Shop New York in East Village where I met Lolita, the owner. I spent 45 minutes in an engaging conversation with her about the history of her shop, her occupation, the power of smells, what they can do, her travel stories and much more. I went back to her store again, this time to find 2 distinct scents for our project.

The task : to find two distinct scents that are easily recognizable (or familiar) – so the user doesn’t spend time familiarizing himself with the smell. This way they can focus more on the experience.

After an intense testing session – I zeroed down on 3 scents.
1. Pink Martini – a citrus smell with a mix of grapefruit, oranges and lemon
2. Woodstock – a spicy, earthy scent.
3. Lemon essential oil – a condensed lemon smell

3 Scents

To make these scents project more in the air, I bought a cologne base which is essentially an alcohol that is to be mixed with the scent in a 60:40 / 80:20 proportion.

Cologne Base

Next, we needed to get some spritzer bottles that would fit our scent dispenser framework. We found these at Muji.

Muji Bottle
In the Muji Store we also came across some Aroma Diffusers that were noteworthy. The interesting feature is that one can see the smell being dispensed (in the white foggy form). It can work as a good and subtle visual feedback.