I have always wanted to build an underwater escape room experience and the unity assignment was the perfect foil for doing the same. When I started on the assignment, I knew that I wanted the following things:
Build a terrain of my own in Unity.
Play with lighting of the environment.
Create a responsive environment that acknowledges your presence.
Use the unity animator tool.
While my initial plan was to build an experience in which the person is trapped inside a ship with a shark trying to break in from the outside, it was out of my technical scope. So, I pared it down to something manageable that I could do with the finals madness going on.
Making the terrain in unity was pretty trivial but the models proved to be a huge pain. Initially, I used extra detailed models that I found on thingiverse and other places but they kept dropping the frame rate. So, I had to keep paring down my polygon models to make them work.
The schools of fish also proved to be a big problem. I had 7 schools of fish in the environment but they dragged down the frame rate. So, I had to write a script that they will be triggered only when I am near their start location. Then, I wrote one more script which makes them hurry away from you in a random direction when they come near you. Once the fishes were done, I focused on the sharks.
I created 2 parts of the shark. One was the body and the other was the mouth. The mouth was animated in unity to snap back and forth as it moves towards the player when one ventures into their area. (Basically, a big hit box triggers the shark to come towards you).
You can see the piece here:
The last assignment. And probably my worst. What I wanted to build was a box with a button and a title saying “How will your day at ITP be today?” and when you press the button, the windmill moves to a random answer.
However, with the final week madness and my P.Comp project being a tangle of wires and a mess of code, I didn’t have the mindspace or energy to make it. So, I probably pulled off the worst job of all time and used hot-glue, reclaimed wood and mounted the motors with basic screws to create something that works.
I did not get time to either program the random behavior or do anything else with it. The only thing that I did in this that I was remotely proud of was to turn the circular piece of wood. I call this piece “Turning it in“.
For this assignment, I had to use two materials that work together. I have always to build a lamp which is part silicon and wood. However, I have been working with silicon a lot lately and I decided to go with epoxy so that I could try out a new material for a change.
I wanted to interplay between the relative hardness of the 2 materials. Epoxy is fluid while wood is hard and I wanted to make something which reflects that in its form. After sketching multiple ideas, I honed in on this one.
I wanted to build a lamp which had wooden supports but had epoxy in the middle which lights up. Once I had a direction in mind, I got started on the wood and making the base. Wood from the junk shelf, an angled cut from the miter saw and finishing on the sander gave me 2 pieces which were perfectly matched in form.
The next step was to make holes. I made a through hole in the center for the LEDs to pass through and two small holes for the epoxy to flow and harden so that it doesn’t slip. (Remember Mark, you said no screws!) Once I had that, I tested an LED strip and pared it down to size.
The next step was to draw an outline and start forming the mould.
So, one would expect that it would go smoothly right? I had the mould ready, the epoxy in a glass and it was all looking good.
But I made 1 big mistake. The epoxy I chose doesn’t play well with the plasticine I used. The epoxy also undergoes an exothermic reaction which basically made the plasticine more sticky and ensured that my structure failed on me. I had taken a small nap after pouring the epoxy and woke up to the epoxy fluid seeping all over the floor in the shop and creating a huge mess everywhere. Thankfully, there weren’t any people around as it was 6 in the morning and I scrambled to clean it before John came in. I finished all the tissues in the kitchen and the loos to mop that mess up and had to walk back home smelling of epoxy fluid and defeat.
While the end of the process was a complete disaster, It was education in itself. I hope to repeat this all over again soon and make something that really works and sets.
So, this week out mission was to build an enclosure. (Rubbing hands in glee)
WARNING: Long blog-post ahead.
TL,DR: I build a box for my physical computing project. It was very pretty.
This assignment segued neatly into my physical computing project. for more on the project and it’s background, please go here.
The first sketch of the box was this:
Initially, I did not have any idea of the dimension or the scale of the box. So, the first stage was to finalize the puzzles and the ergonomic size which was in line with the theme. So, I did just that and finalized the puzzles and the tentative layout.
I knew that I wanted the box to be big but still fit within people’s hands. A quick test with people on different sizes on the floor and I fixed on it being about 3 feet wide. I also wanted it to resemble the control panels of old on the outside. The insides needed more careful consideration. The box had to be sturdy enough to handle people playing with it so it needed enough cross-bracing. Also, the top panel needed to be swapped in and out so that the box layout could be iterated upon and also adjust the circular screen in the center. Once the dimensions were fixed, the box was built in 2 phases:
Phase 1: Acquire the buttons, knobs and dials and figure out their mounting. This part was probably my favorite in the semester. I spent a good part of 3 days deciding on the buttons, knobs and dials and trawling through the depths of the internet in acquiring them.
Phase 2: Once the dimensions were figure out, the next task was to build the box. Initially, the idea was to use hardwood but we fell back on ply as it was easy to obtain in the dimensions we needed (and cheaper too). A combination of the miter saw, band saw, sander and voila!
Once the box frame was done, we spent a week or two play testing and getting the layout right. A week before the submissions, we started the final mounts.
All ready for the show! Let’s see how this goes <3
As mentioned previously, my ICM finals was to design a Pose-Karaoke experience. For my motivations and background on the project, please go to the post here.
While I had major issues with getting the ICM code to work on the ml5js platform, much of it has been rectified by the maintainers of the code-base and this current example solves pretty much all the problems.
But this was the easy part.
While starting the project, I did not realise the complexity of comparing 2 poses. A lot of what I had to do relied on being able to compare 2 images and it wasn’t a trivial problem. Trawling through the depths of the internet, I came across this post by Google research where they had worked on a similar problem. This post is a wealth of information on how to compare poses and it was outside my technical ability to be able to incorporate everything in my work. But the chief things that I could incorporate were:
1) Cosine similarity: It is a measure of similarity between two vectors: basically, it measures the angle between them and returns -1 if they’re exactly opposite, 1 if they’re exactly the same. Importantly, it’s a measure of orientation and not magnitude.
2) L2 normalization: which just means we’re scaling the vector to have a unit norm. This helps in ensuring that the scale does not play a factor in comparison and the 2 images can be compared normally.
The cosine similarity helped my code run faster and the L2 normalization ensured that the relative distance from the camera won’t play a role in the comparison.
Getting these 2 things to work proved to be a big challenge and once that was done, the comparison went pretty smoothly as seen in the video below:
I ran out of time to build a complete experience for the users which involve an engaging UI but that gives me something to do for the winter break. While I could not match the scope I had set initially, I am very happy that I could dive into algorithmic complexities and solve those issues to make something working. This gives me a lot of hope for the future and my coding abilities. All in all, time well spent!
For animation class, we were asked to read ‘Understanding comics‘ and reflect on what we have learnt. I have read the book many years ago and it was great to pick it up and go through it all over again after so many years. I had originally read the book before I started design school and I realised that I forgotten so many aspects of the book which emerged to me the second time around. Here is a list of my reflections that I noticed and picked up on my 2nd read-through:
The narrative is more immediate as compared to film. While we demand narrative coherence in film as we respond to the flow of ‘time’, a comic is free because it can move through time and space in a matter of few panels. I hypothesize that its one of the reasons why comic book plots don’t translate well on screen where the audience responds more to the flow of events across time rather than the space of a comic book.
The role of a narrator: I believe that the narration is the anchor which hold the comics together. Which is why I haven’t seen many works where the narration and visuals are at odds with each other. It will be interesting to see a narrative where the visuals and the narration start diverging and running at total odds with each other.
Panel to panel transitions: This was the biggest part of the book that I had totally forgotten. Scott Mccloud does a great job at explaining the various ways in which a narrative can work across time and space using the 2 dimensional paper grid. This got me thinking about the forms that I see back home. Is there an inherent structure to how a story manifests in a mandala or on the wall of an Indian temple? Do similar rules apply> I believe that there should be one but hopefully, I will find a book that talks about it in detail.
One of the biggest things that struck me while reading this book was that the constraint of the medium squeezes out the narrative style and structure. While the boxes might be seen as constraints by some, artists used it to tell their stories in unique ways that have now become representative of the medium. I wonder if there is a similar story with VR. While VR does not have any control over the user’s view-point, what are it’s unique constraints from which VR-only narratives will emerge?
For the after-effects project, I am teaming up with the mighty Cara. During our first brainstorm, I fell in love with the spaceman character that Cara had come up with and we decided to use the character in a museum dedicated to Earth after it has been destroyed.
An anonymous astronaut visits the Museum of Earth, the last surviving vestige of the planet destroyed in the water wars of 2050. The screens consist of short montages of what it means to be human. The animation cycles through 4-5 animations.
Pretty excited to see where this goes. After effects is daunting AF but we can power through this.
This week’s assignment was to make something out of acrylic using the laser cutter.
Easy peasy. I have been meaning to do a Voronoi lamp for the longest time and finally, it’s time to scratch that itch. So, off to p5 I went. Now, things are much easier compared to the last time I worked with voronois and guess what? there are libraries for that now.
After playing around with the shapes and size, I brought a few patterns that I liked into Illustrator. I played around with the dimensions of my box and tried to find a cross-section that won’t have very tiny shapes that might mess up with the laser cutter.
Once I had that, it was time to find the material. Now, my original plan was to have 6-10 colors but looking at the costs and the availability of plastics, I brought it down to 4.
I separated the colors into individual files for each color and off to the laser printer I went. Cutting the pieces was pretty uneventful except the part where I lost a nice slab of acrylic to the 75w printer which refused to work. (The cutting gods always demand a sacrifice) and within 45 minutes I had all my pieces. (easy peasy!)
Sticking the acrylic was a different monster altogether. The adhesive that I had was so runny that it was making my life miserable. But thankfully, Lydia got me out of a soup and loaned her rubber cement to me which made my life so much easier. And voila, within 2 hours I had a lamp!
All I need to do is find some LEDs to light it up and it shall be AMAZING!.
For my final in physical computing, my initial direction was to continue working with soft robotics. I wanted to explore the material more and create data-driven experiences that with softness, slowness and reflection as guiding principles. I brainstormed on multiple ideas and none of them felt satisfying. I was tired of being a one trick pony with silicon and nothing really felt like it was adding up to a meaningful experience. I spent a lot of time going around in circles till I gave up and focused on everything but Physical computing.
And that’s probably the best thing I did.
Not working on P.Comp gave me time and distance to think about it and combined with the happy coincidence of my friend Nun continuously going “I wish I could use all the buttons!”, it led me to a happy place that became the start of an idea that seems promising for the final project.
For the 2nd fabrication assignment, we had to make 5 repeating shapes of the same dimension. I was still hung up on making lamps and I wanted to combine wood and silicon in such a way that the 2 materials are interlocked with each other. My idea was to split the wood into multiple sections and then fill silicon between them as you can see in the figure below (the black area is the silicon):
I wanted to illuminate it from the bottom and the angled cuts appealed to me more. I came up with the straight cut option as a backup plan (2 months in ITP has taught me that at least!) and set off on my merry way.
I found a piece of squarish wood from the shop spring cleaning that I cut into wooden blocks using the miter saw.
The next task was to create equal blocks which was achieved using a pencil, ruler, miter saw and the sander.
I forgot to clamp the first piece and lost the whole piece as it flew away and smashed on the wall. Never forget to clamp the wood, kids!
Thankfully, I had extra wood and the breakage proved to be a minor inconvenience.
I traced the diagonal cut shape and went at it with a band saw and sander. I got the shape I wanted but problems were immediately apparent:
1) The sander eats through tiny pieces of wood. The piece on the middle-right became smaller than the rest in no time.
2) With the diagonal cut, it becomes very hard to keep track of the perpendicular surfaces and tracking the relative position to each other.
In the interest of time, I decided to go with the straight cuts and eliminate the smaller pieces once they are cut.
Measure and draw clearly marked go-no go lines.
Put your trust and full attention in the band saw and sander.
Cut small dowels and glue them in.
Wait for a few hours and voila! You have your wood shape.
I did not get any time to pour and cure the silicon but that’s next on the item list as soon as the class is done!
1) The first prototype sucks. ALWAYS. Everything that can go wrong will go wrong. In hindsight, I should have made one and then made the others.
2) The hardness of wood varies so much that it’s not always possible to predict its actual behavior on the cutter and the sander from the sketch.
3) MEASURE TWICE. DON’T MESS UP YOUR PERPENDICULAR SIDES. DRILL SLOW.
4) Sanding eats into wood fast and measurements go for a toss. The sandpaper is slower but you are in so much more control.
Bonus picture of leftovers:
I had quite a fortuitous moment that my project was to focus on machine learning while our final projects were due. I have always been quite wary of using anything that does machine learning as I don’t think I am don’t have enough programming chops but ml5js laid that to rest. Picking it up was so easy and getting posenet to work was trivial. It gave me the confidence to approach these algorithms and start thinking on those lines.
Back to my concept.
I have always wanted to work with the body as an input (Since I don’t move so well in real life.) Until now, most of the work related to body, gait and posture have been with a kinect or a specialized camera that is currently beyond my scope of technical ability. But posenet makes body-tracking so trivial that I realised that I could work with it.
The posture is so iconic for a person that people are defined by it.
And who hasn’t done this!
So i want to take the energy and goofy fun of dance games, karaoke and combine it with posture and movement.
My idea: People are given famous posture to imitate one after the other and they are supposed mimic it. The computer compares their posture to the original posture and scores them on the basis of it.
To get it working, I realised that I had to work with ratios and angle instead of absolute numbers. My explorations with the posenet library was to use it to calculate angles between various body parts but I was thrown off by the vector method in p5 which refused to work for me. So, here is a sketch of my failure:
For our first “Intro to Fabrication“ class assignment, we had to make a light-switch. The rules were pretty simple:
It should be portable.
It should create light.
I went through all the other weeks’ assignments and I realised that it was going to be the most open ended assignment that we have in Fabrication. So instead of working with wood, acrylic or any other material that is going to be used in the class, I wanted to use something that I will not be able to use again. I had some silicon lying around after my mid-term adventure and I was like “Why not?“. What I wanted to do was make a silicon shape that looked with no exposed electronics outside but it lights up once you touch or squeeze it.
As a first step, I cooked up a batch of silicon and added acrylic color to it to create a batch of colored silicon.
Adding the color to the silicon made it cure faster. Weird but I wasn’t complaining. Some That ITP friends pointed out a thai sweet called Khanom chan. The next step was to figure out how to cut it and that’s where I made a big, big mistake. Cutting silicon with a cold razor was a bad idea and I completely messed up the cuts.
Physically hurts to see this disaster of a cut. Ugh.
But Anyways, I soldiered on and cut a small air-pocket inside to fit the LEDs, battery and wire. the Leds were packed on top of a battery and i made a copper contact that hovers over the battery and gets connected when you squeeze the top. With some trial and error, i got it working and slathered some fresh silicon in a clamp and prayed.
4 hours later, It worked! Here’s the video:
Woohoo! Pretty satisfying except the shitty shitty cuts. Oh well, Onto the next one!
For our first meeting, we spoke about what we liked, enjoyed and wanted for our project. We both agreed that we need to make something which involved some fabrication as we had not done any of it till that point. We also spoke of our love for TeamLab and how we enjoy data turning into Art. But for Halloween, we both agreed (It was a pretty agreeable meeting!) that we both wanted to make something which was highly interactive and involved multiple people playing together. (It’s a people festival, after all!) One of the ideas that emerged was a wand-duel game recreating the Harry Potter-Voldemort face-off in Goblet of the Fire. It would look something like this:
Two players would grasp the 2 ends and the middle section will fill up with light which gets more intense with vibration and the brightness of the LEDs. The first person to let go off the wand loses!
We decided to head our way and think of more 2,3 or multiplayer games and meet again in a couple of days. Once we came back together and started finalizing the idea, we threw across ideas to each other but nothing seemed to stick. Shu-ju mentioned how everyone was doing scary things and it would be fun to just do a happy candy dispenser. We left the idea behind and started thinking of something else to do.
During a break in our brainstorming session, We started talking about weird and funny interactive objects and Shu-ju showed me this video: https://vimeo.com/52555492. We both were laughing through the whole thing and realised that we could actually incorporate softness in our Physical computing project and the idea of a candy dispenser that needed to be played with emerged.
We thought of a candy machine which dispenses candy only if you press it nicely. If you press it too hard, it gets cross, shrivels back and refuses to dispense any candy.
(It’s quite interesting to see how the final idea did not veer too much from the first sketches.)
Now that we had an initial concept in mind, we quickly realised that the sensor was the hard part. One idea was to put a force sensor inside a soft material and call it a day but we were not very happy with it. It felt like a cheap way out and we spent a day without moving ahead. Thankfully, Shu-ju was sitting next to one of the residents, Lola who had worked with silicon and using air pressure to measure intensity for her work in soft interfaces. (Wish, universe, law of attraction yada yada). We quickly booked an office hour with Lola and she gave us everything on a platter. What we needed to do, how to do it, plenty of encouragement and a loan of her sensors so that we could get started asap. Since we were still figuring out how to work with silicon, We raided the junk shop and found a discarded silicon tube. The thing was gunky, dirty and full of holes but a bad repair job with some hot silicon, we had a dirty sensor which looked like a dismembered finger (hey! it’s the flavor of the season) but it worked well for us to start coding.
Looks so gross, works so fine!
The lighting of the LED went on without a hitch and using an MAX7219 IC reduced the number of wires into the arduino to free up spaces for more wires.
This was trivial!
We were feeling pretty good about our progress as our temporary sensor and led array worked and it let us program the logic of the candy dispenser based on difference in pressure through the air pressure transducer (MPX5010 for life!) and we started our foray into making the actual air pressure sensor. We first 3-d printed a shape and then fashioned a cover to create the negative space for the air-pocket.
We printed a total of 3 times as we kept getting the wall thickness wrong. Each wrong silicon experiment set us back by 12 hours (considering the 3d-printing and curing time) and we lost a lot of time but we finally got it right and it felt so nice to squish it!
Once we had everything, Shu-ju quickly put together and enclosure and I got to work on the final coding and assembly. We had thought the structure through during the never-ending wait for our silicon sensor we really didn’t face any issue while putting it together. (Touchwood!)
The sound, mechanism for the candy dispenser and everything else fell together really well and we thought we were at the finish line all high and dry!
Or so we thought…
Because Murphy came visiting at the last moment and the nose broke with the sensor ruining the circuit (broken pins et all!) and we had to hastily put it back together at the last moment before the final presentation (Panic and duct tapes are a match made in heaven!). The thing worked but the sound wouldn’t work and it was finicky.
Here are a couple of videos which show it working (without sound):
The project went surprisingly smooth expect the last minute hiccup. It was a blast working with Shu-Ju and I would love to do it again soon. We have ordered all the materials we need for repair so I expect the poor box to live again soon. In hindsight, we should have started fabrication a bit more early to prevent the last minute rush job and relying on a functional form because we didnt have time to think it through. Overall, I was very happy with what we had and it was a great start to ITP. Onto the finals and Happy festival season!
For this week’s assignment, we were assigned to go through the video and sound lectures and then form a cohesive project around it. The videos went pretty smoothly this time (I think I prefer the visual aspect of programming over the DOMs and APIs) so I was feeling pretty comfortable with working on the assignment.
As my first idea, I thought of 2 ideas:
Taking a video and playing it from the end to the start while the sound plays normally. I specifically wanted to making a video play backward while the sound plays normally. Think Coldplay’s Scientist playing reverse to front while the audio plays normally (It would make a boring video but would have been cool to code.)
Splice the screen horizontally into 8 parts. Each splice plays the same video with a time difference.
Pretty simple right? This is where I ran into the first problem! To achieve my first intended result for idea 1, there is no way to play a video backwards in p5js. The other
First month in USA and I already fell ill. Not the greatest start to school and I was horribly behind on all the videos and assignments. But gradually, I managed to dig myself out of that hole and here is my combined blog on the Week 3/4 (digital and analog projects) for Physical computing.
As I was very short on time, I did not go with a big concept but chose to focus on making something which demonstrated my learning of the topics for week 3 & 4.
The project that I chose to work on was:
If you press the buttons in the correct order one after the other, the LEDs light up together.
If you don’t, the middle one lights up (Literally showing you the middle finger).
To make things more fun, I added in 1 motor which rotates a full 90 degrees when the buttons are in the correct order and gives a small shake when you don’t.
I did not run into major issues with the circuits after following the labs videos but the code to count the correct order of buttons was a bit tricky. My final code is a jumble of if/else statements but I think there is a more elegant way to do this. I shall speak to some residents and see if that can be done in a better way.
I did not get time to try out the speaker and tone (also, I did not have any speakers with me) but I am glad that I am not as behind as everyone as I was 1.5 weeks ago.
Onward and upward!
Currently listening: Drive-Incubus
For the first assignment of Physical computing, we were asked to interpret a switch and come up with a creative version of it. My first reaction was to make some kind of a Rube Goldberg like contraption but as all my tools and kit was still back in India (Stupid US shipping times! 😠) I had to scale down my idea.
While chewing on the topic to come up with ideas, I started playing a mobile game (The shitty kind whose name I shall refuse to take because of total embarrassment!) I remembered the Oasis puzzle in Legend of Zelda: Wind waker. That puzzle was emblematic of the countless hours I have invested in computer games in finding buttons and switches and I thought that would be a good start to my Physical Computing journey.
So, armed with the enthusiasm. I fist went and got a pattern, cut it out and stuck it on cardboard, stuck silver foil below the cardboard in a way so that the pieces line up when placed in the correct pattern.
Did this work? Yeah, kinda, sorta. the pieces slid properly once of twice before the sliver foil below ripped off and it was unworkable. Thankfully, I managed to get a picture before that happened.
I was really satisfied with the idea and the first prototype. I just wish it had worked a bit longer. I think it will make an interesting fabrication project and I would like to take this ahead when Intro to Fabrication starts next month.
Looking forward to the next one!
Currently listening: Switch-Will Smith