Final Project: Ghost Baby Presentation P.2 & Reflection

We presented our final project today, after adding the finishing touches to it. We decided to project the clip of the Ghost Baby onto a large projector and have Katlynn, our demonstrator for this presentation, do a short demonstration at the front of the class. We also turned the lights out to add to the creepy atmosphere.

I think we addressed the main issue in our project: to add to what we initially had and make it have more baby-like characteristics (and at the same time straying from literal baby noises). There is a short clip of the final project being shown on Stephen's blog here. (Realplayer seems to be able to handle it, and not much else).


I really enjoyed working with this team. We were well-balanced, and everyone did their work well. Using both the Arduino and accelerometer was a great learning experience that I would like to further explore in the future. Thank you, team!

Final Project: Ghost Baby Enhancements

The team has decided to go along with the plan to add visuals to our project. We have filmed a short clip of the Ghost Baby (right now an ugly plastic doll picked up at the dollar store) with the camera coming in at different angles -left, right, up, down, and middle (middle being the camera is steady). The Baby has two states, angry and happy, and spikes/dips in accelerometer output can trigger either state. The directions refer to the directions we're giving the illusion of; we plan on analyzing the numerical output of the accelerometers and assigning a video to each one. This way, we get clips that look as if the person interacting is actually interacting with the Baby instead of straight out controlling its actions.

While Stephen focused on some MAX/MSP programming, Oliver, Katlynn and I worked on the videos. We separated the long video footage into the separate directional chunks and added numerous effects. We made each one black and white, and utilized effects that made the Baby's head look bulbous, made the Baby blur and increased the contrast so the haunting, creepy feeling could be achieved.

Also, we fixed the wrist band so the Arduino and accelerometer were more secure. We put the Arduino (with the accelerometer wires happily attached to it) in a separate sturdy pouch that we affixed to the wrist band. We put the accelerometer in a small, conveniently located "hole" (rather, a small square depression in the wrist band) and secured that down too.

To see a sample of what the Ghost Baby video looks like, check it out at Katlynn's site, over here. I certainly hope it's creepy -it is a ghost, after all.

Greg's Comments

Here are the comments Greg made in an email to Oliver:

"I like the fact that you have stayed away from literal sounds and I'm intrigued that you have based your sounds on baby noises. It might be an idea to let the uses hear more of the original sound. Right now the word baby refers abstractly to a virtual collection of sounds. by giving hints of the original sounds the virtual space may take on more characteristics of "baby".

As I said in class, the main thing I would like to see you work on is stabilizing the space more. this means reducing any latency of response or drift of coordinates."

___

As I stated in my previous post, having something to support the sounds would make them more definitive and set more of a space. I hadn't considered the childlike characteristics of the Ghost Baby (though I should have, considering the project title). Maybe adding a baby-themed video can help here. As for coordinates and latency of response, I'm pretty sure we can work these out after a lot of programming (with the coordinates, this is likely trickier).

Final Project: Ghost Baby Presentation P.1

For today's lab, we presented our final project to the class. Before coming to class, the team met up early (10 am!) and worked pretty heavily on the Ghost Baby. For the most part, Stephen and Oliver worked on getting the Ghost Baby set up, responding to the gestures and playing the correct sound; Katlynn and I worked on getting the sounds, editing them, and testing/debugging the Ghost Baby. What was really neat was getting the Ghost Baby to respond from a very far distance (we had the project set up in the open lab and I went downstairs right below this area to see if it still worked -which it did).

As the sound designer, I took my cue from Oliver and went for distorting the baby's sounds. Simply picking clips that sounded like a baby would be way too literal for our project -we wanted it to be more abstract. The sounds were imported into Audacity and modified heavily; we used many different effects, such as amplify (to make it sound more scary), reverse (this reverses the waves' pattern completely) and fade out (brings the waves down gradually so the sound gets softer).

As the demonstrator, I had to show a large range of the Ghost Baby's reactions. From wide motions to small ones, each gesture had a unique sound. Before the presentation, Oliver gave me a quick set of instructions of what to do regarding how to get the aforementioned range. When Greg asked, at the end of the presentation, if I thought it was intuitive, I was rather surprised at the question: I'd worn it so much and did the gestures automatically without thinking about what kind of steps I needed to follow for it to work. Thus, it must be a pretty intuitive device. Maybe we can get it tested on someone who doesn't know how our project works beforehand and have them figure it out.


To improve our project, the following suggestions were raised by Greg:

1) Have MAX/MSP get the readings of the accelerometer's position in space.
As Oliver mentioned during the presentation, we have tried this before but didn't get it working. It's hard for us to program these and hard to get proper output from the acceleromenter that are stable enough to use. However, we'll try it again for presentation number two.


2) Take out the delay time of the 'boom' sound effect when slapping the Ghost Baby
Greg noticed that when I performed a slapping motion (from the top left overhead to the bottom right at about hip level, as similarly demonstrated in the first image), there was a bit of a delay between the slap's sound effect (onomatopoeiacally, a "boom" sound). I'm not quite sure why there is a delay, but the team will work on it later.


Sound-wise, maybe the team should work on making the sounds more clear...?? (They weren't car crashes!). This is a minor issue though...I think that the sounds would be better defined if they had something to support them (like visuals, maybe; just a thought).

On a side/frivolous note, I think we should perhaps make the gauntlet-glove a little more secure and more aesthetically appealing. Currently, we have the gauntlet (really, it's a rollerblading wrist protector) with a sock overtop of it. Resting snugly between the gauntlet and the sock is the Aruduino and accelerometer. We have electrical tape keeping the sock attached to the gauntlet. To make it more secure, we could have the electrical components encased in a black, rectangular plastic box and tape/glue said box on the gauntlet.


Also, it has been mentioned that it was a good idea that we abandoned the H-bridge/motor idea. It would have probably put too much pressure on the motor, and the elastic thread would have snapped. Also, it is counterintuitive to our idea of having large gestures.

Here are some images of the project:

This is me testing how high it can go.


And this is a close-up of the controller.

H-Bridge & Motor

For today's lab we focused on wiring up the H-bridge that would enable our motor (to pull the hand back, of course). We put everything together, put all the pins in and tried testing it. Nothing. Puzzled, we looked at all our connections and tried to figure out what was wrong. We re-plugged but nothing worked; we decided to ask for help from Greg, but that was of no use because everything should have, theoretically, been working: each pin functioned, the wiring was okay, etc etc.

The result? We have decided to try putting off the H-bridge for now. We'll try some more problem-solving later (asking for help, rewiring yet again, etc), but the decision to put down the H-bridge is looking rather final.

Meanwhile, I've started collecting some sounds for the Ghost Baby. It's going well -I don't know what I'd do without those sound effect library cd-roms up at the library. Hopefully, we can get a large range of sounds we can pick and choose from.


Edited to add: We spent extra time talking to Greg and trying to get the H-bridge to work. The H-bridge rather thinks we're asking too much from it -again, it refused to function as it should. Thus, we've given up the H-bridge/motor idea. We've also found that Arduino is not responding to any of the code we're feeding it (not even when we run the built-in LED code and make very small changes). We're working around this idea by simply sticking to MAX/MSP.

Block Diagram

I've created a (very quick) sketch of how I imagine the interaction of our project to be like. This might be wrong (I'm certainly no expert at drawing block diagrams) and it might be inaccurate (we might change the project to fix problems) but this is the skeleton of its functions.

Here it is, in all its crooked-line glory:

Merleau-Ponty

Merleau-Ponty makes interesting arguments about the body and the senses in his work, "The Visible and the Invisble". He mentions that the senses of touch and sight are interchangeable; we can use our sense of touch to "see" objects and our sense of sight, our gaze, can allow us to consciously know of the objects’ existence, to almost tangibly “handle” them. What is especially interesting is his theory on the body as a chiasm. It is the way we are able to “[cross]-over” and it bridges the gap between “subject experience” and “objective existence”. To me, I felt this referred to our own, however flawed or wonderful, experiences in life regarding an object and how this influences (or is influenced by) the object’s simple, unbiased state of being.

Embodiment, I believe, is an example of this duality, as proposed by Descartes and summarised for us in the first week of class. As Wikipedia defines it, embodiment is “the way in which human (or any other animal’s) psychology arises from the brain’s and body’s physiology”. So, the person’s experiences and mindset about something are based off the brain and body’s (objective) “mechanical, physical, and biochemical functions”. In our project, we can try to further make this link clear. We can try to find an isolated experience and try to make it, somehow, independent of its physiology.