interaction lab @ wdka

I have been trying to draw the bodies at the bottom of the boundingbox but I think I am overlooking something. Right now I have the bodies drawn on top of the boudingbox and trying to offset them with all kinds of combinations with this code:

somehow I always end up beheading my characters, at one point or another, when I move too far away. I’ve even had the bodies suddendly upside down. Anyone any suggestions?

Another example of weird characters:

Ok just a backtracking update…

I finished porting the Johnny C Lees Framework to openframeworks and OpenGL last week. Have to say that it was pretty straightforward as far as OF is concerned. I didn’t incoporate the ofx3dutils addon til now, simply because Lee’s datastructures are more complex and i want to leave in the possibility of using the wiimote.

First screenshot after the final rewrite

I also got simulated viewers implemented which i will post next when i got the screengrabbing working. At the moment what happens with those viewer is that the view gets recalculated from left to right as often as possible with the intention to get to images (or more) at once through the persistence of vision, which is going to play an even bigger role when you are projecting it on a big screen but more about that in a later post.

first steps in the algorithmic architecture department

For the simulation i will use more objects and they are going to be based on primitive three dimensional forms… with which im still experimenting.

another view of the same thing just to give you the angle

Anyway to sum it up, im currently working on:

• using the blob-detection instead of simulating the viewers (practically finished, will upload an animation this weekend)
• incorporating the tracking and the GL-View into on application so that i can load the example (people walk into the installation setting) and at the same time see and manipulate the 3d-view behavior… and thanks for the FBO example because that is absolutely the right way to do these kind of things.

ok, thats it for now.

55 mins - 100mb http://lab.impssble.com/videos/ParticleClassTutorial.mov

I’ve got the bodies under the heads. the problem now is that you behead the characters when you move to far away.

The hands (Butterflies) can now be stored, and drawn out of an array. They are drawn on variable coordinates, so they move around freely. I will use simplexnoise to make things move organic. To differentiate it now gets random colors.

the shark with thanks to jurjens animation skills!

its divided in pieces but we’ll need to test if the blurry black underneath every piece is convincing enough.
i also made a sound scape what is built up from 3 loops. for the different areas.
heres the first draft!

I just made a small change to my video labyrinth. Instead of having both a sound labyrinth and a video labyrinth, then i will just put the sound and video together in aftereffects and then i only need a video labyrinth. And that looks like this:

so that is what i have been doing!

And this is without the bird…!

I’ve been trying to connect a soundFader to the center of mass and to change when entering a new zone,
I got really stuck and Alex is also taking a look. Besides that im busy making sounds for when the installation is working, the design of the shark and a testvideo we could work with.
I’ll soon update them but i’m having some uploading problems.

Hi everyone,

here are a few stills from what i’ve got untill now. at the moment i’m trying to get the bodies to stick to the heads. i’ve thought of some ways to do it and where to write it in my app but it’s not working at the moment so i’ll have to look at it a few more times;)

Hi, I’m sort of better again… and here’s a little thingy I worked on today. It’s a test of some water like structures I did, to find out the best way to get a watery surface. The framerate of the video is a it messy, but good enough to get the point across. I’ll clean the code a bit, and put in some comments, and post that too.

On another note, I’m pretty much stuck with the particle fish thing. As yesterday was the first day at could look at code again without my head starting to twist at the sight of small letters, I haven’t put much effort in it, but at the moment I don’t see exactly where to go from here, so I’m looking forward to tomorrows particle class…

A little inspiration:

http://www.insertsilence.com/paganpoetry/

So, I’ve been in bed all week, flu got me pretty bad. Sadly it also means that there hasn’t been a lot of progress. I’ve been working on a particle tracking a point, combined with a simplex noise number, which is currently working. Also, I’ve been trying to give it a solid sort of “tail”. Which worked to some extent, but not the way I want it to work…
Anyway, here’s a little video.

The framerate is a bit whack, but it’ll get the point across. Goal of this whole thing is to get the shark to make more realistic looking movements towards the target, and not rush in there in a straight old line.

I’m off for tea and bed, goodnight all

edit: can’t seem to get the vimeo thing to work…

Alex

We’ve got incoming messages into our OFapp!

Me (Peter) and Jurjen decided to work together,
because we both couldn’t find the “working drive” to complete our projects.

As we both we’re talking about our projects being too ambitious,
we thought working on one project together would solve a lot of problems.
After a brief discussion, we agreed on working on Jurjen’s SMSproject.

So, to get things going we started on trying to receive SMS messages on a computer.
Since the “SMS Receiver” didn’t work, we scanned the internet for other possibilities.
We switched from PC to Mac after we’ve found an amazing program, BluePhoneElite 2:

After some settings, we’ve received our first SMS Message:

Quick after that, we figured out that BluePhoneElite 2 is able to set Reactors.
Which is something like an “if”-statement.
So, we set it to export every incoming SMS to a .txt-file:

What happens now:
You send a SMS.
BluePhoneElite2 receives it and open the Message + a “Export to Text File”
But you have to manually input the name of the TextFile + click [Save].

As a solution to that, we’ve found out that Mac comes with Apple Script.
Which is designed to automate every single action you want it to do.
So like opening files, closing files, hiding windows etc.
One thing we couldn’t get it to do is: Input a name and Save a file!
So, we’re looking for someone who is able to write a bit of Apple Script! (Theo?)

We’re now working on importing the exported .txt-files into OpenFrameworks.
After that, animations and the interaction-part is going to be designed.

Blob tracking
Persistent blob tracking example - lets you track blobs over multiple frames by assigning a unique id.
For 005 Xcode - http://lab.impssble.com/files/misc/blobTracker-XCode-005-FAT.zip
For 005 CodeBlocks - http://lab.impssble.com/files/misc/blobTracker-CB-005-FAT.zip

Advanced! - FBO - off screen rendering

This lets you render your openFrameworks calls into a texture or even multiple textures - allowing you to split a projection over multiple displays as well as perspective correct a projection. Has an effect like this: http://vimeo.com/806183
http://www.openframeworks.cc/files/FBOManagerExample-Xcode-00573.zip
http://www.openframeworks.cc/files/FBOManagerExample-CB-00573.zip

Addons
There are a lot of new addons coming out for OF. You can see a list of addons projects here: http://addons.openframeworks.cc/projects. These include addons for doing 3D, physics and fluid simulations, even using the sudden motion sensor in your apple laptop.

Awesome project called: NeoVisus Gaze Interaction
With this program you can control an Interface, just by looking at it.

openEyes

As we talked about before via email…

Technologically the idea is a mixture between this http://www.youtube.com/watch?v=Jd3-eiid-Uw and the infrared head-tracking you started last time. At first it seems only as if i would use a different technology to make the screen bigger (which from my personal wow factor is nice already). But as he said, his technology and concept are not able to track two people….

but this is exactly where it gets interesting, not only would i be capable of tracking two or more people, i want to make the conflict between the different POVs the theme of the work…

so the idea is to start out with a single person, probably only one (or few) three-dimensional basic shape(s), black background… when a another person enters, the object from the perspective of one person morphs into the perspective of the other person, but since both are still there a new object (the synthesis of both perspectives) is created, the morph is less of a new object but much more a visual oscillation between the two (inspired by the gantz graf video but less shiny, less shading, less NURBSy, but i can also imagine a more awkward jodi game-mod variant of polygon dumps, i am less on the slick side)… i have not too much details here, because first i want to figure out of course if that works at all, and mental images always get altered by reality. The sound is going to be based on a granular synthesis were different grains (whose size and distance is defined by the positions of the people being tracked) fight for volume.

A change to the project I’ve made is adding an extra camera to the space,
for motion-detection on the entire bed.

The procedure will now be:

Step 1. Set the alarm. And go to sleep.
Step 2. At the set time, the alarm will go off.
Step 3. Look into the Webcam #1 above you, to silence the alarm.
Step 4. The alarm will go off every 7/10minutes.
Step 5. Once you go out of bed, to turn of the alarm, a question/puzzle/task will be there to be completed.
Step 6. If you complete the task, Webcam #2 will check if you don’t crawl into bed again.
Step 7. If so, go back to “Step 3″.
Step 8. Webcam #2 will be active an hour after you completed the task at “Step 5″.
Step 9. Get yourself a Cup of Coffee/Tea!

For this class we worked on making a test video of our workarea.

I filmed Vinesh awalking from side to side on a white background and he started twiching at somepoints for no obvious reasons.

This we can now use to create a bounding box that serves at first as a simple blob to be targeted by particles. Later we cant to make the particals into 2d kitchen objects and finetune the blob into a more complex contour for the object to stick on.

As explained we need to forge simple appz that solve part of our puzzle, so making the shape more complex and an example to show the attraction treshhold.

Since we’re also doing another project that’s related to parabolic speakers, we figured let’s do this one together - and simultaneously work on the Harmonograph-project too, see pictures below for the book I showed in class.

In the first project we’ll make parabolic speakers (probably 2) that ‘follow’ the people walking around in the installation. So, beams of audio will be directed to the person, and he/she will hear it while other people (not standing in the beam) hear much less - if anything. 

In the Touch Harmonographs project, we’ll try to make an ‘environment’ where people can create harmonographs on the walls (floor?) by drawing a line with their fingers/hand. A corresponding sound will be created.

Week 1. Try to make some harmonograph graphics (hooked to mouse dragged). Upload a screen-capture.

Week 2. Make some sounds correspond with the graphics. (Octave, Chords, see harmonograph booklet). Finish watching Theo’s VideoManager Class Tutorial & start writing classes ourselves.

Week 3. Begin making tools for a touchscreen (Ir camera, Ir lights, surface?). Decide whether to let the walls be as they are - or to put something there to deform the projections. Reserve a beamer (Rental)!

Christmas break?

Week 4. Finishing up the code - testing & trying to make it crash, etc. etc.

Week 5. Making the complete test setup. Fix last problems that occur.

Week 6. Complete project.

And a video: Harmonograph tryouts in O_F - screencapture

———————————————————————–

Tracking Parabolic Speakers:

week 1 - Looking up examples of parabolic speakers - tutorials maybe? First dish tests (speaker + dish & one speaker in tube - picture + video coming soon!)

week 2 - See how to make the arduino + 2 servo’s work with O_F. How to make 2 arduino’s work at the same time? Look for steppermotor-chips. Post more stuff on the blog - pictures/video/progress.

week 3 - Start writing code for face/blob/? -detection - how to track people the most efficient way, etc. 

Christmas break?

week 4 - Make first test-setting - complete writing the code & assembling servo’s + arduino(’s). Final decision which audio to use.

week 5 - Finishing up - testing.

week 6 - Complete project

Week 1:

Dividing the tasks
Starting on getting a convincing looking shark working
Figuring out how the Simplex Noise works.
creating areas in the camera sight

Week 2:

Continue working on the shark
Link sound fading to the areas in the camera sight

Week 3:

Finishing the shark
Checking out the camera tracking over time

Christmas holiday:

Catching up on stuff that’s not done yet, making usable test video’s etc.

Week 4:

Making the shark “bite”
Connecting the shark to the blob tracking
Water and trail in the water

Week 5:

Test setup  & troubleshoot

Week 6:

Final presentation

Cheers, Alex & Marel.