imp table
prototype 1 updates

______________________________________________________________________________

Prototype 2: Narrativity, Structure and Sensing
March 31, 2006

This time around I'm getting more in depth with the table imps. A story will play itself out, the outcome dependent on the users' actions. The narrative development updates are coming along. I've also talked to Kate about the project and wanting to allow users to move the objects on the table without disrupting the animation, and she recommended using video sensing. Christian is helping with this, and recommended infrared for the camera to detect, so the imp shadows won't confuse the video - we just need to detect people-shadows and the objects. There's also progress on the structure of the table, which in the prototype version was looking a bit bulky and obtrusive. A good clean kiosk-like design needs to be used, so nothing distracts the user from the table top.

______________________________________________________________________________

Prototype 1 Presentation
December 19, 2005

I took video of the table in action. One issue with this sort of presentation is that people aren't sure what they're seeing - which speaks to the fact that this project really confuses 2D and 3D perception. In any case, the video is at least a taste of what the table project is about. I'm not sure if I have the videos loaded correctly online - I need to get more technical help on that.

______________________________________________________________________________

Almost to full prototype
December 14, 2005

I worked on my motors and PIC code. For some reason, SEROUT was causing my chip to stop working all together. So finally, I went with just having the motors react when the sensors are touched, instead of when data comes in. The pulseout widths are better now, giving the movement I want. I cleaned up my processing code and got my qprox sensors recalibrated. I also got the motors fully mounted, with extra extensions so they can push up the plate. And the table a bit more readied. And a test with the projection. I found out that my calculations were off with different projectors - I had no idea they varied so much with the projection size and brightness. Anyway, all these things took from 10am till 9pm. Another thing I've learned through this process is that everything takes 3 times longer, at least, than you think.

So, on to processing. In ActionScript, needed to combine movies with loadMovie (not attachMovie, which didn't export). Exported to Quicktime, but found out that Processing needs a Quicktime Video export from flash. This finally worked, but Processing won't allow a display at full screen size, which I need to get the animation to lay out correctly on the table. Argh. 1am.

When I talked to Dan Shiffman today, he said that I could try loading a series of images too. So I did that - I loaded 50 images, which makes for a short animation. But they needed to be renamed because of Processing wanting a straight image sequence numbering system, and Flash exports as 4-digit.

Anyway, I have my Processing code at the ready, and my PIC code is set too. Now to get back to the presentation part of all this... In order to show all the animations, I'll need to do it with SWF, but now they're all in separate movies. So I'll save and attach them with ActionScript.

______________________________________________________________________________

Serial to Flash
December 13, 2005

10am: I think I have it all figured out. With the code I researched and wrote about in the last entry, it seems like the Serial to Flash thing will actually be easy.

12pm: Oh wait, it's not so easy. I can't get the mac to recognize any data. Todd recommends that I talk to Dano, because he's the one who knows about all these Java classes and such, but Dano is in meetings.

2pm: I installed the following:
-rxtx serial java library extensions - http://www.rxtx.org/ (version rxtx-2.1-7pre17) - java extensions that may help get serial info in
-zterm http://homepage.mac.com/dalverson/zterm/ - which is like hyperterminal - simplest way to see if you're getting serial data in and best to test with before getting data to flash?
-USB adapter driver from Keyspan http://www.keyspan.com/support/macosx/usa/.
I'm a bit confused about what parameters to use to send the data. I'm getting different info from various places about what baud rate to use. Also, I'm not sure about the whole inverted situation.

3pm: Okay, Leif just said he had lots of problems with Zterm, and finally got data into Processing directly. So I'm trying that first, so see if the mac can see any data coming in.

5pm: So, finally I went back to Dano's serial server solution. I think the main reason I avoided it before was that I didn't feel comfortable with the code and running terminal applications, etc., but in the past couple weeks I've been knee-deep in all that, so it's not so intimidating. And the Cobox Micro waits sadly on the shelf, seeking revenge someday. But anyway, since nothing else was working, I installed the classes and the serial server and got it working in processing! It's great to see those qprox values show up on the mac.

7pm: Can't seem to get the data into flash. On Dano's Serial Server, I get a message saying I'm connected to the right port, but no data coming in. I've tried 7 variations on my PIC code, thinking it might be the data I'm sending. So I tried sending Hex and then binary, tried with various delimeters or none, tried different baud rates, etc. Looking for Dano, but he's super busy today.

8pm: Talk to Todd about this, and he says that everyone he knows who has tried Dano's solution has had to take a week to figure out how to configure it. And he says that Cobox will take me at least a day to get working. He recommends that, since it's working in Processing, why not load the movies into that. Aha! I may have a problem in Processing with my file sizes, but I can at least have the thing work when you touch it.

Wow, a long day of futility with the serial-flash issue. But hey, I learned a bunch of stuff:
- how to set up and configure ports on a mac
- how to work with zterm, rxtx, serial server
- lots more about PIC programming
- that flash is a stubborn old man who doesn't want to listen to serial code
- that the firmware lab makes you insane, but focused!
- processing saves the day

Oh, I also did some user testing and found that people 1) don't know it's an interactive table, 2) when they do, they pick up the objects, don't just touch them), 3) if they do just touch them, it's on places on the objects I wouldn't have guessed.

______________________________________________________________________________

Serial to Flash Communication
December 13, 2005

Okay, so the 18F452 PIC microcontroller settings need to be 19200 Bauds, N, 8, 1. Some good info on what the code should look like is on DanO's site. The code will send out the qprox values as text (DEC qproxVar1, etc). But, as Dano mentions, Flash won't like the 0 value, so I'll need to have qproxVar1 = portb.7 + 65 or something to make it readable. And then 0 at the end to signify that all the bytes have arrived. So, something like:

qproxVar1 = portb.7 + 65
qproxVar2 = portb.6 + 55
qproxVar3 = portb.7 + 45
serout2 portc.6, 16468, [DEC qproxVar1,44,DEC qproxVar2, 44, DEC qproxVar3, 0]

The microcontroller will be connected to the serial connector then the USB Keyspan serial adaptor to my Mac. Tom has some notes on this on his site.

Then, the server script I got from Hernando Barragan should get that incoming data from the USB port. Then, ActionScript code should detect the incoming data and put it in a string. (See Tom's notes about this serial data).

I need to take that data and use it to trigger .swf files from within my main movie. So, incoming data that looks like 165 (showing that qprox 1 has been touched) should trigger movie1.swf. The data should come in looking something like "66, 055, 45 ". So in my Action Script, I'll need split the data. Something like (purely pseudocode):

function handleIncoming (messageObj) {
trace(messageObj.toString());
data = messageObj.split(",")
if(data.length >= 3){
if(number(data[0]) == 66) {loadMovie("table1.swf")}
}

______________________________________________________________________________

Sensor-ality
December 12, 2005

I calibrated the Qprox sensors. 0.1uF capacitors and the exact right size of copper mesh seems to hit the sweet spot, but it does seem to adjust over time. So, it's working, but when I have everything else in place I want to go back and refine it. The cup is harder than the plate to get a good touch ratio happening - maybe because it's ceramic. My code is getting closer...

It's also time to realize the whole microchip-to-cobox-to laptop-to flash extravanganza. While researching this tonight, I came across this Hernado Barragan's serial to flash solution. Oh wow - it's possible I don't need to do the whole Cobox setup? I'm working on this now - figuring out what code I need to send what data out from the chip, what data flash can receive, and then how to code ActionScript to play the right movies depending on what sensors are touched. Barragan also has an interactive table project that is good research material for this. Oh wait, he also wrote Wiring, which Arduino is based on... suddenly I'm seeing the bigger connection - and how my little project fits somehow into this wider movement to combine moving graphics and physical computing. It's like being at the Sex Pistols' first show! Well no, maybe Bowie's 20th show. Anyway, if I learn more about this whole networking and data transfer piece, maybe this piece of the puzzle will be much easier a year from now.

______________________________________________________________________________

Table Mundo!
December 10, 2005

I built the table with rack and matched the plate and cup placement to the animation. I drilled some holes for the parts the motors will use and experimented with various ways to mount the motors under the plywood. The optimal way would be something that would allow me to easily detach or adjust the motor. I got the right amount of pulse width for the motors too, so they can push a little part up through the table to tip the cup a bit.

I'll have to see if I'm able to use the magnets - with the metal of the rack, it might be problematic.

______________________________________________________________________________

Imp-a-Limba Land
December 6, 2005

So, so much animation. The processing of images takes a while, and then moving them around the stage so they look natural. Gotta say, though, I'm loving the dance sequence. I really should be on a faster computer to do this work, though. Around 900 frames each movie. Through a lot of experimentation, I've gotten the video-to-flash system working well... it's modified from what I thought would work (listed down there on Nov. 27).

The new system - for each clip:

1. In Final Cut, cut short clips from video and apply brightness, contrast and levels filters to get pure black on white image.
2. Export using Quick Time conversion - Photo JPEG, 8fps.
3. Import short movie to flash, embed video in swf, resize to 550
4. Export PNG sequence
5. Run through custom Terminal application which saves every 3rd image into a new folder
6. Photoshop batch process to resize to 275, adjust canvas size, name in 4-digit serial
7. Import PNG sequence to flash
8. Move all images to correct size and location on screen, using onion skinning to select all frames
9. Adjust, tweak, add animated drawing, etc...

Also, on the PComp side, I got the QProx's all soldered with coaxial cable and copper mesh, and started building the table.

______________________________________________________________________________

Progress
December 1, 2005

Got wood for the tabletop, copper mesh and biscuits to potentially use on the movement. I wired my board up with the basic parts and did some coding (starting code here). I got the servos turning and pot responding.... next steps on the PComp front:
- finish hooking qprox up to copper mesh, experiment with different capacitors under plate, cup and spoon
- build rack and wood structure, finalize object layout
- servo mechanism for object movements
- attach servos, objects, plates

_____________________________________________________________________________

Slow Goin'
November 28, 2005

Found a hardware store that cuts lumber to size - Home Front on 29th and 3rd - open 24 hours! Also talked to Michael about servos and cable and the wire rack - which he said shouldn't cause me a problem. Briefly talked to Tom about the Serial Server solution, which he said is very easy, but that's coming from an expert who can do all of this in his sleep! Fine for someone familiar with Java and comfortable with coding in general. Makes me realize that more programming or networking classes would be good for me. So against his and Michael's advice, I ordered the CoBox Micro. Maybe I'll have time to figure out the Serial Server solution, but at this point I just need to get my breadboard set up and functioning, and some actual animation together. That stuff is taking longer than I'd like. I had to make modifications to my process of converting video to flash - the steps listed below - because Adobe Bridge isn't quite to the point where it can do the kind of batch processing of vectorizing files that I need. So I'm using bitmaps in Flash - one for every frame. Which is just huge for a 5 minute animation and keeps crashing my machine. But it's what I've got right now until I find a better solution. I'll vectorize the frames I need to adjust and keep the others as bitmaps for now. In the meantime, I've got to refine my system of processing and placement before I can really get to animating. It's times like these that I wish I were doing claymation with a couple LEDs thrown in! A little instant gratification would be nice right now.

_____________________________________________________________________________

Converting Video to Flash
November 27, 2005

After many days of trial and error and software research and asking of questions, I have a system for converting the video of me being the imp to vectorized frames in Flash.

1. In Final Cut, cut video into short clips corresponding to specific actions (running, dragging, laughing, etc).
2. In Final Cut, apply brightness and contrast to get almost pure black on white, some cropping as necessary
3. Render and export as Quicktime .mov files at 8 fps, greyscale, photo jpeg
4. Import a video into flash 8 using the new Import Video wizard, embedding video and using default settings except change size to 400 pixels wide in Advanced Settings
5. Export as png sequence
6. Delete every 2nd and 3rd frame by running Terminal Script on the folder.
7. Vectoring every image - Illustrator CS2 does this - use new Adobe Bridge to do a batch live Trace. This process isn't perfect. I'm sure Illustrator will improve it after a while. For now you can only batch 10 at a time and it takes a while. (I looked into Wildform and Siame software for this before coming upon the Illustrator option).
8. Change .ai files to .png with batch. Since flash will only import all if #s are sequential, use the batch processing to get sequential #s.
9. Import png sequence into flash
10. Resize and adjust placement of objects as needed, add blur and alpha effects
11. Repeat for each video clip
12. And finally, work on the animation!

For step 6, I was actually deleting thousands of files manually until Nathan wrote me a script to run in Terminal which does it automatically. Thank you Nathan!

Here's the code:

#!/bin/bash

n=1
for i in ./test/*;do
if [ $n = 1 ]; then
filename=${i#./test/}
echo copying $filename
cp -p $i test_destination/$filename
fi
n=$[ $n+1 ]
#echo "n="$n
if [ $n = 4 ]; then
n=1
fi
done

______________________________________________________________________________

Gears and Motors
November 26, 2005

I talked to Todd about gears because I haven't been finding what I need - at Radio Shack there only seem to be circular gears whereas I need linear movement and I haven't found interesting gear parts elsewhere yet. Todd said that all I really need to use for much of the movement I want - since I only need things to move less than an inch - is a servo motor with some of those little plastic parts that come with it. Also, for that cam rotation for the cup, he said I could get a wood oval biscuit. Michael also sent me this good list of parts suppliers. The phidget site has lots of servo motors and parts. I'm wondering if I can get the servos cheaper elsewhere. Their list of hinges and parts is a good reminder that I have to think about the hardware to connect all this together, but I think that cheaper hinges from a hardware store will work just fine.

______________________________________________________________________________

The Physical Setup
November 25, 2005

I realized that, in order to do the animation and really understand how to deal with the physical aspect of all of this, I need to think about the thing as a whole for a bit. I need a table with holes through the top in various places, plus, the projector overhead, specifically located. I found a metal kitchen rack counter top which will be perfect for the basic table. I'll get a fairly thin piece of plywood at about 20"x16" for the top. With the metal rack setup, I can extend the 4 columns upwards to another rack which I can modify to hold the projector, which will mean that every time I show it, I don't need to carefully suspend the projector from the ceiling. I don't think it's great aesthetically, but it's good for this first pass. I also need to be careful about the metal, I think, since I'm using these very sensitive sensors and magnets. But parts of the rack can be cut out and it will still be stable enough to hold everything. I created a new layout diagram for the pcomp page.

I got a projector and my 4 objects - plate, cup, large spoon, and tested out how high above the projector would need to be (35"), and worked on translating the size of the objects into the pixel dimensions I'll work with in flash. I'll do a 550x400 animation, so the 10.18" plate will be 279.9 pixels. I have to be really exact about this in order for the animation to look good with the shadows.

______________________________________________________________________________

Writing About It
November 23, 2005

I've written a short proposal based on Michael's documentation guide: Goal, Context, Problems, Proposal, Target Users, Design Criterial, User Scenerio, Hardware Design.

______________________________________________________________________________

Hardware
November 22, 2005

I've seen the light and realized that Tom Igoe and Dan O'Sullivan's Serial Server solution is just not going to work for me on this project - too finicky and complicated, using java classes and all sorts of stuff I have no experience with. I'm already full up with trying to accomplish things I've never tried on this project. So, in the interest of getting this done, I'm gonna order the Cobox Micro - I think CO-E2-22AA with two TTL channels. Vincent was nice enough to send me to a Parson's student music project using the Micro, which further links to this ITP Networked Objects project. Great stuff. And, Tom has a very helpful page on his site about setup and code for the Micro, which also lists the parts I'll need for setup. And another resource for hooking up. Anyway, things being what they are, it's not as easy as going to a website and ordering, it seems. I have to wait to hear back from a Lantronix sales rep.

For the qprox sensor, I'm got the QT150 since the QT160 is out of stock till the end of December. I worked on my breadboard some today, and got some more parts from Jameco and Radio Shack. More drawings, too. I'm a bit unclear about how to deal with the motors and gears I need. I keep doing lots of research, expecting to find resources to buy cheap gears - like the rack and pinion and cam rotation stuff I need, but I'm just not finding it.

______________________________________________________________________________

Specifics and Parts
November 21, 2005

I met with Michael about the project and got lots of good advice. He said that since I want 6 sensors, I should use the 6 key QT160 instead of 6 separate QT113s. I'll need a 10mHz oscillator for this. And for the toggle, I'll want to set it to 60 seconds so it can recallibrate. Also, he said that the wires between the microprocessor and QT160 will be sensitive to proximity, so those should be shielded. He recommends cable TV cable, and solder wire from silver shield in there to ground and don't let it touch inner connecting wire.

Update - the 6 key QT113 is out of stock and not available until the end of December, according to the distributer's website. So, I may need to use the QT113's after all...

For the movements of the various objects, Michael showed me this great site, flying pig. Lots of fun mechanism types. Details about some options, with motors and gears and magnets, plus an animation I did for my cup are all updated on my pcomp page. The process of making the cup animation was great in helping me understand what I need to make and test. Taking time to sketch out the layout with every component I can think of helped me make a parts list too.

Serial Connection to Flash
My nemisis at the moment. I have some questions via email racing out to various resources - Dano, some alumni who've done this before. Tried Dano's Serial Server solution, but not getting the java bits to run correctly yet. It would be nice to get the free solution running - I believe the Cobox is $89 and the Teleo module is $159. As Dano mentions on his page, it seems ridiculous to spend that when a cheaper built in solution is most likely on the way.

Update: Got an email back from alumni Jack Chang, who said he has used CoBox and that it's good because it's stable and once it's set up, it's set up. He hasn't tried Dano's solution. I found this Cobox code from Tom Igoe...

Animation
I've also finalized my storyboard for this first phase of the project. I don't know that I'll have time to integrate all of the content I want to and my intention for this coming 2 or 3 weeks is to get it to a place where it's working. I want to get all the parts working and the animation looking good, if not the extended narrative with text and facts and 25 separate reactionary animations. It's something I would like to continue with, though. Anyway, for now, I want 5 separate animation sequences that react to touch: 1) Plate Peas Chicken, 2) Waking Dance, 3) Cup Swim, 4) Fork Fling, 5) Globe Spin. The extended Single Pea animation is the type of thing I envision doing in Phase 2.

I shot more video this weekend, using the Japanese Room for more horizontal space to move in. However, I didn't get as much of an angle as those first videos. I'll play with all the video I have and see if I can use the more angled stuff for animations in the foreground (plate in particular), and these recent ones for animations further back on the table - which makes sense anyway.

To Do in the coming days:

- Serial to Flash - figure out which solution - Dano's java solution or buy CoBox and figure that out
- Buy the rest of the parts and materials I need
- Video - edit the 3 hours into short sub-segments, adjust contrast, bring into sorenson squeeze, and then into flash as .swf, adjust frames, trace each - prepare to edit
- Flash 8 - experiment with effects and animation techniques
- Refine map of all physical elements including breadboard, motors and gears, object layout
- Prep Flash movie with exact layout plan for shadows and movement

______________________________________________________________________________

To Do Now
November 18, 2005

To do this weekend:
- finalize storyboard
- re-videotape all imp sequences, with best possible lighting
- video - edit clips, adjust contrast, bring into sorenson squeeze, and then into flash as .swf, adjust frames, trace each - prepare to edit
- sensors - test and calibrate qprox
- serial to flash - try DanO's way of doing it, try to get qprox data into flash

Dan from class tipped me off to this interactive bar in the Ars Electronica space - luminescent worms on projected on a table. And right now, while I'm finalizing the storyboard in order to launch more fully into production, I'm remembering that simple can be beautiful, and for this version of the project, I should not get bogged down by too much narrative. (See my pea storyboard). Instead, I should get everything working in an elegant way. First I'll get my video into flash, and see what the movements of the imps tell me.

______________________________________________________________________________

Feedback and Answers
November 17, 2005

PComp and Animation classes today. In animation, Marianne said that Sorenson Squeeze should help solve some of my Video to Flash issues - I should be able to save as a .swf out of there and bring into flash as a sequence that will populate the frames. Also, she recommended that I get the Japanese room to refilm, so I have enough space to move. It's also a good opportunity to really get my script and movements figured out so I can tape exactly what I need.

______________________________________________________________________________

Flash - Video to Vector Questions
November 17, 2005

I got helpful feedback from the flash email list about getting video into flash frame by frame and vectorizing it.

eric:
bring video into processing. saveFrame() every 3 frames with something like:
int frameCtr = 0;
void draw () {
if (++frameCtr % 3 == 0) {
saveFrame(); } }
note syntax here is not correct, this is just pseudocode.
then open all those TGAs in photoshop and make an Action that automates the white selection (magic wand, 0 tolerance), clearing, and saving as PNG. then bring those PNGs into a video editor that supports transparency (FCP? QTPro? dunno about this part) and export as a video. import the video into flash and let it make keyframes for you. no way to automate the bitmap tracing, i don't think. looks like a *bit* more automation than you have currently...?

liana brazil:
i'm also not totally sure what you want to do, but here's some info that might help;
-in QT Pro, you can export the movie as a image sequence with whatever frame rate you want. so if you do 10 fps, it'll automatically pick 1 frame every 3 frames. then this image sequence will be named like: sequence001, sequence002, etc.
-photoshop has actions that you can record and apply to a batch
-flash will import all frames sequentially if you import the first frame of sequence001.jpg, sequence002.jpg, etc. it'll ask if you want to import a sequence and it'll place the frames one after another (so no need to drag each image on individually)
there used to be a program Adobe Streamline, it has been discontinued and they say Illustrator substitutes it. not sure how it works with a batch, take a look:
http://www.adobe.com/products/illustrator/pdfs/ creating_vector_content.pdf

dan:
I'm not sure what you're trying to do for the final result, but flash 8 has
alpha channel in the video now for blue screen effects (those are called
Luma Key right?). And 8 also has a new bitmap api that you can use to mess
with pixels, although I'm not sure if you can grab frames of the video
itself. Seems like you can use the copypixels function to do that.

______________________________________________________________________________

Video - Luma Keying
November 16, 2005

I worked on the video in Flash. I found some info online about chroma keying, but that's mainly for blue or green screens. But I found that I could apply the same principles for the white screen using Luma Key and the Matte Shape Mask. By reducing the tolerance to 0 on the Luma Key, I was able to get a pretty good shape.

The Luma Keyed clips take a long time to render. I wasn't sure about the best way to export from Final Cut to bring into flash, especially since I don't want to use the video as is but instead just use the shapes and vectorize them. I decided to try a basic QT mov file. It imports just fine, but I find that I can really reduce the frame rate on the original file - probably 5 frames per second will work fine. Also, I need to figure out how to change the video to vectorize it - it doesn't treat video like bitmaps, but maybe I can convert it so it does.

Later.... my desperate email to the flash email list:

Hi everyone,
I'm looking to bring video into flash, but what I want to do is take every couple of frames of the video and trace an object so I can manipulate it further.

I've done a mask and Luma Key in Final Cut so the object I want to animate is black on a white background, brought the video into flash, exported it from flash as a PNG sequence, deleted every second and third image, brought the rest into Fireworks to get a transparent background where the white was, saved as PNGs, imported into the Flash library, and proceeded to add a keyframe to every frame and drag each image on individually, and then trace each bitmap individually.

This is crazy, right? Please tell me there's an easier way!

Thanks,
Rebecca

Next steps on the animation: cut up the video into clips for each animation sequence according to storyboard, luma key better, and get into flash. Then I'll move on to mapping all this to the table layout.

______________________________________________________________________________

Narrative and Physical Elements
November 16, 2005

I also thought more about the sort of narrative to pursue and talked to a friend about it. I want this to be educational, about food issues, but not too didactic or fact-driven. I want to speak to the fact that the everyday choices we make have a broader impact. When you pick up a strawberry, it has this reverberation, a long shadow cast that spreads further than you might imagine. The roads it traveled, the person who picked it, what it was sprayed with, where it grew.There's a whole story behind that strawberry, and I'd like to tell those stories in interesting, entertaining ways that also get people thinking about some of the problems with our modern food system. I want to talk to some more people about this idea to get a sense of what they might find interesting in that story.

I also did research on the sensors and moving parts for the table and updated the pcomp page on this site. I'm getting a better full picture of the thing as a whole and how I want it to work.

I think I should start with a plate, bowl, cup and globe for the people and imps to interact with. There will be sensors on each object. I'll need to test to see if touching the top of the cup will set off a sensor underneath the cup. Also, whether 2 sensors under either side of a plate will get data independently or if touching anywhere on the plate will give the same output.

______________________________________________________________________________

Serial Data to Flash, Sensors, Video Taping
November 14, 2005

I've been wondering about getting data from the sensors into Action Script. DanO'Sullivan's web site is super super helpful - all the steps I need, hopefully. Also, someone on the list said, "I've used the Telo board from making things: http://www.makingthings.com/ They have an XML server you can use to pass info into Flash. It's probably the most widely accepted solution for sensor input into Flash. I haven't tried Dano's solution yet. They're similar in concept."The thing is that these Telo modules are $189. I have heard that the more manual solution is tricky and even programmers with experience have a hard time with it. But the best bet is to find the people with experience and see what they have to say. I was given contact info for some alumni who have done serial to flash work and would be willing to help, and also told that people who have taken Networked Expressions should be able to help with the code for the server. It's good to know there are resources out there!

Also, I took a look at my QT113HD sensors, and was surprised. I thought they would be little pads that get input about the pressure. But... they look like chips. So, obviously I have some reading to do.

I videotaped today with good lighting and white background and next will prep that in Final Cut and get ready to bring into Flash. I found this article about chroma-keying out the background.

______________________________________________________________________________

Diving In
November 13, 2005

The QProx sensors I ordered last week have come in. I'll pick them up tomorrow and try them out. Michael mentioned that they take time to calibrate, since they're very sensitive.

I've also done some work on my process and schedule for the next 5 weeks:

Week 1: November 7-13

• Research other similar projects
• Buy sensors
• Schedule and to do list

• Test and calibrate sensors
• Research motors and other movement of objects
• Find out how to get chip to send data to action script
• Video tape all imp movements, import to flash
• Full storyboard complete

• Build prototype with sensors and motors and do user testing
• Research for content, finalize all
• Finalize plan
• Get touch sensor to action script part working
• Get laptop out to motor working

• Finalize animation
• Refine final table and object pieces
• Get everything working well

• User testing
• Final presentations, documentation

______________________________________________________________________________

First Feedback
November 10, 2005

Today I presented my preliminary ideas and research to my animation and physical computing classes and got good feedback on a number of things.

First, on the animations. People liked the shadow exercise I did with transparencies, saying that it was great that the little imps have exactly the same quality as the shadows of the plates. They also liked the character development video, and recommended that I use that as a basis for my animations. In other words, retape the video and import into flash to refine it and add more. This means cutting down on animation time and I think it will go a long way in creating really natural movement.

For the physical environment, people liked the idea of being able to interact more fully than just touching an object. I had mentioned spinning a globe, and they thought spinning a plate too. They liked that in the Kobito project, you could push an object against the animations and they push back. I could consider doing that. It would involve magnets. Or, if I had the plate or something move just a couple inches, it could be on a track sliding under the plate. I would want to come up with something in the narrative where you're invited by the sprites, or challenged, to push the plate. So people have an idea of what objects are able to move in what ways. Otherwise people will try to push or rotate everything, and may get frustrated with few options.

______________________________________________________________________________

Beginning
November 8, 2005

I created the project outline and overview, with some preliminary research.