Today i graduated at Communication & Multimedia Design in Breda, The Netherlands.
After 4.5 year of full time dedication im released to enter the market.
In my final presentation i showed the projects i worked on during the last year. In between the different projects i generated backgrounds with processing based upon the graphics/photos of the project. This script created a number of columns based upon every tenth pixel of a random horizontal row. Delevering some nice results. And the most important part is that people who knew the works recognized these in the backgrounds.
During the development of my multitouch application i stumbled upon the problem that the TUIObject didn’t perform quite like how i was expecting. So i jumped into the AS3.0 TUIO bandwagon and made some adjustments.
The issue was that the way how rollOver and rollOut where triggered where not quite like how it is triggered in mouseEvents. So i re-arranged a part of the TUIObject.as with the result that i am able now to move my finger over the stage onto a picture which will trigger an rollOver event, as soon as i continue sliding and slide from the current picture onto an overlaying picture i get another rollOver event, and a rollOut event for the other picture. As soon as i release my finger i get a upEvent but not an rollOut event because you wont rollOut when you release a finger. With this modification is was able to get my flipping of the pictures to work.
At this moment i am not aware of any problems, but this modification needs heavily testing though. But maybe this helps someone else in progressing TUIObject as3.0
Yesterday i had my presentation as a part of my graduation.
All went well and they want the screen for a public viewing of the school, so right now the screen is located at my school and i am unable to progress it further untill the 20th of november.
Also this day delivered the first quality tests with multiple people jamming on the system at the same time.
The following screens are captures of the multitouch workspace, i made an gesture for the letter S which enables one to make a capture/save of the current workspace. It delivers some nice results.
During the building of my multitouch screen i’ve been working on my multitouch image research software also.
With this application its possible to collect images from a Flickr user, i.e your own account. And flip the images, on the back of the images the title can be found and also some relative tags. With these tags you are able to conduct further search actions. In this manner you come across images you couldn’t imagine you would come across (nice sentence, like the television show which sell magic carpet cleaners. anyway..)
Either by making a W gesture you are able to call up a rescalable keyboard and put a search on flickr. And again you can flip the resulting images. Also with the keyboard there is a possibility to make labels in order to organize your research.
Im working on a function in which you can export all the pictures with tags to an xml file for further use. Right now yo can save the canvas to a online directory.
Not all functionality is seen in this movieclip, but as soon as i am able i will make a better movie showing the features.
My multitouch screen has reached a state in which it is completely stable.
I can walk away and come back an hour later and still have a working multitouch.
The wiggle in the blobs is completely eliminated. The only bad side effect is that i lost a small portion of sensibility but thats way better than wiggling blobs. Thanks to Seb from the Nuigroup forum. Additionally i added some code so i could adjust the wiggle sensibility from within a xml config file. No more C++ compiling for now.
Last night was my first experience with C++. Actually it looks alot like As3.0 or is it the other way around ;).
The case with the shaky fingers bothered me so much that i opened up the source code and with good help from the Nuigroup i was able to eliminate most of the shaking. The side effect is that a finger isn’t tracked when it moves extremely slow. But with help from the Nuigroup community i might be able to fix that also.
Still C++ feels like your dealing with “The real thing!”. Very Mature, or so. Zipp it.
Some progression, mounted the mirror in the box and fixed the infrared led clusters in the box. With a bit of tweaking i managed to get the screen evenly lightened by the infrared light. I did some first photo app tests on the new setup and i am pleased!
The blob detection is still a bit shaky but i hope to fix that later. Next step is to get rid of the hot spots from the led clusters, i’m gonna put a reflective soft filter on the inside of the box and point the led clusters towards it.
