Success has been achieved with our multi-touch build. Success being the calibration of the screen and getting the flash examples to work. There is still a long way to go before the screen is everything it could be. As you can see from the pictures the setup is fairly… um… rustic. Currently we are using the table without a projection surface. This can be painful on the eyes when you look directly into the table. The mirror, camera and projector are not mounted resulting in frequent calibrations when anything gets nudged accidentally.
Lots of other testing has gone on since the last time I posted. We found a clear polyurethane rubber and poured a sheet on the 4″x4″ sample. It seems to enhance our blobs quite a bit. The rubber allows us to use paper as a surface but the paper sticks and the blobs stay. Interestingly the 3M Vikuiti mounted to a sheet of clear lighting gel does not generate any blobs on the test screen. This was a major blow as the Vikuiti rear projection film has impressive detail and color. Instead of Vikuiti we are going to use Rosco grey, which should be here next week. Overall we are still making adequate progress and are on track to having it all together by the middle of July.
Here is our list of improvements:
Mount the projector vertically
Make improvements to the box fit and finish
Mount the power supply
Solder up a final circuit board with low pass filter
Run the wiring through the table top so it is out of the way
Get an optical mirror to help with ghosting issues
Pour the polyurethane rubber sheet on the acrylic
Mount the Rosco grey on the surface
Mount the camera in a fixed position (possibly order a new firewire camera)
Applications, such as Graffiti, have been appearing on Facebook. Here are some of my recent drawings that were created using Graffiti. It was a challenge to create the drawings given the limitations of its brush tool and the absence of a layers palette.
Sebastien showed me this the other day and I shrugged it off. Now that I have watched the presentation it really is worth sharing and discussing. Microsoft research seems to be doing some crazy things these days.
Things are progressing at an acceptable pace. We have the acrylic, LEDs, projector, and a sample of rear projection film. I am still trying to find a clear silicone rubber from a local supplier and someone that can sell us a full sheet of rear projection film. We currently have a sample of 3M Vikuiti rear projection film. During some testing I determined that the image clarity and sharpness are exceptional. However, it is a specialized film (Vikuiti film that is) which 3M only sells in rolls starting at $3K+. Now we are trying to figure how to get a sheet. If we can’t get it soon we will just use tracing paper until we can.
Most DIY multi-touch screens mount the LEDs on the surface of the acrylic. Personally I thought this was a lot of work. In my research I found a wiki that mentioned drilling holes into the acrylic to mount the LEDs. This seemed simple and effective. So using one of our 4″ pieces I tested the drilled holes with 2 LEDs. Results showed that drilling is an effective way to attach the LEDs. On the large piece of acrylic I have drilled 26 holes on one side and soldered up all the LEDs for it. Testing on this sheet has gone well but without a table it is hard to tell how well. One thing I saw mentioned on a forum was using a metallic tape on the edges to reflect the light from the edges. I am interested to see if this will help so that is the next major test. I may also polish the holes for more efficiency.
As shown by the guys at http://www.multitouch.nl the addition of a layer of silicone rubber can really increase the sensitivity of the display. Right now a local supplier is trying to figure out if their clear silicone will stick to acrylic. They also have a flexible polyurethane that may work. Hopefully we can try both.
Things left to do: Build a table, test some silver tape on the edges, get the silicone rubber and convince 3M that giving a sheet of Vikuiti would increase their karma and sales
