Day 10
The Arrival of the Glass
Two weeks after ordering the glass I get the call to say that it has arrived, after arranging for delivery and paying for the glass it arrived and was put into place by the delivery men who cleaned it and seemed quite interested in seeing the working system. Here are a few shots of the non-laminated glass.
Lamination Applied
Six weeks after the glass top arrived, six weeks after the order was placed, six weeks after the specifications were given to the glass tinting company they finally came around to apply the laminate to the table top after many follow up enquiries to find out what was taking so long. It took longer to get a tint applied to the table top than it did to develop the whole thing from concept to working prototype and final frame complete with glass top. Then there was the... professional service... the laminate was cut to the wrong size despite make follow up measurements for themselves and the laminate used was not the one requested after I had spent a significant amount of time testing different laminates to select the right one. There was now a laminate applied however and I was eager none the less to get the table top in place for a little testing.
The laminate applied to the table was far more transparent than the one I had requested and as a result the web camera can see through the film too easily and as so it can see the glare of the IR spotlights off of the other side of the glass top. This reduced transparency also means that the majority of the projected light makes it through the laminate onto the roof. The glass tinting company had said that it was going to take measurements for a second layer of laminate, this time of the correct type. In my tests using the prototype screen I had found that two layers of the milky frosting laminate (which is what I had asked for) was as good as one milky frosting with one of the same type as already applied to the table top.
The Last Problem
The new laminate material caught the projected image well and it also let IR light pass through well enough to see reflections off of your hands when interacting with it however since it is plastic based it has a slight shine to it. The shine is not immediately noticeable to the naked eye however it is extremely obvious to the webcam. The material used for the prototype screen, the double matte drafting film, has no noticeable reflections at all and lights up evenly under the spotlights for the webcam. The hot spots produced by the bright IR spotlight reflections interfere with input as the standing reflection is brighter than the reflection of light from your hands on the top side of the glass.
As a solution to this problem I considered several approaches. I could affix a piece of the drafting film to the underside of the table top which would stop the reflections however the air gab would make reading input hard and bring the total number of diffusing layers to three making it quite hard to for the web camera to see through and would also highlight a region of the table top where the screen would be projected making the surface non-uniform. I could attempt to sand the laminate with a coarse sand paper in an attempt to remove the lustre from the surface or I could try applying a number of different chemicals to try and damage the surface of the laminate to give it a chalky texture.
I had decided to try a number of different chemicals on sample pieces of laminate as I thought it would be easier to apply uniformly across the entire laminate and would be more consistent than sanding. However before I could proceed I was inspired with a better way of applying the drafting film. If I were to apply a liquid agent to one side of the drafting film, as you might a screen protector for electronic gear, I would eliminate the air gap removing the inconsistency of the air gap and bringing the diffusing layers very close together. The liquid would also server to make the drafting film more transparent meaning that it would reduce the input quality far less than if it were to be applied dry.
I added a quick test piece to check and all seemed well. Later with a helping hand I managed to use water with a little detergent to adhere a new clean piece of double matte drafting film to the underside of the table top, removing reflections and only slightly reducing the input image quality. In the below photos you can see the diffused reflections on the lamintate compared to the glass (left) as well as the difference between the diffused reflections and the matte (centre) and lastly a profile shot showing that the matte even seems to work with the greater room natural light.
The only downside left to this approach was the non-uniform appearance of the laminate that alludes to the existence of something special about that region of the table top while the projector is off. The water holding the pieces on with surface tension eventually evaporated and the segment additional film came loose after a few days however it lasted long enough to show that it would solve the reflection issue and give a good clean input. The hunt began for an appropriate adhesive that could either be coated on the film perhaps with a spray-pack or roller or that could be mixed into the water when mounting the film that would allow it to stick to the surface longer.
After a number of test sections on the old test screen from the first prototype with various glues I managed to settle on a bottole of water soluble Clear Gum from a local art supplier. I applied it directly to the table top in excess, positioned the film and then used a squeegee to work the excess glue out from underneath the film. It took several days to dry and the effect was nearly as good as the water adhered film though the effect lasts much longer.