Iteration 2 - Dissasembly and Delay

I've been radio silent for a couple weeks now, but I am still here, and still working on the MT screen. My job and school are over an hour away from most of my tools and parts though, so I have very little time to work. This will likely not be a fast project. In the mean time, some progress notes and thoughts to chew on. Feel free to comment and discuss here or on NUI group. I'm not greedy as to where the talking happens, so long as the ideas are flowing.

Progress thus far:
---
I acquired two broken Dell LCDs (one with a bad power supply, the other with a cracked display surface), and dismantled them, using the parts to make one working one. If you can get a broken monitor to play with, then definitely do so. Firstly, it gives you an opportunity to see how the guts work without risking hundreds of dollars. Second, you may be able to salvage spare parts from it.

Note on LCDs:
They appear to be composed primarily of three sections..
1) Image surface (liquid crystal matrix)
2) Control card
3) Power supply.

I almost gave up on building a working monitor from the parts of the two junkers, as numbers 2 and 3 appear to be pretty much inextricably linked, however, the very last output from the control card was identical on both monitors, so either PSU/Control card combo could be used with either liquid crystal matrix. I don't know if the same is true across brands, but I wouldn't think it too farfetched to say that two monitors of identical max resolution could share that last connector.

Second, I have made some decisions on a table design. I am going to have a flat top, as it is more multi-user friendly, even if the screen is only 17". Doing some simple math, the table will need to be roughly 22" high at minimum to accomidate a 17" screen with just one camera. This could theoretically be decreased to just under a foot using four cameras, but I want to build a single-camera table first.
----
Next, I am primarily shooting for function over aesthetics with this first table, but I will try to look into the possibility of finding some sort of uniform surface that covers not only the touch area of the table, but the rest as well. IF such a surface could be set up, along with some nice capacitive touch switches for power and monitor controls, then you could have one slick looking piece of living room furniture.
-----
Lastly, I have been discussing with a friend the possibility of combining FTIR and DI techniques in a single table. This is definitely furthest down my "later on list". The primary method for this sort of technique would involve doubling up on cameras, and using light sources of different specific wavelengths for each set of illuminators.

Next, you outfit your cameras, each with a band-pass filter matching the corresponding illuminator wavelength. A cheaper alternative would be to use high/low pass filters, and have illuminators to either side of that dividing line, however this may suffer more problems due to external light. If you have the money, throwing cold mirrors behind the lenses may help minimize non-IR interference. The hope is that these filters will help us deal with ambient light well enough that an IR blocking top-surface is not needed.

All this said, the present table will be using ONLY FTIR for now. I feel that it is a technique better lending itself to natural usage, as all this talk of fiducal recognition seems primarily limited to parlor tricks. Fiducal systems require that you custom mark each and every object placed on the table, this is not "plug and play" style functionality.

For now, If someone places a bluetooth enabled object on the table, I see no reason why not to just let a user claim it as their own via a popup interface. Sure, it's not as cool, but until one of us comes up with a better solution than painting dominoes and curliques on the bottom of our expensive gadgets, it seems a far more sellable idea to both the home and professional crowds.