Hello,

I am trying to build a LCD DSI setup.

I have bought Endlighten plexi and led strips (850nm )which I mounted on the edges. I put a camera under, and a glass over, but i cant see the DSI effect. If i press the finger directly on the plexi i can see the FTIR effect. The edges of the plexi are milky so i think there are endlighten. I testes this with light leds and the light is filling the whole plate. What am I doing wrong? I cant i see the fingers on the touch surface?

Please help!!!!

Same thing here. What happens if you have a small gap between the glass and the endlighten? I had the same problem and just stopped using a glass. Now I directly touch the endlighten. Obviously that’s not how it’s supposed to work, but I actually don’t think DSI is that well researched. It’s just some guy who put it on his blog as a possible solution. To my knowledge, nobody has really scientifically measured any results with it.

That leaves us with two conclusions:
a) DSI is bogus (The only thing indicating this is not true are some people who claim it works for them. By magic I suppose.)
b) You don’t have enough light. (Since you’re even using the LED strips I doubt this is the issue.)

or c) somthings wrong with your camera

I use a PS3 EYE. I removed the filter and I put band pass filter from a floppy.

Can I do anything with it?

if ur leds are ok - then i would say yes.

According to some spectrum analysis i’ve seen in a thread in this forum, floppy disk is the worst bandpass filter existing.
You should try negative photo film or even better: real filter from omegabob ebay or Peau shop.
What do you mean by “The edges of the plexi are milky”, maybe you need to polish these edges ?
DSI-LCD works perfectly for me, it’s not bogus, or magic you certainly have done something wrong.
Can you post pictures of your setup & bad results if you want more help.

I’m already using a proper bandpass filter and still have the same effect as the original thread starter.

Speedman: You got some other material such as glass lying directly on the Endlighten, yes?

I’ve also tried, like many of you, in order to have a better touch sensation to put glass on top.
But No ! .... Never put glass on top of your Endlighten even a very thin (i’ve tried a 2mm) no IR light passes trought it.
Maybe special glass could be a solution: http://nuigroup.com/forums/viewthread/1119/
but what i really think is the DSI effect needs the finger to be in contact to work correctly.
A simple test: You do a very small rise of your finger from the surface, you will notice that the blob intensity fades very very quickly.

So you have the same effect that we described, which is just the plain old FTIR effect and nothing special to Endlighten. You can use any acrylic for that.

I hence conclude that a real evidence as to the non-bogusness of DSI remains yet to be provided.

@dennda
i sat at two DSI tables… what do you mean with bogus?

Do they work if you don’t rely on the FTIR effect? (I.e., do they still work if you don’t touch the Endlighten directly but touch another layer on top of the Endlighten such as protective glass?)

In case of FTIR you need an “additive” surface above the acrylic, often named “compliant surface” (silicone rubber ie. Sorta Clear 40, Elastosil, Lexel, etc.)
By doing this, in my opinion, the touch is less pleasant, you need to realy push on the surface, if you only soft-touch you don’t have any blobs.
In case of DSI, no need of compliant surface, and you could have fiducials, a thing you can’t do with FTIR.

I think you need to read this: http://nuigroup.com/forums/viewthread/1982/

That is not correct.

A compliant layer is optional in FTIR setups. Its use was suggested by the original author (Jeff Han) to prevent the acrylic from attracting dirt and improve touch sensation. The TIR effect is a well-known physical phenomenon that in no way depends on something like a “compliant surface”. The compliant surface and the lack of coupling between objects other than greasy fingers is what prevents the usage of fiducials, not the acrylic or the TIR effect itself.

I have read the thread you mentioned and even built my own “DSI” setup. The problem is that several people claim that you can use an additional layer of glass (or similar) to protect your Endlighten and/or improve touch sensation. That, however, have I not been able to reproduce, since any material on top of the Endlighten triggers the FTIR effect still present in DSI and you lose most of your precious light due to that.

I suggest you read the original paper on FTIR by Jeff Han.

@dennda
yes, they were using fidicuals and tested different fidicual sizes - i saw it myself, it works not only wih FTIR - BUT we were touching the endlighten directly… im really worried if my setup is going to work...???

@speedman
i was planning to put my projection layer (7D006) on top of the endlighten and to cover both sheets with a thin adhesive scratch-foil , or 3mm clear,anti-reflective acrylic… you dont think thats gonna work?

but how do you protect the endlighten from sctratches? do you have your proj. layer under the endlighten?

Hi guys,

I also have a working DSI setup, but of course I don’t have glass on top since glass absorbs too much IR. At the moment I have that plastified tracing paper that architects use. And with that I have a good DI effect, since if I set amplify high in CCV I don’t even need to touch the surface to get blobs. So now blobs are great.

What no one mentioned in this thread is that at what resolution do you have your cam? For me it was quite better as soon as I put it at 640x480 @ 50fps. With more resolution fainter blobs are better detected.
I also plan to use 7D006 on top of my setup but I also plan to add one more cam at 640x480, to make sure that I will have a high resolution image of the surface.
So DSI is definitely not a bogus, neither magic. It has its limits, but if you stay within, it should really work.

The second thing is setting up CCV. To get it really sensitive do it like this:
- When CCV launches set Image Threshold to 1, Min Blob size to 1 and activate only the Highpass filter with Blur at 0 and noise at 0. (if CCV crashes leave Blur at 1)
- Do a background remove while you are not in front of the screen. (there will be a lot of noise but thats OK)
- Set Blur to 2 or 3 and noise to 1 or 2
- Background remove
- Now adjust the image threshold to filter out the noise, once there is no more nois
- Background remove
- Try tracking
And after this you should play with Blur, Noise and Amplify. Probably you won’t need to get Nois much more up. You can crank up Amplify at 130- 190.
Tell us how it worked. (If you haven’t been doing like this already)

Greg

I also have a working DSI table… with a projection surface above the endlighten (ACRYLITE Rear Projection Sheet: 7D513 RP). This provides a nice touch-feel to the surface.

When illuminated, the entire surface glows with IR illumination… it does work much better with my ps3 eye running at 640x480.

I have done some testing with reacTIVision as well. I was generally happy with the results.

http://www.brianlucid.com/blog/