INTRO skip if too boring!
Hi Guys (n gals), my name is Stuart, I’m from the UK and work in IT, luckily I know a thing or two about C++ programming and have done quite a bit. I’m bursting at the seams to come up with a MT display, its v exciting stuff, I myself have performed some MT experiments with sheets of perspex, a modified (crap) logitech webcam about five years old and my very old projector. I have yet to ‘marry’ all the bits together but have successfuly tested blob detection using TouchLib config tool with a small area illuminated by IR LEDs.
/INTRO

I’m new here but am looking at staying some time. I haven’t posted on the intros page yet because I’ve got about a thousand questions to ask, this technology is too exciting!
I’ll ask several here, to keep me going for now, many thanks for your feedback.

I’ve seen Gravanos’ Tutorial PDF on building a multi touch display, excellent work there by the way, it must have cleared up the previous thousand questions I had, all in one go.
However, I’m still a bit confused as to the purpose of everything in the display surface. Firstly, am I right about your construction? see below.

Air
-----------------------------------
Acrylic (Perspex in my case)
-----------------------------------
Silicon
-----------------------------------
Rear proj surface (is this the ‘compliant’ surface I keep reading about?)
-----------------------------------
Air

1 - Okay, I’m reading on the forums about how some people have multiple acrylic surfaces? Whats the reason for this?
2 - My main question though is why do you need the Silicon layer? I’ve performed some tests already and can see no reason for this layer because I’m still getting finger-blobs right through simple (but good quality) tracing paper, the paper is in direct contact with the perspex layer, I’ve actually tried on top and underneath and both are effective(ish).
3 - I can see a need for a better rear-projection surface, as even high quality tracing paper shows a graininess, so whats the concensus on the Rosco screen material, which ones would u recommend?
3.5 - I’ve just thought that you could perhaps combine the projection surface and silicon layer with just the silicon layer which has been mixed up with something that diffuses the light- turn the silicon layer into a projection surface too. Is this feasible?
4 - My testing has revealed that its difficult to achieve a good bright contrasted blob without pressing reasonably hard on the perspex, I have several thoughts as to why this might be -
(A)-IR Illumination coverage and brightness not good enough.
(B)-Perspex material is bad (but all have been very clear in appearance and I’ve polished the edges).
(C)-Perspex is too thin (so far I’ve used 2mm and 6mm, I have 8mm still to test).
(D)-My poor Webcam (Logitech QuickCam Express, removed IR Filter, put in photographic negative).
The problem could very well be a combination of some or all the above, and more, any ideas specifically to improve on finger-blob contrast?
5 - Whats the point of the baffles? Are they mounted in the corners generally or somewhere else?
6 - Can somebody put forward a decent camera recommendation thats likely to cost < 100 Euros and can be easily set up for IR use?

On a side note, I’ve made an interesting discovery for mounting IR LEDs alongside the edge of the perspex - I use what are called slide-binders, normal stationery items for holding together lots of paper sheets, I happened to have some in black. Along the spine drill 5mm holes for the LEDs, poke them thru, then slide the thing onto the edge of the perspex sheet, so far I’ve managed to slide them over 6mm perspex, but I’m sure you can on 8mm too. Works a treat for blocking out IR from shining anywhere except inside the perspex. I’m sure I can then frame it all later on by mounting the whole lot onto a wooden frame. Of course I’m only working with letter/a4 sized sheets of perspex so it’s about perfect for prototyping/experimentation.

Many thanks for anybodys feedback, I’ve still yet to do a lot of expermentation but its going to be great fun. I’ll set up a site when I start to have something concrete.

Stuart

Clever thought with the slide binder - I may have to look into it.

You answered your own question re: the compliant surface.  A good compliant surface alleviates the need for strong pressure on the acrylic (making it easier to track drag events).  If you can find a good projection surface that is both low friction (to protect your fingers) and will frustrate the IR, then yes it should work.  I’ve been testing with a yoga mat.  Too high friction on the fingers and it distorts the image too much.  It does work though.

BTW, use silicone, not silicon.  One makes computers - the other makes boobs.  We want the boob kind.  Silicone was chosen because it sets off the FTIR well.  If you can find another concoction to do so, please share. =)

Thanks IllustratedLife.
Yeah I really meant boob Silicone spelt it wrong is all. I’ve mixed up opaque Silicone for mold making in the past.
After your description, I realised what I got wrong after reviewing the how to guide.
It seems somewhat strange putting the silicone layer on TOP of the acrylic, I thought it went the other way around, that also means I got the compliant surface wrong too. Am I right now in saying that the silicone is the complaint surface, not the projection surface?

At least, this is Gravano’s composition, actually contradicting the FTIR explanation graphic of Jeff Han’s (http://cs.nyu.edu/~jhan/ftirsense/), pasted into the how to guide too. His doesnt mention a rubbery surface at all, and the diffuser projection surface is on the BOTTOM.

I see now this is why I’m getting confused.

I’m presuming I need to do it Gravano’s way if I want better finger blobs?

And whenabouts are you guys going to sell more of the silicone mats? Cos I think I’m going to be in need of one soon!
Thanks

Stu

Oh one other thing-
Given the problems with using FTIR screens in the light, how good is the Rosco stuff in allowing you to use the touch screen during daylight hours?
This is the IR Camera FTIR systems biggest let down in my opinion.
Thanks

Stu

If you want a compliant surface, you put the projection surface on top (because you couldnt see the projection if it was beneath the rubber).  You can also do it Jeff’s way, but your diffuser will likely also diffuse the IR light, so you wont have clean edges.  Not a problem for most applications, but if you need precision or shape, a diffuser underneath is the bad way to go.

Like I said, I’m using a yoga mat right now, but I’m not having any problems with ambiant IR.  I’m in a room with a 96” window (with thin white fabric blinds over it) and am not doing anything special to block out the light.  I have Rosco Grey, but I don’t have a compliant surface to test it with.  One thing I have noticed in my brief tests with the Rosco is I get very obvious shadows.  It could probably work for Front Illumination.  It seems to diffuse everything but very close shadows.

Rosco is as thin as a few pieces of paper and has the texture of a Fruit Roll-Up.  It’s made of vinyl.

I just got back from Office Depot.  I couldn’t find any tracing paper that set of FTIR.

@ Heliospan, you can order the silicone rubber compliant surface overlay from us, the NUI Group multitouch webshop will be alone in the near future, but you can already order it. You can decide your own dimensions, price is based on 99 EUR per m2 excluding shipping. Depending on the size shipping will costs between 5 and 10 EUR

2) better coupling, brighter blobs
3) rosco grey, alot of people use rosco grey (you can buy a silicone rubber rosco combo from us as well, just mail me at )
4) silicone rubber
5) baffles are along the edges, it seems that if you dont use them that the FTIR is really hard to produce or not to produce at all
6) i suggest the philips spc900nc and buy a non IR block coated lens with it, thats < 100 EUR

theillustratedlife - 07 July 2007 03:49 PM

..  One thing I have noticed in my brief tests with the Rosco is I get very obvious shadows. ..

what you see is probably the ghostimage. maybe this comes from the mirror.. look here under ‘Verbesserungen’

I don’t even have a projector set up.  I get shadows from my fingers on the Rosco.

Hey all.


@Illustrated life - I bought my tracing paper from an expensive arts and crafts shop (not necessarily good quality though) in the UK called HobbyCraft. However they had some great quality tracing paper, at least the best I’ve ever seen not being an artsy person, very thin stuff.
I get as good an IR finger-blob as I do without the tracing paper at all (perhaps with a small loss but nothing significant), this stuff is not your typical cheap as crap 1970’s primary school toilet paper (oh the memories! ).


Yesterday I tried the idea of coming up with my own bouncy surface on the perspex by using a thin layer of latex rubber, applied the same way as described in Gravano’s how-to - applying silicone. Used purely cos I had some lying around. Of course, I dont plan on using it in the end result due to its ‘poor optical qualities’ but I just thought I’d try it out to see if I get a better FTIR effect.
I’ve yet to wait for it to properly cure, but I’ll keep you all posted as to its qualities. I looked at it this morning, and the surface of the rubber is not pleasant to touch. Its not wet (Still curing tho) but its kind of slimey textured and wont allow free finger sliding (Oohh missus). I’ll check tonight for its IR touch properties.


Thanks for your assistance guys. I’ll probably pick up that camera you suggested within a couple of months. I may shell out on getting some of the proper silicone from you, cos I learned I’m no good at applying the stuff to the surface. However Wacker chemicals operate in the UK, it would make sense for me to acquire the silicone from them and do it myself.


The baffles are a bit odd though - you say its nigh on impossible to get the FTIR effect working without them? Then why am I getting it without using baffles? True it is a poor IR image but you guys say this is due to no silicone layer? (Havent tried this Latex thing out yet).


Hmmmmm. Methinks I’m getting conflicting opinions here.


Stu

Perhaps that’s where I misunderstood.  Are you touching the acrylic or the tracing paper?

read the 4 a4 jeff han paper regarding the baffle, i think there is something mentioned in it, i’m not sure tho since its been ages since i read it

the silicone rubber is indeed not nice to touch, thats why you place a projection material on top of it to project the image on and to touch it

you might want to check out the multilayer surface composition of tim roth, which seems to work very good.

@illustratedlife - I’m touching the tracing paper, the tracing paper is on top of the perspex, the ir cam is underneath it all. It works!


@tangible - Blimey, Tim Roth uses a lot of surfaces, I’m thinking sourcing and getting hold of all those layers could be quite difficult, esp for all the custom sizing you’d require, but could very well end up being better for use during the daytime. Thanks for showing us that tho.
But whats the point though of using tracing paper AND a projection surface? Sounds a bit redundant to me.


@gravano - Do u have a link to Jeff Hans paper mentioning this baffle? I only saw his very old website and the various vids of him on the net. Suffice to say, FTIR is definitely working for me without but I’d consider using it if it helps.




Well guys I’ve done some tests with a rubber layer, not using silicone, but with latex. Yes I know, it’s rubbish, in fact it bathes the camera with a lot of IR light due to it’s internal cloudiness and crap optical qualities (it cures only to a semi-transparent appearance).
But the reason was to check to see how well it improves on touch detection, and sure enough, it confirms all your findings. I get a much better image of a finger blob with much less pressure needed, in fact it is much better for tracking finger motion too.


My next step is to get hold of both the rosco surface (I read your blog Gravano- you are testing other rosco rp surfaces?) and some silicone, probably purchased from Wacker directly over here, but I’m still thinking of buying a silicone layer from you to get the perfect finish - all I’m doing at the moment is testing.
However I’ll email you soon about getting hold of some of the rosco surface (maybe I’ll wait till you’ve done these tests of the other rosco types, maybe they are better). I cant find anywhere in the UK that sell it.
Thanks chaps!


Stu


P.S. Are you guys still doing ventrillo sessions on sundays? When is the next one, I think i’ll pop in n say hello!

or just use the most easy setup, having a tracing paper under the acrylic as projection surface, and using a drop of extra vergine olive oil on top (directly on acrylic). i know this is not the method for all apps, but it is real smooth to operate like this. you also have very little loss of light because of just one layer.
the only odd point is that this setup is not good for use with fiducial markers (besides the oil film, markers too diffused since paper under the plate).
i did not find a tracing paper that reflects the ir (for top mounting). the drag feeling on paper is soso. sill looking forward to get a compliant surface from gravano

regards