IR Illuminators test

About a week ago I got the illuminators in, but surprise surprise, no adapter or what so ever included. So I plundered through my house and found a couple. After connecting the Illuminators I uhm.. got disappointed. They worked and all, but the beam angle was too small and there was no way to adjust this; I tried aluminum foil wrapped around a piece of cardboard where the light should be bouncing off at an increased beam angle, but it only made things worse. I've tried to reposition the illuminators to the side, from the top, adjusting the height of my screen. Nothing.

All in all, the illuminators can't get the entire screen lightened up and diffused enough for TouchLib to see my fingers. Bummer.

So, these illuminators themselves won't suffice. Buying more of them would be like killing a fly with a shotgun; it can be done much more efficient, cheaper and easier. So.. where have I left these 940nm IR LEDs..?

These will do I suppose. I bought them for an FTIR setup but they didn't work with that, but I guess they should work for DI. Maybe not alone, but combined with the illuminators the screen should work. I'm going to do some calculation on the beam angle for these (as I can easily adjust it when, and maybe even after, I've soldered them in a circuit board) and solder them tuesday or wednesday at school.

No pictures or movies of the tests this time, because my mobile phone camera deceased a couple of weeks ago. And, there isn't really anything which I can show with exitement. :-(

However I did make a picture of the working illuminators. Enlightening huh?



Another setback, but I'll keep looking for solutions. I know it's a proces of trial and error, and I also know that I'll get my screen working.. Once..

IR Illuminator

It's been a while since I posted anything about the progress of my Multitouch screen and that has only got one reason: it didn't make progress. Untill now that is. I just called my uncle about there IR illuminators I found here in the Netherlands on a website and once again he said he would try to order them for me! Thanks man!

I've ordered two Illuminators because they only contain 28 ir leds and I though that just one wouldn't be enough to cover the whole screen. For the Dutch folks out here who need these things ordered aswell, just search the http://www.conrad.nl/ website.

Here's a picture of what the Illuminators look like.

The website said that the product should be delivered within 4 days so I'm hoping they'll be here before the weekend. I'll make another post when I get them and ofcourse, I'll test em right away! ;-)

Basic test

I've just done a simple test after making the setup. As you can see on the vids below my fingers have the IR LEDs on them so the real toucheffect isn't there yet, but I hope I can finish that off soon. I thought it would be worth posting this as I enjoyed making these vids. :-)

IR LED test

Today I did a test with one of the IR leds I got last week. I found a little led mounted in a plastic finger holder with a swich on it and which holds onto your finger using an elastic band. I disassebled it, cut the led off and taped one of the IR leds onto the remaining wires. To my surprise it did actually work. Below is a picture.

I initially wanted to solder some leds together in school to check if it would work but now I could check the FTIR effect with just one led. Although I didn't expect it to work it was still a disappointment.. There were completely no blobs at all, not even on the rims. Using it for DI was pretty good though, but I didn't have a diffuser and only 1 led didn't do a lot of magic.

What I could use this for though, was to just hold the led above the webcam and run the smoke app. It gave nice looking results but you would have seen pretty much the same with an ordinary flashlight. Here's what the app showed.

That was about it. I will be ordering an IR illuminator soon and I hope to speed things up after that. Untill next time..

IR-LEDs arrived and change of planns

First off, I would like to wish everyone a happy and healthy 2008. And I'm sorry for being away from this blog for so long, I was kinda busy around the holidays..

Anyhow, I got my IR-LEDs today from my uncle. As I mentioned in the first post he bought them on eBay and voilá, now they're here.. But like I mentioned aswell these are not the right ones, since the ir/mW is not high enough for a multi-touch table according to several people from the NUI community. Though I won't give up easily and while I have them here I might aswell give it a try; I'll ask my Physics teacher if I can connect a couple and see how it reacts to the plexiglass. Here are some photo's by the way:

Now about the way I planned to build my multi-touch screen. Originally the idea was to make an FTIR system but after looking at the pro's and con's I decided I'm going to build a DI setup first. After I finished it I might make an FTIR screen eventually though. A DI setup is just easier to build, doesn't require a compliant layer, works in ambient light and it's a bit cheaper.

I found this IR illuminator on eBay but I don't know yet if it has the specifics I need for my screen, nor do I know if just one will be enough. Any help about IR illuminators is greatly appreciated.
I know that gravano is using a 140 LED IR illuminator and I'm actually waiting how it's gonna work. I might order the 48 LEDs illuminator next week though.. more about that later.

Plexiglass arrived

Around 6:00 pm today a friend of ours came by and brought some plexiglass for me from work. I gave him the measurements 54x72 cm (which is a 4:3 ratio) but he kinda lost them so he had to guess a little. He gave me two piece of plexiglass; one 51x72 and the other 54x74, both 10 mm thick.

As you can see I'll probbably use the 54x74 since that one comes closest to the 4:3 ratio. I can use the other one if I want to build another prototype after this one or simply as a backup, incase this one fails. The best part is that I got two pieces of plexiglass for free! So all I can say is thanks Jan! :-)

Here are some pictures I took with my phone, so the quality is bad. I will post new ones if I get the chance to use my sister's camera. I will be sanding the edges soon, so I can get on with the LEDs when I get those. As you can see btw the 3rd picture shows some white stuff on the plexiglass. This is the foil that used to cover that side, but since these pieces have been laying around somewhere for quite some time the foil kinda shinked and now it's hard to remove. It took me an hour orso to clear one of the screens, so I guess tomorrow I'll have to find some time to do the other one.

The first two pictures show my first setup (right..). I used the light of my phone's camera to shine through the plexiglass on the right side where I took some of the blue foil off and I place my webcam underneath. It was kinda cool to see for the first time, but other then that it was pretty useless to do. :P

Getting in contact

Today a Hungarian called Ferenc Szekeres contacted about software for multitouch devices. He told me that he and a friend of him set up a company and they were planning to make an application with blob detection, similar to Touchlib. They asked me if I was interested in joining them to create such applications using either vvvv or Programming.

Since I do have a basic understanding of programming in Java, which will probbably mean that I can learn the other languages quite fast now, I agreed with him that they would contact me if they needed help on small things and small aplications. This way I could learn and be of use. Here are some pictures of their product presentation:

I'm glad they're giving me this opportunity, since making multi-touch software seems to be a hot item and I would really like to create some apps of my own. I'll keep this blog updated whenever they assign me with something. :-)

In the meanwhile, my setup itself isn't really coming together, since my dad had to get on the road for his work and I can't order anything if he isn't here.. ah well, I hope I can order the LEDs by the end of the week. More information about that later.