Hey, I just want to say hello and introduce myself to this striving community.

My name is Daniel Fischer, i’ve been working with “realtime video processing” on linux (mostly gstreamer) for a few years now. Since about a year, i’ve been on-and-off developing a FTIR-based table. I’m still missing a proper mirror and base construction, but the plate is a well-made prototype (although the next prototype will likely be very different). Also, i’m using a plain RoscoGray-on-acrylic surface construction, which doesnt perform too well. It works well for medium to hard pressure, but for “zero-force”, it’s hopeless.

I plan to experiment with silicone rubber overlays in the next few weeks, but alternatively will go with the current model anyway.

I’m very interested in the software and ui design issues. I’m developing my own software based on gstreamer and my own xinf project. Initial results are very nice-- in addition to the “common” blob detection method, i’ll also go for rectangular “sensors” which might give more sensitive results for things like sliders. Until now, there’s enough CPU load left on my (nowadays average) 1.8GHz laptop to do some application stuff in addition to the analysis-- so i hope to avoid having to get two computers for this.

I think we’d need some “professional” physics adivce-- i saw some people put the LEDs itself at a 45° angle to the surface edge-- isn’t that pointless? TIR is about the angle of individual light beams, and your avg LED has 20-60° angle… so straight in should be best, no?

Also, some insight into the silicone rubber thing (how the Frustration of TIR actually happens, and where) would be great. I’ll try to post my questions to the “compatible surface” thread…

i saw your site, it looks interesting.
i think aspecially joobs is glad another person is looking into the overlay topic
regarding the angle of the leds, some say it helps so say it doesnt really improve the performance a lot.
i choose to do a “normal” approach since it saves alot of time not having to angle those leds in a 45 degree angle
interested in what you’re coming up with
good luck

this thread contains some valuable insight into the angle issue. seems something like 20° for a 30° LED makes sense-- if you put a narrow-angle led at a straight angle, you loose a lot of the center beam..

should this kind of info be accumulated in the wiki?

yes it should, the wiki is by some unknown reason low profile. there aren’t alot of people adding content. i wanted to add some more stuff myself, but i can’t for some reason edit the pages. but nui is looking into that. i’m also a bit short on time lately.