I started building a multi-touch table a few years ago with the help of the community, but after many failed attempts I abandoned the project. I recently starting playing D&D again and now have a reason to finish it! It is very closely modeled after SpinalTom’s 42” Table. I really liked the design and presentation. That, and using 2x4s as a frame was brilliant! Saved money and time in the assembly process, plus added the much needed rigidity.
Without this community I would have been completely lost and I probably not have even thought of doing something like this, so I thank you all very much for your vast wisdom and suggestions.
Here’s the specs of the table:
Camera: PS3 Eye with 2.8mm lens and 850nm filter from Peau Productions. (Completely worth it. It was like going from VHS to Blu Ray when I removed the floppy disk and installed the 850nm filter.)
Surface/Diffuse: 30” x 30” x 1/4” acrylic with vellum sheet.
IR Lights:4 x 48IR LED illuminators 12V 0.5A 850nm
Mirror: 12” x 10” from ACE Hardware
Projector: Epson S4 (sometimes a Dell 3200MP)
I don’t have many pictures of the assembly process as that all happened a few years ago, but I can take pictures of what it looks like currently. Any suggestions are more than encourage as I still need some help here and there fine tuning it and getting it to perform how I would like.
The acrylic looks like it’s flexing but that is just the picture making it look that way. Very strange.
This one shows roughly where the mirror will go. I used to mount the project on the side wall, but have since started putting it on the ground. It has a very short throw so it all works out!
I’ve been using the compliant surface on the underside because when I have the vellum on top I get vicious hotspots that come back every time I touch the top. Is there a better way of fine tuning CCV to ignore them? I’ve thought about dynamic subtraction, but that doesn’t seem to work as well as I was hoping.
Finger Size: 15
Finger Sensitivity: 200? (Not sure how much it affects the overall tracking)
Reactivision Camera Settings:
Source (vellum on top):
Tracking (vellum on top):
Source (vellum underneath):
Tracking (vellum underneath):
My blob detection is decent I suppose, but configuring ReacTIVision is nothing but easy.
I have some questions for the community too!
Would switching to 1/4” glass instead of acrylic be a better choice?
How can I reduce the glare when I have the vellum on top? I’ve tried angling the lights, but every time the acrylic flexes it detects new blobs regardless of orientation. Putting vellum on the lights themselves doesn’t seem to diffuse the light as much as I was hoping either.
The box is not completely light-proof. Would taping up the joints, edges and gaps improve my readouts that much? Same with lining the inside. I’ve read mixed things about both, but nothing conclusive. I also recently found out my power supply is 12V 1.25A and not 2A. I ordered a 2A so hopefully the additional amperage to each illuminator will increase their output and improve tracking!
I cannot seam to get the Flash XML to work either. I always get dreaded “Could not bind or send TCP to Port 3000”. I made sure the global flash settings were correct and that I had the standalone flash projector. I load up CCV, calibrate it, return to menu by pressing x. I fire up a flash demo and then I go back to CCV and press F. Nothing.
Someone mentioned the mouse driver interfering, but I didn’t know how I could prevent that. Any help would be much appreciated.