A bad day at the box factory. from jrc313 on Vimeo

As the title suggests, this little 600 frame movie took rather a long time to render. I think I need to invest in a new machine before I make any more attempts to render pointless, but beautiful movies.

You can have a look at a realtime version of the sketch over here and the source for the P5Sunflow version can be downloaded here.

Sadly, my hopes were crushed.

While the silicone layer appears ultra smooth with no air bubbles and is pretty clear (I hesitate to call it crystal clear), it still requires a lot of force during movement to set off the FTIR. I’m not entirely sure what the issue is. It could be that my inaccurate measurement of the liquid silicone and catalyst have resulted in something that is too hard or that the optical properties of my chosen silicone rubber aren’t ideal. Either way, I’m going to need to apply a new layer before I can continue. Fortunately, I’ve found a couple of distributors of the more popular silicone - Sorta Clear. Most folks seem to have pretty good results with Sorta Clear, so I’ll try it out and see what happens.

Pouring the silicone was pretty nerve wracking. I knew that the compound had to be mixed very precisely (5:1 by weight in my case) and that others had had some issues with air bubbles or their layer remaining tacky after curing. Fortunately it seems to have worked out quite well, although I won’t know for sure until I try it out on the table again.

Anywho, on with the pics …

The two part compound. The big tub on the right is the silicone rubber itself, while the small bottle on the left is the catalyst.

Weighing out the components with the only scales I could lay my hands on.

20 minutes of stirring later. Getting slightly worried about the number of air bubbles.

The spreader. Made using some of the MDF off-cuts and a piece of pine dowel. The MDF is 9mm thick compared to the acrylic’s 8mm. So, I should end up with a nice 1mm silicone layer. Underneath the spreader is a big ass sheet of cellophane to protect the dining table.

Silicone applied and ready to stashed for curing. At this point I was already chuffed with how smooth it was looking. As you can tell from the large puddle of silicone, it was pretty difficult to judge how much to apply.

Should be testing the compliant layer early next week and I’ll post an update here shortly afterwards.

The U rail is fixed in place with No More Nails, which is frighteningly strong. One side is left open for the moment so that the acrylic can be removed for access to the monitor. Once everything is ready I’ll close the gap.

Another view of the underside. Things are looking a lot busier under there now. The LEDs are all wired into a molex connector which will plug into the PSU for the machine that will be powering the table.

Here’s the first test grab of interaction with the surface. This was before I had added a visible light filter to the camera, which is why there appears to be colour in the image.

Although the blobs show up really well here, they are much less visible while dragging. Going to have to add a compliant layer to the acrylic to resolve that.

First the legs are attached to the table surface. The panel at the back with the white hinges is the base of the table.

Base is attached, now it’s time to sand the hole for the monitor … sand, test, sand, test … repeat ad infinitum.

Monitor mounted. Cat is happy with the results and endorses the drinking of a celebratory beer.

The monitor still works … huzzah! Time for another celebratory beer. The panel to the left of the monitor contains the controls for the monitor. They’re going to remain as they are for a nice lo-tech look. The pencil outline around the monitor shows where the acrylic will be placed. As you can see, there’s plenty of space to scale up to a larger monitor … *scours ebay for cheap 19″ widescreen monitor*

Underneath the table. The terminal blocks are there to extend the backlight cables. The corners are from a piece of hardboard and are used to hold the monitor components in place. Going to have to replace that white backing sheet as it blocks IR light. Tracing paper is candidate number one at the moment.

Fortunately it was “relatively” easy to dismantle and all of the circuitry was attached by a ribbon cable secured to the top edge of the panel. Once it was prepared, it was laid out on the dining table and ready to be plugged in to make sure that I hadn’t damaged anything during the deconstruction.

Stood well back at this point and powered it up. As luck would have it, the damn thing worked! Only one of the backlights are plugged in at this point, which is why the top of the screen appears much brighter than the bottom.

A while back I was messing around with using a webcam in Processing to track movement and ended up happening across the fiducial system employed by Reactivision for their rather spiffy Reactable.

I’d originally planned to build a fiducial driven coffee table to use in reception at work for displaying staff profiles, case studies and the like. However, it quickly became apparent that the fiducial markers would obscure quite a lot of the screen and I’d need to make a massive table to accommodate both the markers and the information I wanted to display.

After a spot of googling research I  happened across the rather splendid work being carried out by NUI Group and the many talented denizens of their forums. Here were a group of individuals who, while everyone was marvelling at seeing Jeff Han’s TED Talks demos, were actually building their own multitouch systems.

Inspired by what appeared to be a return to the heady days of building your own computer (I don’t mean going to dabs and buying components. I’m thinking more of the type of thing Woz was up to in the days before Apple was incorporated) I felt compelled to join in the fun and build my own multitouch system.

The majority of multitouch systems use a projector for the display, but projectors are expensive and I wanted something a tad more reasonable for a my first attempt. A couple of the NUI Group forum members were building multitouch systems with LCD monitors instead of projectors and I was fortunate enough to acquire a free one from work so I was off and running.

Just draw with mouse and the objects will drop into the scene. I’ve not added any checks for complex polygons, so try not to draw any shapes which intersect themselves as they tend to freak out a bit. Press R to reset the scene.

Source

The track listing is updated every two and a half minutes while Dashboard is displayed. In order to conserve resources, the tracks aren’t updated while Dashboard is hidden.

There are a number of different colour schemes to choose from. So, if you want to keep multiple copies open to watch more than one person’s tracks, you can easily spot which is which.

Mac OS X 10.4 Tiger is required. If you’re using Safari, click the download link. When the widget download is complete, show Dashboard, click the Plus sign to display the Widget Bar and click the widget’s icon in the Widget Bar to open it. If you’re using a browser other than Safari, click the download link. When the widget download is complete, unarchive it and place it in /Library/Widgets/ in your home folder. show Dashboard, click the Plus sign to display the Widget Bar and click the widget’s icon in the Widget Bar to open it.

Download Now!

Screenshots

Instructions

Drag any of the vertices around to control the “cloth”. Hold down the space bar while releasing the mouse button to “pin” the vertex.

Source

You can grab the source code for this if you really want to. You’ll need the traer.physics library if you want to run it in PDE.