Sucess! Basic functionality test.
September 7, 2008
I think I’m just about as excited as a six year old on Christmas Eve night. I have great news to report! I now officially have limited basic functionality on my multi-touch table! Granted, right now it is neither multi, nor touch. I am using an IR-pen until I get my illumination set up so all I’ve got is single-point touch.
I am using the marvelous Touché framework and I can’t say how awesome this framework is already. It’s only in it’s second beta incarnation and already has a very beautiful UI, intuitive setup process, and from what I can see without having written any code yet: good development capabilities. Setup was very quick and in about 10 minutes I had everything tweaked how I wanted it. Hopefully in the future we’ll see fiducial support and the ability to tie actions to custom movements instead of being limited by touch, double tap, and pinch. But again, Touché is only in beta 2 so it’s got plenty of open road ahead.
Right now everything is running off my Mac Mini. It makes tracking just slightly laggy, but it’s really not bad at all and I was expecting it to be much worse with such a small computer. The final product will be run off a Mac Pro and as such should have no problem with lag.
I’m now moving on to finishing the design of my IR illumination setup. It’s going to be a hybrid of LLP and FTIR using much safer LED’s instead of lasers. Hopefully in the not-too-distant future I’ll have touch ability done. But for now, this gives me plenty to work with in starting to develop my software.
Covering the back side of my LCD with tracing paper. This diffuses the backlight and gives perfectly even light. My camera now is completely un-noticeable and does not cast a shadow on the screen. The seams of the backlight enclosure have now completely disappeared.
The Unibrain Fire-I camera inside a white enclosure that holds the Thorlabs filter in place. On the left is one of my reflective backlight enclosure panels (thin composite board covered with a sheet of Mylar) The other three sides don’t have the enclosure panels installed yet.
Time to get Touché! Notice how the image has completely even illumination even without the rest of the backlight enclosure in place–thanks to the tracing paper. Also notice how the white areas look grey and a bit dirty when viewed at full size. This is because the tracing paper has a slight texture which you can see from the top. Not a major deal at this point as it just looks like you are looking through a dusty lens. However, this will obviously have to be replaced with a different sheet of texture-free tracing paper.
Touch the glowing white dot! Calibrating the screen. Extra important with the 2.1mm lens on the camera. Touché did an excellent job with calibration and once I had my filters setup correctly I only had to calibrate once and everything was spot-on.
Playing with the thumbnail app. Right now to rotate you have to grab and drag on the corners of the image. Hopefully a future version will allow you to use the two finger “twist” approach that is a bit more common.
Coverflow baby! Oooh yeah!
The little things…
September 5, 2008
I finally received a portion of the exterior paneling that will be bolted on to the aluminum frame and also provide the surround for the backlight. I thought I’d update on the little projects that are keeping me occupied while the rest of the paneling is finishing up it’s design/manufacturing/application process.
One of these is the camera enclosure which will serve to hold the Thorlabs IR filter in place over the camera without having to use poster tack around the lens. It will also completely enclose the camera and hopefully cut down on the dark footprint in the middle of my backlight. The enclosure is a simple box which will be covered with Mylar just like the walls of the backlight.
My second little project was one I just completed today. Since I don’t have my IR-illumination fully designed and implemented I wanted to be able to calibrate and test out the table while I finish the infrared LED frame. My solution was to take an ink pen and convert it into an infrared pointer that I can use as a “single touch” input method.
I pretty much just followed the instructions that I found on this website: http://www.terracode.com/IR_Pen/DIY_IR_Pen.html
I used one of the SFH485 LED’s that I already had and bought the largest barreled ink pen I could find at Walmart (a whopping 98¢ … gotta love Walmart!) I ended up using a AAAA battery which was still 1.5V but would actually fit inside the barrel of the pen with just some minor modifications for the wiring. I also bought a much smaller momentary switch at RadioShack (part number 275-0008 for anyone interested)
I was able to solder everything and fit it all inside the barrel of the pen. The rubber grip on the pen was removable which allowed me to cover up much of the cutting that I did to the barrel to accommodate the wiring. The good thing about this was that it also allowed me to completely cover the momentary switch which gave me an invisible, pressure sensitive on/off point.
Here’s a short video of my results, the second half shows what the pen looks like with my IR bandpass filter over the webcam. As soon as I finish my camera enclosure I can start actually testing it on my table!
Update on backlight enclosure
August 18, 2008
Last week I was at Siggraph in Los Angeles. First time for me, and a very enjoyable experience. Now I’m back, refreshed, and back to work on this project. I thought I’d post a quick update on my backlight enclosure and where I stand right now.
My first material of choice was Styrofoam insulation panels from a local home improvement warehouse. The pro’s were that it was both lightweight, and also white in colour giving me a good starting point. The light reflectivity of white averages to be 84%* which is good, but still gave me some pretty noticeable differences in light as you can see from my previous post.
My next idea was to use chrome paint on top of the styrofoam. Styrofoam and paint is a bit tricky. You can’t spray paint styrofoam because the accelerants will eat away at the styrofoam leaving a mess. The two choices are to use Krylon H2O spray paint, or to coat the styrofoam with a product called Gesso and then using regular spray paint. Both of which can be found at a local craft store. The H2O paint doesn’t eat away at the styrofoam and the Gesso provides a protective coating before painting.
There wasn’t any chrome spray paint available under the Krylon H2O brand so I was left with the Gesso. Unfortunately the Gesso takes a long time to dry. The bottle recommends 24 hours but on the styrofoam it took more like 48 to dry completely.
Unfortunately, after using Krylon Chrome spray paint I ended up with only a slightly shiny grey surface. Not anything near Chrome whatsoever, and much worse than just a plain white surface.
The best option now is just to use an already-reflective surface as my enclosure. Three possibilities that I have considered are Mirrors, or using either Foil or Mylar over the styrofoam.
Mirrors, besides being expensive and fragile, are also heavier which would contribute to the weight of the table. They also aren’t very efficient at reflecting light since the light much pass through glass before hitting the reflective surface. The solution to this is to use first-surface mirrors which have the reflective surface on the top instead of underneath glass. However, this would be incredibly expensive.
Foil or Mylar are the ideal option. Both can be spread over a surface like the styrofoam yielding a lightweight reflective surface in a very short amount of time. Between the two, Mylar is the better option. It is thinner than foil and much stronger. Foil is very easy to wrinkle and tear, both of which would mess up the even reflection. Also, Mylar is more reflective than foil at 92% average* vs. foil at 72% average* Mylar is a bit more expensive than foil, but should theoretically yield much better results.
So, Mylar on top of a lightweight material is what I’m working on. I’m actually getting the pieces for the enclosure custom cut out of a material used to construct tradeshow booths instead of using styrofoam that I’d have to cut and glue myself. I’ll post some pictures once I’ve got everything built and installed.
By the way, Riley Dutton commented on my last post recommended using a sheet of tracing paper directly underneath the LCD to diffuse the light if there is still a noticeable light difference. Great idea and I will definitely try if needed!
* Numbers according to an indoor gardening manual (Gardening Indoors by George F. Van Patten) and quite a few indoor Marijuana growing instructional websites … lol No guarantees how accurate they are, but they sound plausible.
Backlighting the display
July 30, 2008
Time for a much-deserved update! Been busy at work the past couple weeks and then went on vacation for the last half of last week. Back to work now!
So, the replacement LCD arrived and everything works wonderfully again. I have everything hooked up and have moved on to working on the backlight enclosure. As I posted previously, I’m using the original LCD backlight but need to add reflective walls to the sides since the backlight now sits about two feet below the display.
Right now I have begun work using styrofoam insulation sheets. They are white which means no painting and they’re very light and easy to work with. They won’t add a ton of weight to the final product. I have two walls built so far to test and see if things will work like I had thought.
Upon turning everything back on, I was very pleased to notice that this had made a huge improvement over not having any walls. However, there was still a noticeable brightness difference in the area with the backlight and the areas where the walls could be seen. It is very faint, but definitely noticeable.
I happened to have a medium-sized mirror nearby and placed it in front of the styrofoam walls to see how such a surface would look. When the looking at the display, the area with the mirror had no noticeable difference in brightness from the actual backlighting panel.
This is great news. My plan now is to buy a product called Gesso to seal the styrofoam so that I can paint it with some chrome spray paint. This should produce about the same result as the mirror without the added weight or cost of purchasing actual mirrors to use in the wall panels.
I’ll post again once more progress has been made. It should come a bit quicker this time!
Building the skeleton
July 10, 2008
It’s been a while since I’ve updated. This is due to an unfortunate issue that has delayed my progress. I was waiting to see if I could get it resolved and post pictures of the table with the display on. However, it looks like it’s going to take a bit longer than I thought.
For some reason my LCD just stopped turning on for no reason. No relay click when plugging it in, and nothing powering on at all. I had everything mounted except for the power supply board before things stopped working. So, I assumed in the mounting of the power supply I had done something that damaged the board. I bought a new power supply which I received today and plugged in. No luck. Still nothing.
So, now that I know that the display not working was not caused by something I did I’m just going to have to put it back together and send it back in. Guess that’s the chance you take by buying refurbished.
In any case, I thought I’d post some pictures of my progress. Enjoy, and I’ll report back soon!
Notching one of the aluminum pieces. Couldn’t get the snips in at the angle I needed so out comes the trusty Dremel!
Screwing the LCD frame into the aluminum to keep it from flexing when pressure is applied.
An earlier shot of the boards mounted on the back of the table (before the TV died)
Currently you can see through the sides of the table. I’ll be adding white reflective walls surrounding the backlight to fix this issue.
Closeup of the display. LCD frame screwed into aluminum frame attached to the top of the table. Tempered glass above for a nice rigid surface to touch without worrying about someone messing up the LCD. The glass I’m using is temporary pulled from one of my desks. It has a blue tint so I’m going to have to buy a new sheet for the final product.
The finalized rear of the table with the LCD circuit boards attached.
Voila … the final product. This is just the basic skeleton to which the exterior paneling will be attached.
Seagate D.A.V.E.: perfect MT compliment?
July 6, 2008
One of the biggest things that is thought about when designing multi-touch is getting rid of the keyboard and mouse. I think the second biggest thing should be getting external data onto the table easily without using cables or a secondary computer for uploading.
Microsoft’s Surface has done some of this in a very neat way by interfacing with your Bluetooth enabled camera or cell phone and displaying the contents on the table. However, what if you want to add other files beside pictures on your camera? Time to pop open the CD tray or plug in a jump drive.
I was reading a bit about some new technology from Seagate coming out later this year that appears to be the perfect compliment to multi-touch. D.A.V.E. is a small form factor hard drive that includes Bluetooth and WiFi built in.
I can’t wait to get my hands on this when it’s released. I can imagine it would be pretty simple to check for a large blob from the hard drive (or a fiducial marker) and then interface with the Bluetooth to grab and transfer files to and from the drive. Plop the hard drive on the table and interact with it’s contents wirelessly without having to plug anything in!
If I’m able to get ahold of one of these I’ll work on developing some kind of interface (Objective-C/Cocoa based) that I’ll make public open-source via another part of this site that will hopefully be launched in the next few months at http://developme.ntal.ws/
Taking apart the LCD … pics
July 2, 2008
Some pictures of my progress the other night. Completely separated the LCD TFT from the backlight, removed the power and input source boards, and reconnected everything. My Mac Mini is connected via HDMI giving me full 1920×1080 resolution. Things look really nice.
Working on the skeleton of the case now. I hope to post progress and/or pictures tomorrow.
It will work!
June 28, 2008
It’s 3:30AM … way past time for me to be asleep. Sorry if this post isn’t written well or makes very little sense. However, I had to post something real quick as I’ve had time this evening to dig a bit deeper into my display. I was able to rip more of the frame apart and see how this thing is constructed.
The excellent and exciting news is that there looks to be absolutely NO FFC issues! The LCD is attached in two places to the top of the display and as such should cause NO problems working it into the table.
I also plugged in my Mac Mini via a DVI to HDMI cable to see how things look. This was very exciting too because things look amazing. Text is crisp and readable and everything looks super good looking down on things like a table.
I’ll post some pictures of my progress tomorrow. But everything looks good here and my Vizio purchase so far seems to be an excellent choice.
By the way: interesting note … when ripping apart the Vizio display I found out that the company who makes the actual LCD part of the display is LG Philips. Very interesting considered I paid $800 for this Vizio that has the same LCD panel as a $1,200 Philips.
The LCD arrived!
June 26, 2008
I received my 42″ Vizio LCD on Tuesday. Unfortunately due to work I haven’t been able to dig in and start tearing it apart. Last night I had a bit of spare time and decided to take a quick peek at the insides.
I plugged it in and turned on my Blu-Ray player just to make sure everything was in good working condition and there were no dead pixels. Obviously I didn’t want to tear it apart and then find out something wasn’t working … especially since any warranty would be very much void.
After that I tore off the stand and the back of the display. Not too much to look at since it’s all pretty encased in metal. Getting the metal off looks to be a bit tricky as I’ll need to disconnect quite a few wires. Better not dive into that quite yet.
One thing I wanted to look into was whether or not I would have to deal with an FFC issue. While I wasn’t able to do a full investigation I did loosen one piece of the metal casing and lifted it up to see where the wires led. It looks like it might be promising as I see two flat cables running to the top of the display. I couldn’t find any other place where cables were running to the display so hopefully things just attach to the top and make things easy.
I’ll post again once I have time to start disassembling everything for real.
Projector or LCD
June 21, 2008
Quite possibly one of the biggest puzzle pieces toward getting this table completed is the display. This has been proving to be a bit of a headache for me.
My initial plan was to use a projector inside the table to display the image on the surface as the user was touching and interacting with it. This seemed to be the most popular way to go based on the other setups I’ve seen. I ended up purchasing a Toshiba projector (the TDP-T45U) for about $600. It was DLP technology, had 1024×768 resolution, and was a decent price and a recognizable brand.
Apparently I didn’t do enough research on the projector. I expected to have to use a mirror to decrease the throw distance (obviously I don’t want a six foot tall table) Well, upon actually playing around with the projector I realized I would need two mirrors to get the throw distance small enough to work in my setup.
Unfortunately, since my setup is being used in a live music environment this isn’t exactly ideal due to the heavy bass possibly shifting the mirrors around or in worst case scenario cracking them. One mirror would be fine glued to the bottom of the case, but two at odd angles just wouldn’t.
So, I’ve decided to go the route of using an LCD instead. I’ve purchased a 44″ Vizio LCD with full 1920×1080 resolution. A bit more expensive than the projector at just under $900 with shipping. Here is a quick comparison between using a projector and an LCD inside a table:
- Projectors are cheaper at lower resolutions but more expensive at high resolution.
LCD’s are cheaper at high resolution but smaller size and more expensive for larger sizes. - Projectors usually have a large throw distance, so if you want a bigger image you’ll have to place it farther away from the screen–or use mirrors.
LCD’s have no throw distance so you can make your box as small as you want it. The only thing you’re limited by now is making sure your IR camera can see the whole screen. This can conceivably be taken care of by using an array of IR cameras and stitching them together in software. - Projectors require a projection screen to be placed underneath or on top of your table surface for the image to project onto.
LCD’s obviously don’t require any projection screen. Everything’s there! - Projectors require expensive bulbs. For my Toshiba these bulbs cost $350 and have a life expectancy of about 5,000 hours.
LCD’s have a life expectancy of around 50,000 hours. - Projectors require fans to cool the bulbs and aren’t recommended to be put at angles more than a few degrees. Mounting them at sharp angles (which is most often required in table setups) may lessen the life expectancy of the lamps.
LCD’s shouldn’t really require any cooling whatsoever as they don’t generate an abundance of heat.
With these things in mind I’ve decided that LCD is the best route for me. There are some downsides to using an LCD though: you have to tear everything apart and move the electronic components around. You also have to figure out how to backlight the display: whether you use the original backlighting or create your own.
I’ll post more details as I receive the LCD and begin tearing it apart and getting to work!
























