Tearing down Disney’s Glow with the Show props

disneys-glow-with-the-show-teardown

[Andy’s] boss recently returned from a trip to Disneyland with a set of light-up [Mickey Mouse] ears in tow. He said that during the event, every set of “Glow with the Show” ears in the crowd changed colors in sync with the performance. After he and some co-workers speculated on how this was pulled off, [Andy’s] boss gave him a new assignment – to find out how the darned things work!

[Andy] carefully disassembled the ears, sharing his findings and speculations with us. Inside, he found a small flexible circuit board powered by three AAA batteries. At the center of the device resides a TI MSP430G2553 which is tasked with controlling the RGB LEDs embedded in the ears.

In one ear, he spotted what he believes to be a Vishay TSMP6000 IR receiver. Vishay-branded or not, he verified that it does indeed pick up IR signals using his oscilloscope and a TV remote. In the other ear, he found a pair of small IR diodes, which he speculates are used to repeat the IR timing/sync signal received in the opposite side of the device.

The synchronization methods seem completely different than those found in the Xylobands we covered a while back, so we’re really intrigued to find out more about technology behind them.

Stick around to see a video of the light show in action, and since [Andy] says he’s willing to entertain any thoughts on how Disney makes their magic happen, be sure to sound off in the comments.

Comments

  1. ColinB says:

    Very cool! It would have been fun to be the design engineer for those props.

    It’s a somewhat similar to the way that the Synchronizing Firefly works. (Posted on HaD previously.) Of course the fireflies have more random and autonomous behavior, while the Disney props are more rigidly and strictly instructed.

    • steve says:

      “Disney props are more rigidly and strictly instructed.”

      Surprise!

    • steve says:

      I really wonder what the transmitter is and what role the internal IR diode plays. There were some hats out of sync in the video…
      Why doesn’t the IR sender doesn’t show up in the video? One should be able to see it, if it is bright enough to cover the area. Maybe it’s in the mickey mouse in the center or something…

      • cde says:

        You can see the ir transmitters. It’s those big poles on either side of the video. They are bright purple when all the hats are on.

        • Lokivoid says:

          Those are high wattage Dental blue/UV LED’s, Those lights are often on wile the hats are not active. If you look at 2:12 in the video, the UV’s are turned off, but on closer inspection you will notice faint dots of white light (I should also point out that IR shows as white on most CMOS sensors, not blue/purple)

      • The IR diodes are probably repeaters. You can’t guarantee the transmitters’ line of sight in a crowd.

  2. StripeyType says:

    Huh. Well, I’ve got a Disney trip coming up. Time to start reverse engineering the IR protocol for these.

    • ColinB says:

      Brilliant! If you could reverse engineer the protocol, you could make a more powerful transmitter and take over the show… that would be awesome.

    • ColinB says:

      If the IR receiver is indeed the TSMP6000, then it is by far the most expensive component in the product. Maybe $4.00 in 1k volumes according to Mouser? Probably 4x the price of any other component. The IR receiver has a unique-looking package, so it’s either the TSMP6000 or a cheap knock-off.

  3. opcode1300 says:

    Look in the video(or others) for the IR. Cant watch the video here at work, but maybe it in there.

  4. ginsublade says:

    what if the internal IR diode is for repeating the show’s program and extending the IR network….just a guess..

    • Terp82 says:

      That was my thought too. The original source would not have to be very powerful to initiate the activity throughout the theater. For that matter, there could be several low power sources to ensure saturation even if there are gaps in the coverage of the seats (empty seats, someone not wearing the ears, etc). Multiple sources could also permit different effects in various areas of the seating – could that be a coming attraction?

  5. barelystable says:

    I would speculate that the IR has nothing to do with show control. These devices work when you’re in the general area of the performance. Something like World of Color has everyone faced in the same direction, but they work if you’re facing in the opposite direction.
    Also, looking at the video, there is no propagation delay. So either everyone actually got the signal from the primary IR transmitter, or it uses a radio command.

    I think the IR is a programming and testing interface. Think JTAG, but wireless. Each hat just passes the commands on. You setup bunch of these in a row on the assembly line, and blast the first one. I’d be interested to see if you gave it a slow enough signal on the front end, if you’d see the retransmission on the back end.

    The video indicates a smattering of pre-programmed, sync’d flashes with additional random flashes with a set color scheme (Aladdin section). They could get all pre-programmed for the show, but I suspect they get a quick dump at the beginning of the show of the script, and then get triggered via radio signal throughout.

    It is the only way I can think of to ensure kinds standing on the ground at knee level will still get their blinking ears.

    • ginsublade says:

      how can it use RF when it has no RF gear on board….it did not look as if it had any in the teardown?

      • Reed says:

        IR is a type of radio. Line of sight, with high squirt rates. Hell, even LEDs are bidi comm devices, if you reverse bias them within the refresh rate (before they are completely discharged).

        I’m not terribly well versed on circuitry, but all you’d need to do is have a receiver, LEDs for color, and ir output.

    • macona says:

      I disagree, look at 2:10, there is a fade of color from from to back and then back again. Something more is going on than we think.

      • charles says:

        Nah, just trickery. A slight delay in rebroadcasting with saturation of the initial signal can account for much. We don’t even know what the device is programmed to do given various input or how many inputs there are.

    • cde says:

      If you bothered to click on the link, and read the guy’s article, you can see that he took the thing apart, and labeled every part. The msp430g2553 has no wireless programming ability, it must be physically programmed via Spy-Bi-Wire, which is brought out to test points on the flex-pcb. It is a very simple, IR based, control, the guy even tested the iR receiver.

  6. cubeberg says:

    Sweet! You could re-program this using a TI Launchpad. It would be very simple to write your own code and get your hat working however you want.

  7. Dr. DFTBA says:

    If it is a repeater network, wouldn’t that make signals start looping around from hat to hat?

    • redbat says:

      If they had each repeater decrement a loop counter, then you handily stop propagation when the counter reaches zero. Like how IP packets have a Time-To-Live field in the header for the same purpose.

      • ryan leach says:

        couldn’t that still cause loop around?

        I think the solutions more simple then that, the lights are at the front repeater at the back, it will only ever flow from the front to the back.

        you could have some sort of ID for the current command like a seek bar, if the new command is older then the current one ignore it.

  8. nlfairbanks says:

    The folks at doityourselfchristmas have been working on the ears for a few weeks now.

    http://doityourselfchristmas.com/forums/showthread.php?20818-Ear-to-Ear-Networking

    One member was inspired to create his own version for use by spectators of his light show.

    http://doityourselfchristmas.com/forums/showthread.php?20918-Infrared-Communication-for-Light-Shows

  9. messmaker says:

    At 1:34 in the video, above, you can see someone placing their hand over the ear and the color not responding the same as the others around it. I wonder if they were just screwing around, doing some testing or if their ears were just not working correctly.

  10. RD says:

    If it were truly IR then shouldn’t you be able to see the light source on the video? To send the commands needed for all those hats the whole area would need to be able to be covered in IR. On a cell phone camera or other video device the emitters (and I imagine fairly powerful ones) should be visible.

    • charles says:

      All cameras that I am aware of have IR filters. Just look at the IR night vision hacks.

      Besides, as stated the device is capable of receiving and transmitting.

  11. The_Truth says:

    For the record… the Disney night time parade uses DTMF codes to maintain sync between all of the floats. I don’t have the audio sample off hand, but I know for a 100% fact that is how they were working (unless they have been updated more recently).

  12. notmyfault2000 says:

    After viewing the video, I now need to go and see the show for… “research.” Yeah. Serious, not-fun research. It may even take multiple sessions.

  13. CoolMod says:

    That’s weird. I think it is more complicated. I agree with barelystable.
    1) I don’t see any light (IR) coming out from the ears. Regardless of the ear colors I’d expect like a white or blue light coming out, would the ears be repeating the signals.
    2) Check how the colors are different (like different programs) for example around 0:55, where some of the ears turn on in one color (close to the wheel) and the ears in the back of the plaza are off and then turn on with another color and patterns.
    Last comment, the poles that somebody mentioned are the light sources don’t seem to be. They sometimes look lit and some other times change colors and are off. Regardless their state the ears change to different colors and patterns, so those aren’t the sources.
    Very cool how they did the whole thing.

    • mrshko says:

      Not likely repeating. The IR can be out of the spectral range for typical camera, and probably intentionally designed in such a way. People will definately takes pictures and video at disneyland. The spectral response of the TSMP6000 is up to 1100nm, that’s way up there.

      If you look in the video, they have lights and speakers in what look like palm trees. They have suspicious looking glass bulbs pointing in all directions. I bet those are IR LEDs or IR bulbs.

      As for the randomness,MSP430s can be programmed with serial numbers and random behavior can be simply set by using a modulo operand on the serial number: if even=red if odd=blue, if %3 blink at 1Hz, if %4 blink at 2Hz.

      This is the smart and cheap way to do wireless. Look at the $30 IR helicopters you can buy at Fry’s electronics. They use IR and I’m actually pretty impressed with Disney. RF is such a nightmare and power hog. Hats on the head give you line of sight from above the crowd.

  14. rud says:

    maybe the source of the IR signal is not in front, but rather in the back (?) remember these IR diodes that were thought to repeat the signals behind ? maybe they just catch the signal from behind ..
    plus it’ll make sense not to blast the IR signal in front of the audience, as cameras would pick it.. whereas in your back .. nobody is looking in the back when you got such a show in front of you .. my 2cent speculations … peace.

  15. Mrshko says:

    What if the secondary diode isn’t for the show, but to have the hats interact with other hats when they are near each other? Anyone have two hats that they can put near each other?

  16. Galane says:

    From watching the video I can see the system has what looks like a multi channel addressing scheme, or more likely there are a few different ID tags that can be addressed.

    At one point during the Pirates of the Caribbean section, half of the ears are flashing blue and the other half are flashing orange.

    During the finale the ears are flashing multiple colors. I assume the different colors show which ID tag they have.

    Another thing I noticed is they have independent control of the left and right ears. At one point they’re all alternate flashing a single color. At another point in the video the left and right ears are showing different colors.

    The simplest hack would be to re-arrange the milticolor LED connections so that your ears would never match the colors the show program is designed for. “Why are your ears alternate flashing green when everyone else’s are orange?”

  17. JohnK says:

    It can’t use IR. If it did it would ruin all of the videos people were taking of the event. Not sure if you have ever noticed, but if you turn on your phones video camera and point a remote at it and start mashing buttons you can see the IR flash of the remote.

    • cde says:

      Except, it does use IR (see he took it apart and found ir receivers and ir leds) but the main thing is that IR leds are focused lights. Even with the remote example, it depends on the direction, intensity, and distance from the camera. At 6 feet, the remote isn’t noticeable. And at less than 6 feet, it’s still a small dot, when aimed directly at the camera sensor. At disney, they can simply have the ir transmitters up high, and off center so it’s not messing with cameras. Simple enough to plan where they should go.

  18. Angel says:

    They need to bring this to Disney World inf Orlando so I can get my hack on.

  19. jglegos says:

    When the whole crowd’s ears are doing the same thing, they’re using a source higher up in the air with more coverage, in order to hit the majority. When there’s a pattern, they probably use a source closer to ground level, positioned to the front/sides, and use propagation to hit the ears behind – so if you want a wheel-like motion of rolling colors, you only hit the people in front, and it gets passed back, using the propagation lag to roll it out. As for the left crowd/right crowd difference, use the same method, but position the emitters to the left/right, and as it propagates back, the two sides meet in the middle.

  20. John Weber says:

    According to several cast members at California Adventure, the ears not only can sync to World of Color and other shows, but they also have the ability to synchronize between each other when not being driven by show data. I would presume that this is at least part of the function of the IR transmit/receive system – it takes up to a minute for a pair of ears to ‘sync’ at which point they will adopt similar behavior so that everyone in your group gets synced up. And to prevent this feature from activating during shows I’m sure that part of the bitstream being transmitted to the ears includes flags to disable that feature (or possibly it automatically turns off for some time period after receipt of any RF based command). Also worth noting that various bits of info I’ve seen online indicate that there is a ‘random’ capability in the RF control stream to allow not only having all ears have identical behavior but also having scenes where some will exhibit a complementary behavior for more variety in the show.

  21. EarsEarsEars says:

    There isn’t any RF built into the hats.

  22. ihateinternetexplorer says:

    They are IR powered. The transmitters are located on the towers with the speakers. I was there at the end of June, purchased a hat with these ears, and learned as much as I could. Here’s the pictures I took: http://imgur.com/a/W3bKP

    • Jay Sandler says:

      Not sure if anyone is following this thread anymore, but I was at Disney World Magic Kingdom tonight and saw the hats and the transmitters all over Main Street. The transmitters looked very similar to the devices in the pics ihateinternetexplorer posted… They were ALL OVER during the Wishes show. Some were on outdoor fixed poles, near lighting and sound equipment and many others were in the windows at the top of the stores on Main Street – the windows are closed during the day, but they opened during the show to reveal the transmitters (my girlfriend actually pointed them out, knowing I’m a techie, asking if I knew what they were). Many different windows at many different angles on both sides of the street.

      Also could the diodes at the back of the hats also be for IR reception? I’ve seen many IR devices that seem to use what looks like a clear LED (most likely an IR transistor) for reception. It wouldn’t make much sense for something requiring Line -of-sight that has to work in a crowd to only receive from one specific direction. – people aren’t always looking the right way at all times.

      if I recall, back in 2000 or so, WDW had a light uppin that would start to light LEDs when you were in the presence of one of their twinkling light up banners they had all over the parks. I believe the banners contained am IR emitter and the pins would receive the signal to trigger the internal electronics. Though I never took one of those apart, the only think I could see from the outside was one “clear LED” that never lit with the rest of them. Again, Mt guess was that this was the IR receiver- I presumed it was an IR transistor.

      J

  23. davefiddes says:

    There’s a Disney video that goes into some of the details behind what the hats can do:

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 96,693 other followers