For ‘Talking Mickey,’ Speech is Just the Tip of Technological Iceberg (Magic Spoilers)

'Simulated Conversation by Pre-Recorded Audio Navigator'The official Disney Parks blog today announced that commencing today — and barring any technical issues — Mickey Mouse will not only be meeting and greeting with guests at the Magic Kingdom, but conversing them as well, marking a new major milestone in a very lengthy history of attempting to bring characters to life in all new ways ahead of full NextGen technology implementation.

Interacting with characters is nothing new in particular, but it took a major leap in recent years at Disney California Adventure, where Roz in ‘Monster’s Inc. Mike and Sulley to the Rescue’ is able to use the actual talent’s voice (Pixar’s own Bob Peterson) to present a custom greeting to an on-board guest via a soundboard app. Controlled behind the scenes, a Cast Member is able to string together pre-recorded phrases together and make it appear as if Roz is addressing a guest directly. The same technology went on to be used for Mr. Potato Head in ‘Toy Story Mania!’ as well as the meet and greet Lightning McQueen and Mater in DCA’s Carsland, albeit not to as great an effect. There is, of course, ‘Turtle Talk with Crush,’ which uses exceptional technology in its own rights, but there Crush is on a screen and not a fully tangible being, let alone one that can traverse in physical space.

Talking Mickey is not exactly new to the scene either and has been spotted in testing off and on over the past few years, but he hasn’t been quite ready for primetime for several fairly obvious reasons, none of which have anything directly to do with the fact that he can talk.

The evolution can pretty much start with the appearance of the articulated heads of several Disney characters often seen in staged productions at the North American parks as well as videos and television appearances. Much of the time, these characters perform to track and are often completely single-operator controlled through finger gestures that control blinking and mouth movements. The major obstacle with bringing these articulated characters down from the stage to meet and greet scenarios, however, is that much like your common off-the-shelf toy, the servos and other electronics generate noise when activated (or even when idle in some cases), which can prove a bit disconcerting. Other issues including generated heat and wear and tear were also of concern. These issues were addressed primarily in US Patent Application 20130130585, ‘Method and System for Articulated Character Head Actuation and Control.’

Another issue is that in order to preserve a ‘natural’ appearance, characters blink. That problem becomes obvious to anyone who’s attempted to take photos of one of the staged performances featuring the articulated heads — there is nothing attractive about an articulated character with its eyes shut, nor nothing appropriate about a character with their eyelids half-shut, giving them the ‘wasted’ appearance. If the blinking were operator controlled in meet and greets as they are on stage, this wouldn’t be an issue, but since operators can’t physically control blinking for reasons of show, the heads blink on their own, which provides an opportunity for some interesting family photos to say the least. This issue was recognized and addressed in US Patent Application 20130073087, ‘System for Controlling Robotic Characters to Enhance Photographic Results.’ The proposed patent covers several methods including both audio and visual cues to indicate to the character that they should not blink for an impending photo. From what information we have unofficially gathered, it sounds as though this may have been implemented in a way that the character will not blink when exposed to red-eye reduction, or a series of mini-flashes from the camera before the final flash. We were told that Disney’s PhotoPass should not experience the problem, but guests’ cameras will have varying degrees of luck (thus we’d suggest making sure both flash and red-eye reduction is turned on).

And, of course, there is the issue of actually talking. Followers of the Talking Mickey saga have seen two different flavors of this being tested. One employs an actual human being supplying the voice for Mickey in real-time. Although this offers an extremely high level of engagement between guests and the character, it’s impractical to offer on a full-time basis. It’s considered an expensive venture to train a human to voice the character and then contractual agreements and the like can cause scheduling conflicts, leaving the character unable to be voiced. The second flavor involves a soundboard, much like other characters previously mentioned in this article use. While this solves the issue of voice consistency and availability, it offers just a limited number of phrases at the ready and there can be an unfortunate delay in response time of the characters as well as awkward conversation that doesn’t really mesh well. These issues are addressed in US Patent Application 20130226588, ‘Simulated Conversation by Pre-Recorded Audio Navigator’ in which the phrases available on the soundboard can be dynamically altered based on several forms of input including voice recognition, allowing the character to seamlessly converse with guests (for the most part).

Another form of input that Mickey can respond to, although we are told this will not be enabled for some time, is Disney’s MagicBand technology. Although the patent application seems to fall short of Mickey addressing anyone by name as face characters will have an easier time doing, he will at least be able to recognize birthdays and other special events and respond accordingly. Although how Mickey will know it’s your birthday, especially if you aren’t wearing a button, is a whole new level of creepiness.

Enjoyed this post? Share it!

 

4 thoughts on “For ‘Talking Mickey,’ Speech is Just the Tip of Technological Iceberg (Magic Spoilers)

Leave a comment

Your email address will not be published. Required fields are marked *


one + = 7