For example if Mickey were to say "Hi Robert Niles" would the keyboardist press one key (call it Function One) to ask a name and then when little Robert says his name the keyboardist would type in the name and then Function 2.
Comments like "I like you princess shirt" would be generic enough that the keyboardist could point and click with a mouse.
That seems so much easier than the NextGen bracelet.
And don't get me wrong, Adam. The NextGen bracelets will probably trip the video cams in the room to record the event. And you are correct that it will revolutionize the theme park experience.
But I think that mirror (one way glass) is an indicator that this is a guest to "live-ware" experience.
word to the anthropomorphic mouse!!!!!!!!!!!!!!!
Unlike Crush, Mickey has physical contact with the guests and is responding to their non-verbal cues.
And, incidently, this effect unbelievably charming and the inevitable parade of You Tube clips is just gonna blow this thing up.
Watch both of those videos and closely examine the timing of the interaction. Look at how the character interacts, and how timely some of the questions and responses are. Look at the movement of the mouth, hands and head.
If there were a second hidden party directing this, then essentially the character would be a human puppet. However, it would be IMPOSSIBLE for someone to react in such a precise timely manner via voice instructions from another.
The first poster is also wrong since Turtle Talk with Crush is an actor and the voice varies greatly based on who is performing.
This is not a voice actor. This is a computer-generated voice, not an actor’s and I believe the character inside the costume is actually talking.
I suspect there is a wireless microphone inside the costume that the character uses. That input is sent wirelessly to a computer that does the voice translation and it is then output to a speaker that is probably in the room, not actually on the costume. However, it could be on the costume.
If the sound is loud enough then they’ll eliminate the chance of the person inside the costume being heard. With this set-up it would be entirely possible for characters to speak in designated environments within the park.
There may be some limitations on words, but maybe not. Knowing Disney though there is likely an approved script.
WORD to the "MICKTRIX"
I'm assuming that there's a speaker in the head, too, and that the "voice" isn't coming from speakers in the room. (Can't tell for certain only from watching a video.)
The only way that I can see this working as it appears to be is that the individual castmember in the suit is doing the talking... its too dynamic and interactive for there to be a keyed-up pre-programmed response or someone else doing it. Mickey seamlessly gestures and speaks in a simultaneous and spontaneous manner, choreographing such antics would be pretty much impossible.
I Respond: He doesn't need "voice instructions." A microphone on the costume would allow the person behind the mirror listen and respond.
Eric G writes: This is not a voice actor. This is a computer-generated voice, not an actor’s.
I respond: Which is what I said in my first post on the thread.
I am trying to figure out though if my three year old daughter would be less scared to see Mickey talking or more scared. Hmmmmm.....
This is really a super extension of the live character initiative, and an ingenious incorporation of Disney’s existing technology taken to the next level. Culminating in the finest free moving, highly interactive character meet and greet, which deservedly has been bestowed upon the mouse himself. I can't help but wonder how amazed and delighted Uncle Walt would be to see this.
Tiny servos operate the eyes and mouth, with a sensor near the persons own eye and mouth. When the actor blinks, Mickey's eyes blink and when the actor moves their mouth, Mickey's mouth moves. That's why Mickey's mouth moves almost in sinc with the voice.
The voice is reproduced by a computer, picked up by a mike in the costume head. As the actor speaks softly, the mike transmits it to the computer, which converts it into Mickey's voice and instantly transmits it back to the speaker inside the head, making it look like Mickey's actually saying the words. That's why Mickey can interact so quickly with a guest, because it's the actor responsible for all the movements and voice.
Right now they want to use it in a control environment, because of needing the computer to pick up the transmission. Once they solve the problem of outside interference, they should be able to use Mickey anywhere. This is only a guess on my part, but it explains how everything could work.
Anyway no interactive character has ever been the physical entity but somebody off to the side. Check out push the can revealed on YouTube . Very sneaky! Also Mr TH likely has a pretty good source on this kind of things like this. I know he has been right before. TH is a true insider!
I'd have to disagree. You have to take into account the timing between the voice and the interaction. Especially look at the second video, on the line, "Just don't tell Minnie" and when he's arranging them to take the picture. There's no way a behind the scenes voice could time that right.
My thoughts mirror Bob Miller's (above) that the voice is replicated by a computer (think auto-tune, Mickey style) fed by the in-suit CM. That's the most likely scenario, and the most replicable option.
However, surely they could have found ONE cast member out of thousands that could emulate Mickey's voice--and in that case, it would just be the facial animations and speaker for the suit. The suit would be easily made again, but finding that cast member for other parks...not so much. That's why I'm leaning on the first option.
The voice and movement are way too in-sync for the voice to come from somewhere else.
Also, I don't like it when people say "ABSOLUTELY NO WAY" and they don't back up their statement with some explanation. Convince us that you're right, because I think there is "ABSOLUTELY NO WAY" it isn't the cast member in the costume.
It would be such a dead giveaway in the videos if someone else were doing the talking and directing. The movement from the costumed character would not be entirely in sync.
In response to others comments, I'd also like to add that if the speaker is in the costume there would be no worry about having to talk quietly. The time required to alter the voice should be almost instantaneous, so the speaker would overpower the cast member talking. It's also super easy to soundproof the costume.
Has anyone here checked out the Auto-Tune app from T-Pain for the iPhone? If some mediocre rapper can make that garbage app work, then Disney can certainly accomplish this task. I even use voice recognition software for typing and it's amazing how well it works. Talk and it types it out.
I'm going with the cast member in the suit speaks and is converted to Mickey voice. The syncing of actions and mouth with dialogue is too accurate for a wizard behind a curtain.
I'm a little shocked at the length of time it took to implement this. It seems like such a given. You'd think that immediately after the validation case ("Crush") they would have began introducing this.
Also, it might be wise of them not to announce it, but rather slowly phase out the inanimate character faces.