All of these methods for balancing the head on top seem a bit too fiddly to work reliably. Would it be possible to use the split ball method so the head is attached, but turn using some sort of gyroscope?
This looks like a neat idea but the head in the Star Wars video is clearly not mounted like that. If you were to use that idea and still try to make it look like the SW video you would have to have such a small wheel base I think it would be super unstable. The one in your video has a really wide wheel base, for stability I'm sure. Also It gives the "head" a very static and smooth movement, where again the SW video bounces and looks more unstable. I know that bounce may not be realistic to replicate in our world.I'm unclear what was wrong with this idea that was suggested in post #41 - it seems to solve all the issues of building a mostly working replica (at least what we see in the trailer). Since the sphere is totally hollow it also allows scope to have opening panels and so on, provided they only open when they aren't facing directly up or down which is trivial to solve with tilt switches.
I'm pretty sure there'll be a toy version out by Christmas 2015 ;-)
https://www.youtube.com/watch?v=dr5xdpLL58A
Any thoughts?
This looks like a neat idea but the head in the Star Wars video is clearly not mounted like that. If you were to use that idea and still try to make it look like the SW video you would have to have such a small wheel base I think it would be super unstable. The one in your video has a really wide wheel base, for stability I'm sure. Also It gives the "head" a very static and smooth movement, where again the SW video bounces and looks more unstable. I know that bounce may not be realistic to replicate in our world.
Not so, take my idea, counter weight it appropriately then add a spring from a bobble head and there you go, you could even put a micro servo on microcontroller to make it turn it's head every few minutes.Also It gives the "head" a very static and smooth movement, where again the SW video bounces and looks more unstable. I know that bounce may not be realistic to replicate in our world.
Ultimately there are few ways to reproduce the CGI in real life, but I think this is the one that would drive the easiest and be most stable, even if the proportions/look are slightly compromised. It will be interesting to see other attempts though ;-)
Are we 100% sure that it is in fact a CGI character and not a practical effect. To be honest it looks more like a practical effect especially when you compare it to the CGI scene in the background and the look of the ship and speeder effects.
In that case it has to be 100% practical. The chaotic movement of the bouncing head never felt like something that was programed, it was far too natural, probably was a rod straight through both halves of the ball with a similar set up to how I described into a track. Probably self propelled though, since with a rod straight through a gyro wouldn't be needed.Some of it could be practical. Somewhere I read that the ball ran on rails which were digitally removed later or just out of shot in other scenes.
In that case it has to be 100% practical. The chaotic movement of the bouncing head never felt like something that was programed, it was far too natural, probably was a rod straight through both halves of the ball with a similar set up to how I described into a track. Probably self propelled though, since with a rod straight through a gyro wouldn't be needed.
True, but I think the CGI in that scene was to erase the tracks and to make the ball sections movement more fluid, and yes bouncing is not a hard thing to program into an algorithm, but it still behaves in a programed manner and can be noticed it you know what you are looking for. it is a hard thing to put into words what I am trying to convey, but I stand by my hypothesis that this is an almost completely practical effect.Disagree completely - nothing is 100% practical these days ;-) There are some complex physics simulations out there that would make the bouncing trivial to CGI.
Star Wars is fantasy not science fiction. Always has been. You want technobabble, head over to the Trek forums. …
My locomotion theory: The body has an internal structure around which the ball casing moves on ball bearings etc. which are powered to move the droid about (like the mouse in a ball idea)
The head is held in place by a magnet surrounded by 3BBs which run on the inside of the casing while the head itself has a similar magnet/3BBs which protrude slightly from the bottom of the domed head. The two magnets attract each other holding the head in relative position with body's internal structure and the BBs allow the ball's casing to pass between them.
This would allow the spherical body to be solid.
View attachment 410314