I think it’s possible that something was missing from the performance side of things - Frank Oz is a hard act to follow, there. I would disagree that there is a substantial difference in the quality of the puppet itself, though, if you’re objectively looking for “realism.” Empire Yoda’s mouth is still a puppetty mouth, and his hands are very obviously puppet hands. When he “walks,” it’s puppety walking.
I think that the Phantom puppet mainly suffered because we had the original fabrication and performance to compare it to, and where it doesn’t quite match, it registers as “off.” If we switched the puppets around and saw the Empire puppet twenty years after the Phantom puppet, it’s a safe bet that it would be the one that looks a bit “wrong.”
That’s possible, but it doesn’t seem very convincing to me.
You mention a lack of gravitas in the ESB scene you linked to, but there’s no shortage of gravitas there. The puppetry doesn’t detract from the performance, unless you’re the sort to notice a faint hint of imperfect matting or the wiggle of a puppet’s foam ears, I suppose.
I don’t know anyone who’s seen the prequels before the originals. It would be interesting to see what they think and if they have different opinions, but I hesitate at performing cruel experiments on fellow humans.
Unless I’m misreading him, that’s precisely his point. The floppy puppet ears don’t detract from the scene, even though it’s a clear marker that the character is a special effect, because we’re all used to seeing puppets used as special effects, and have learned from a young age to ignore those markers. CGI has its own set of markers that make it stand out as a special effect, but they’re different markers than we’re used to. We’re not used to ignoring them, and so they stand out to us as glaring errors, even when objectively, they tend to be much more minor errors than you’d see in traditional effects. People who have grown up with CGI, on the other hand, have learned to ignore those markers the way we learned to ignore puppet-y walks or matte lines.
Okay, what am I missing here? I stared at that so long I half expected one of those screamer things to pop up. I don’t see the “post” or whatever is wrong. Can someone direct me?
Stunning in its effectiveness. I never considered for a second it was CGI.
But why? Doesn’t this quality of CGI cost a fortune? Is the value of getting whatever the art director dreamed up really worth it for a few seconds on screen, versus going to some wonderful old library or whatever that could produce a similar shot?
No, his argument is that CGI is definitely better than puppetry, and we only don’t think so because we’ve learned to ignore the puppety markers. I’m referring to post #79 here.
My contention is that what stands out as markers to him has never been noticed by the vast majority of people ever. To wit:
Look at the center bottom. You can see a very slight cutout in the light where the post was.
I have an exceptionally hard time believing someone instantly noticed that and was instantly taken out of the movie by it. I call shenanigans. CGI doesn’t suffer from those kinds of imperfections, true, but they’re also insignificant imperfections people don’t notice because they’re not noticeable, not because we were somehow trained to not notice them.
In the pilot episode for the show Firefly, at the end there is a scene where the ship’s pilot sits back in his chair while everyone cheers at having survived.You can clearly see he’s pantomiming holding the controls; nothing is actually in his hands.That’s a pretty major booboo which is clearly on screen, yet nobody ever notices it until either they listen to the commentary or have it pointed out by someone who has. Such things are accidents, but we don’t overlook them to be polite, we genuinely don’t see them.
I love CGI. So much better than that stop motion and model shit and aliens whose only alienness is face paint and bumps on their forehead that look like they could have been done at any county fair’s face painting booth (sorry, Star Trek!). It all looked SO fakey! CGI is so superior to the old techniques that it just isn’t funny, most of the time. I’ll grant you the CGI in “King of the Lost World” was bad … you are not supposed to be able to see through King Kong. But most of the time, CGI is a huge improvement over what came before. That’s why it’s so popular.
Anyway, the point is, no one ever looked at the original Yoda and went, “Holy shit, how did they do that?!” Yoda in ESB is very clearly a puppet. And we’re okay with that, because we all grew up with puppets, and we’re used to how they move and interact with their environment. Yoda in AOTC is clearly CGI - but if you didn’t grow up with CGI, CGI looks wierder than a puppet. For some people, this jerks them out of the movie. But if you grew up with both puppets and CGI, then one doesn’t look any weirder than the other. They’re just two different ways of doing special effects, neither of which perfectly mimics reality.
Maybe I’m splitting hairs here, but I see a difference between “CGI is objectively better, people just trained themselves not to notice the issues with puppetry” and “CGI and puppetry both have issues, we’re just used to puppetry issues but not to CGI issues.” I think he’s saying the former, you’re saying he’s saying the latter.
I grew up with both and I stand by what I’ve posted in this thread thus far (granted, I had a few more extra years of exposure to puppetry, but I strongly doubt that made more of an impression than say the CG in Jurassic Park–which btw, I still think is one of the most convincing uses of cg for characters yet).
It may be a “Larry” thing - I recall Lawrence Kasdan going on about the problem that Yoda’s ears presented in this scene. Earlier takes were unusable because the ears wouldn’t take direction, even when the puppeteers were doing their best to hold them still - transmitted movements made the foam rubber wiggle in appropriate ways. Eventually, they sacrificed the articulation of the ears for this scenes in order to put reinforcing material in there to try to minimize their wiggling.
Obviously, I’m not saying that the original execution of Yoda was a failure - it was a triumph of its time - just that there’s no rational basis for objecting to a well-executed CGI character on the basis of “You can totally tell that it isn’t real,” because you can totally tell that a puppet isn’t real, too - and in ways that are generally much more readily apparent than the ways that CGI characters appear “unreal.”
You are right in that I’m saying CGI is objectively better at presenting a “realistic” character that doesn’t challenge your suspension of disbelief, (with the caveat that practical effects may be more appropriate in certain circumstances.)
With the current state of the art, most objections to the “realism” of CGI characters comes down to “I know that can’t be real, so it looks fake.” This is sort of an uncanny valley effect. As a thought experiment, look at the CGI Yoda and try to articulate what it is about it , specifically, that blows the realism for you. The articulation is good, the lighting is good, the material modelling is good. The animation is very good. What is it about him that might “pull you out of the movie?” It’s a safe bet that it’s largely just because he’s a three foot high, tri-toed, pointy-eared little green dude that’s walking around just like he was a living, breathing creature, and this puts his address somewhere near the uncanny valley.
Currently, when CGI is used for the run-of-the-mill, ordinary things which your brain has no objection to sharing a set with live actors, people who are very quick to say that CGI generally looks “fake” will usually not even blink.
Are all those CGI? It looked to me that some of them could just be composited actual footage to save on location shooting. How much does it cost to buy a stock photo of the San Francisco skyline? I’ve found sites that will sell HD video shots of almost anything quite cheaply. I had a friend that wanted to shoot a commercial on a tropical beach. It turned out that getting the rights for the footage he wanted was quite cheap.
Let’s all just agree that Green Lantern had terrible CGI. Like horrendous. The whole thing reminded me of the Mars Attacks aliens. (Great movie btw, Mars Attacks, not Green Lantern)
Mentioned before, but these are all Virtual Backlot shots from Stargate Studios.
They’re differ from straightforward composite shots in that the virtual location is shot with a 360 degree panoramic rig. Using camera tracking (and various other variables such as lens type, etc) from the on-set cameras, the location is rendered to seamlessly integrate with the footage from the set. No individual frame of the rendered location shot comes from a single camera; it is Computer Generated Imagery.
Watch this more current reel and pay attention to how the camera moves in the finished shots. What makes this tech special is that they don’t have to carefully set up the set shot to match the location - they can move the camera with a relatively large degree of freedom and computers extrapolate the frames needed to match that motion when rendering the location.
In terms of puppet vs. CGI compare the finales of *Aliens *to the first Alien vs Predator. I just rewatched AVP and its alien queen, while not fake looking, is just wholly uninspired and not all that scary. But her appearance at the end of *Aliens *is the stuff of nightmares. And in all the shots of her including the climactic hand-to-hand, close-in fight with the loader, even though it was shot with both full-size and small-scale puppets of the queen & the forklift, not one shot makes you think “puppet”.
Actually that’s not true. There is one disappointing shot, the very last one of the queen flailing its limbs in outer space. Why? Because its the only one that’s (obviously) stop-motion!
Keep in mind, the main reasons for these differences aren’t technical, but artistic. It’s more than two different SFX methods, its attention to story, pacing, editing, lighting etc.
Okay, you guys use the term “Uncanny valley” in this thread, but it seems you don’t understand what it means, as it explains this perfectly. The closer you get to realistic, the more we will notice the imperfections. And, yes, that means that, in a finished production, you should shoot for coming out before the valley until the point where you can actually come out on the other side. In no other artistic endeavor do you get to put out technically more advanced items that look worse. You can’t define realism objectively–you must go subjective, along the curve. Or, as mentioned earlier, do what they did with puppets and hide the fact that it’s CGI.
That said, if you want really unrealistic CGI, go look at Santa Buddies. That’s true cheap CGI, rather than this stuff that’s just barely wrong.
While using CGI for this specific purpose is rather… redundant, and while there’s still minor flaws (though personally I think it’s one of the first CG humans that successfully crosses the Uncanny Valley), I think this video does an excellent job at showing what CGI can hypothetically do. There are pretty good robots, but they’re going to be much more expensive to get this realistic, and you can’t shoot a robot in the head and pass it off as a human on screen.
Obviously this kind of quality isn’t quite ready for movies yet (after all, the best we can do is mimic a human which is, at best, redundant, you need to DO something that you can’t do with a real actor for it to pay off). But I think that while for lower-end purposes puppets, makeup, and costumes can work well (and in some cases, even better), I don’t think they can beat the theoretical upper limit of CGI.
I watched and I can see what you are saying, but I’m not sure this is really CGI or computerized video editing. I can see a few CGI objects that are inserted in some scenes, but most of the scenes are actual real objects that actually exist outside a computer and not what I think of as CGI, which is virtual objects that never existed outside a computer.
Look at how it’s being used, though “he’s a three foot high, tri-toed, pointy-eared little green dude that’s walking around just like he was a living, breathing creature, and this puts his address somewhere near the uncanny valley.”
I think most people find that quality CGI fantasy characters are already outside the uncanny valley, and no amount of “more realistic” is going to remedy that, because the subject is something that is intrinsically unreal. It can’t be real, so it looks fake. The only solution is to avoid telling stories that involve impossible characters, while detractors insist that going backwards would yield a result that actually looks more real.
It’s kind of like this: “Hey, I booked a hotel for our trip to California - the Bayfront in San Diego.” “Ugh, I don’t want to stay in the Valley - it’s so fake! Can’t we stay at the Burbank Marriott again?”
I think you are probably thinking specifically of 3D modelling - many people think 3D modelling when they think “CGI,” because it is the most obvious sort of CGI. But any computer generated imagery is a CGI element, and this is computer generated in the sense that it simulates the motion of a single camera by producing the necessary frames by interpolation and anamorphic manipulation of frames the array. The rendered images show a real scene from angles which were never actually photographed. You can think of the camera as a virtual object being modeled in this case.