3D animation: digital Tom Cruise?

This thread might be a bit off topic, but I couldn’t find anywhere to fit this, so here goes:

Lately more and more movies use 3D animation to fill in parts of the movie which would otherwise be extremely difficult to realisticaly create. We think of the dinosaurs in Jurassic Park, amongst many others. There are even movies out there which have been completely created by a computer and 3D artists!

Will we see an age where actors will lose their jobs to the 3D artists whom can create models who are always in perfect shape, don’t charge any fees, don’t argue about some scene they don’t want to play and above all: are simply perfect for their roles (limited to the talent and imagination of the 3D artist, of course)?

It seems that way (like depicted in the movie “S1m0ne”) and sooner or later, even voice casting will be unnessecary as it looks like pretty soon, we’re not only going to be able to make a computer say whatever text we give it, but we will be able to simulate emtions through an artificially created voice!

Looks like we’re going to have a great oppurtunity whilst Tom Cruise and his fellows’ll be out of a job.

What are your thoughts on this?
Are you pro? Contra?
What would the consequences be, you think?

Give me your thoughts, I’d like to hear them!

im no actor, so i wouldn’t mind.

yeah, i would never make a 3d tom cruise.

and it still costs a lot to make 3d movies. that hopefully will go down as programs get better.

but that would be cool. of course movies aren’t the only things actors and celebrities are known for now. tom cruise again.

those days are definately coming. but i think audiences won’t give up real-actor movies. but they will all have special effects and maybe some seamless integration with 3d actors.

A lot of 3d forum visitors have most likely seen this.

My comment: How many pros could actually do something like this? Most prefer to do generic faces which are presumably easier to model. It will probably be more inspiring to see a lot of well-known faces in the WIP section, instead of the usual generic ones, monkeys, monsters, uglies, etc. I wonder how many could subd model a famous female celebrity, for example.

This is pretty amazing



Hmmm, Stahlberg’s work’s looking very plastic these days. That animation was ok but I’ve seen better. It seemed a bit jittery.

I think 3D replacements for characters is a technical possibility. All they have to do is the same thing as King Kong with a human character model. I don’t think they will replace actors though because people like meeting the stars of the show. What would happen is the Lara Croft effect where studios hire look-alikes who often don’t really look anything like the virtual models and people are disappointed.

Real actors also have personalities (not all of them mind you) that they project into their roles across different movies. That’s where some of the appeal to see a movie comes in and why they advertise the cast names with the movies. It would be quite difficult to build up that kind of personality into a virtual model without some serious AI.

Are you saying that the skin in that animation looked ‘plasticy’ to you ? :eek:

It looks perfect to me.

I don’t see anything jittery about the animation, I’ve watched it many times, at first I thought it was a hoax, that he’d just filmed a real person.

If you have any links to realistic human animations, I’d like to see them.

This got a lot of responses on CGTALK, though it’s for a real time game engine, and therefore apparently limited in how realistic it could be.


Not going to happen. Not until computers are sentient. They can’t act. Any jobs lost by live action actors in favor of CG will be gained by a different kind of actors: animators. There’s always someone controlling the performance, one way or the other. Which is to say, there’ll always be acting jobs. (And, as it is now, in CG movies it’s usually separate people doing the voice and the animating. More jobs for everybody.)

But that’s still assuming the worst case scenario that CG completely replaces live action, which it never will. Ever. Live action will always be cheaper and more realistic. Which takes less time and effort, even assuming you have incredibly powerful computers: constructing a fake human and surrounding environment from scratch, or pointing a camera at someone and turning it on?

One thing that I’ve learned from the Blender and Linux experience is that there’s always room for improvement. 3d software features will improve as more developers get into the game. Photoreal characters, photoreal environments and rapid 3d production methods are becoming the norm. I could see a day when smaller based 3d designers/content developers produce media that formerly was only produced by large cg film and game design houses. They will probably do a better job at cg production and bring more advanced features into the process.

Along the lines of 3d content creation I was thinking that more of us could focus on creating better 3d software tools via coded applications. In the past it was not so easy to learn software development. Nowadays it’s just a matter of taking out time and picking a programming language. Artist really don’t have to wait for developers to create new features that are needed. Computer artist can code a feature or two.

If anything, the SOC code projects may inspire more cg folks to learn a bit about computer science. You may not go as far as creating a full fledged 3d application but you may just create cool python plugin, the next great plugin to a hot new render engine or modeling application. In any case who says that we have to wait for any of this tech to show up? If we try hard enough maybe we could make it happen.

Anything is possible.

Blend on!

I remember, someone not from the CG scene recently stated, that movie is going away from the traditional acting and to CG low budget (up to one mio dollars) filmes and the more the CG tools are improved, the more I think he was right. It is impressive, what can be done with CG tools even now, so there may be some surprises in future.

However, real acting will not be replaced this fast, since there is more about an actor than just his films. This may be delivered through CG too, since the actors are lending their voice to the 3D characters and a special touch of personality with it, but this isn’t the same at all. The only real loosers are the stunt man, because their jobs will be replaced by render farms and 3D artists in near future for sure (lets say in about 10 to 15 years), because doing stunts in 3D is much easier, saver and cheaper than really doing them. Think of the fees to blow something of, the explosives, the coordinators, someone to clean up the mess, you are leaving and the stuff, that has to be blown, not to mention the stuntman, who should fly away, grinning into the camera at the end (being replaced at this moment). If anything goes wrong with a stunt, you will have to do it again, leaving you with even higher costs. CG artists and apps will be cheap, compared to that. Scanning somebodies 3D shape, adding hair and textures and then start to rig him, if he does not come with his own 3D model (which future actors may have and will bring with them for an extra fee :slight_smile: ).

Those stuntman may than study copyright law, connected to new bill, gouvernements will have to pass, to ensure nobodies face is used against his will (I think this very good Tom Cruise model is a good example, if someone builds a body to it and rigs it, it would be really believeable).

I was thinking that also. But then I thought about coding intelligent learned preprogrammed responses into character behavior. A database of facial behaviors could be activated by a sentence structure/situation that is funny, sad, angry, etc. Learned responses like a character avoiding hot or cold surfaces could be implemented into behavioral structures. I would imagine that functions like walking up stairs, opening doors, etc could also be learned and stored.

Say your character animation system takes in visual data from a Laural and Hardy clip. It adds in this data in some improvisational manner to it’s current learned response data. Then the 3d director/ producer scripts out a animation course. Along the course behavioral markers are placed to adjust the characters attitude at one point, their mood at another, etc. These markers could be entered as a formal programing language or as common language commands. Then the character would progress along the animation course reacting to each situation as preprogrammed.

The characters behavior may change at some point as the computer attempts to add in the recorded data of the Laural and Hardy clip. Say at one point along animation course the door is locked. The character attempts to open the door normally. Next the character wiggles the door nob. Since the marker for this point in “anger/sadness/delayed” the character is not angry after the first few attempts. Finally the character puts one had in their mouth while it’s facial expression turns to a frown. The character lightly beats on the door with slumped shoulders.

The door opens, the marker happy/ dance was activated. The characters expression is now joyful and it begins to dance. As the computer recorded the data from the Laural and Hardy clip it recognized Laurel’s visual behavior and matched his facial expressions/etc to it’s own data. Then later the computer activated some of Laurel’s behavioral expressions along with it’s own.

I think that we must think of a computer more along the lines of robotics. We may be able to develop advanced applications faster that hoping for human-like or even bug-like intelligence.

Now maybe a 10, 70, 1,000 whateverHZ cpu won’t even cut it for this type of system. Maybe that’s the greatest issue here.

I think that we are nearing the “minds on” era of computing and the “hands on” era is passing away. I remember years back when I complained about manual uv mapping on a cg forum. Some cg artist thought that manual uv mapping was perfect at the time. I wonder what these same artist say now that everyone has access to LSCM mapping?

Less muscle more brain I always say in regards to anything computer related. Let your mind do the walking.

Say a client wants a painted “looking” image of a woman sitting on a bench under a willow tree. You (a) don’t have a willow tree (b) you can’t hire a model in the spot and an artist to paint her © you need it done like yesterday.

There are many methods that would allow you to get this job done in a days time using a computer. Most of these work processes would involve your mind.

Oh yeah I see a great need for better cg actors for creating visual content of one kind or another on the spot.

Just think, we will eventually only use polygon modeling to setup a mesh for sculpting.

Hey! Stahlberg was putting out great character models when most of us were fooling around playing games like Duke or Quake on a PC. He inspired many of us to get serious about using a comupter and start modeling characters.

there was photoshop in there, meaning animating that would not be so simple.

I think actors will have their jobs for a long time.

On the most base of levels, imagine.

Would people enjoy porn as much if they knew it was just the animations of a 45 year old computer artist?

Porn wisdom applies to many things.

Actually forget about a digital Tom Cruise. If I had this marvel of a digital character creation sitting on a holoscreen front of me, hehehe, I just might revive the mothership. Hehe, the funk mothership that is , funky beats, cg funk groups and cg disco dancers. Throw in some multiple camera views, dance move selections, tune selections, etc. We’re talkin’ major throwback saturday night jamz. I would have to hire some soulful singers, dj’s, etc.

Maybe future content creators will go beyond the “movie” entertainment model a bit with this tech.

C’mon, it doesn’t look anything like a real person. I wasn’t just meaning the video though, a lot of his gallery images look quite plastic too. Not that they are not good but some images would look much better with more realism.

There aren’t many. Here is one but it’s not very long:


There was another one I remember a while back that had quite realistic expressions but I can’t remember where it was from.

You’re right and that’s why I expect better of him because he’s capable of it. You can’t just live off your reputation.

I’d love to see if they could place a CG version of a mainstream actor in a porn movie. Imagine Angelina Jolie (although she might be up for it anyway), Katie Holmes and Jessica Alba.

Who would David Letterman interview when the movie is due for release? The animator? Director? Or would we have a digitial interview too?

The Artist known as…

How can we be sure David Letterman is real at that time? We are talking something like the year 2055…

I could see folks saying, “That Davy, he never seems to look a day older. Well the shows over time to catch some owls.” Photonic owls fly around the room while you tire yourself out in an effort to catch some. Before you know it your fast asleep.

This place would be doubly interesting, imo, if we see some real modeling challenges such as likeness of real people or famous personalities. Not just the usual generic faces. Although I appreciate the effort, they no longer seem impressive. There are no rules and references involved, just outright do whatever comes to mind thing. Thus the actual lack of challenge.

Oh, btw, there is now an ongoing competition at SubdivisionModeling.Com with prize. The thread is REALISTIC CELEBRITY MODELING CONTEST.

It was triggered by this thread, I believe.

Hope we’ll have our own here someday.

I would personally like to never see Tom Cruise in any form of media ever again. He is nuts.

I agree completely. Tom Cruise is a perfect example of an actor
that has been “overused” in too many movies. Imho - he’s really
not that great, and as much as I love Steven Spielberg (Still my absolute favorite director of ALL TIMES) - I boycotted the
movie “War of the Worlds” purely because Tom Cruise was in it.