E3 Trailers and the future of CG

Become a member of the CGSociety

Connect, Share, and Learn with our Large Growing CG Art Community. It's Free!

THREAD CLOSED
 
Thread Tools Search this Thread Display Modes
Old 06 June 2013   #1
E3 Trailers and the future of CG

One of the things I am taking away from all the E3 trailers, particularly those for the PS4 titles coming up, is that anything is possible, maybe even actors, and it can, perhaps, even be done in realtime. This leads me to an interesting thought about the future of CG in movies. As old school directors retire and a new crop begins to invade Hollywood, I am wondering if we will see anything even remotely difficult ever done for real again. Looking at the E3 trailer for The Crew, for example, will anyone be able to justify doing dangerous and expensive car stunts for real?

If a movie like Iron Man were made a few years earlier, there probably would have been a lot more time spent with an actor in an actual suit. With each film, though, it seems there is lss and less practical Iron Man, even when he is just standing there talking.

Think about it. What if you never had to bother with permits to close off streets? What if you never had to bother with building or acquiring a seaworthy boat? What if you never had to physically travel to the Amazon, or Africa, or deal with shooting dangerous animals?

Will there be any reason to have a real guy do a twenty story high fall? When there are no stuntmen left who can do amazing horse stunts (I heard they were heard to find even in the most recent Zorro movie) will we even care?

In the movie Angels and Demons, they were not allowed to shoot in some locations because of the movie's subject matter, so they recreated the places entirely in CG. In the future, will they even bother asking for permission?

Of course, all of this amazing stuff we are seeing was incredibly expensive and took a long time to create, but it won't always be that way. At some point the line will be crossed, where it is faster, cheaper and easier than doing tit for real. I remember Stephen Spielberg once commented that you couldn't do Lawrence of Arabia today because it would end up costing a half billion dollars. What happens when that line is finally crossed?
__________________
Terrence Walker
Studio ArtFX
Learn How to Make Your Own Animated Projects!
You don't need millions of dollars or major studio backing!!
 
Old 06 June 2013   #2
Your post is like a message from the past, ca. 2005, and we are already there. Everything you list is already pretty common. When even rom-coms and dopey tv comedies use tons of greenscreen, I think it's safe to say we have reached that point.
__________________
www.artbot.com

 
Old 06 June 2013   #3
Actually, I thought the Iron Man movies have been using more practical suits over the years than before.

I think in some cases it will still be easier to do practical effects than CG, there's quite a lot that's not really reasonable to do with CG. Usually environmental stuff that requires a large number of unique things. CG can make many copies super easy, but making many variations not so easy.
__________________
The Z-Axis
 
Old 06 June 2013   #4
But it's not always a matter of making something that would be too expensive to build as a practical set. Many times what is gained is convenience, and that's a HUGE thing in any production. Not having to close down streets and get permits, or have principal actors on location, saves tons of money. Or when a director is missing some coverage, they can get the actor for a half day in a controlled environment, pop in a bg plate and they have their reverse angle (or whatever) that would have cost tens of thousands to take everyone back to the location, match lighting and make-up, hope the weather is the same, etc.

Now that background plates can be easily added/edited/altered, they only have to send a small 2nd unit crew to shoot them. And of course, there's the world-wide outsourcing of 3D assets, which are relatively straightforward to build. As has been said here many times before, VFX are far more prevalent than most people know, even those educated in VFX. As the author pointed out in that "Victor" article on FXGuide, even the new Arrested Development had 850 VFX shots in it! Sure, a lot were probably paint-outs and clean-ups and whatnot, but that's still amazing. So, yeah, we are already there.
__________________
www.artbot.com

 
Old 06 June 2013   #5
Originally Posted by Artbot: Your post is like a message from the past, ca. 2005, and we are already there. Everything you list is already pretty common. When even rom-coms and dopey tv comedies use tons of greenscreen, I think it's safe to say we have reached that point.

Speaking of 2005, check out the new Stronghold Crusader 2 trailer.

http://www.youtube.com/watch?v=ATAeKRm87h8
__________________
 
Old 06 June 2013   #6
What we need to have is an open library access to unified basic assets. Incorporate those with a game engine. Make things procedural through open source or paid-access. It will make production costs more affordable, and in turn more demand for CG productions. After this many years of CG and countless assets being built, we still see there are studios and artists creating characters / vehicles /animations from scratch - things that are so generic I think is energy not well-spent.

Imagine needing to build a model of middle-age male warrior, or an animation of casual business-man like walking and turning 90' after taking 3 steps? Just adjust the sliders. Do the performances even with real-time, then bake and modify to the production needs.

I am not talking about one software that incorporates 100's of assets, but these assets should come from a whole community. Let's be honest here, there's really no secret to how edge loops should be built, how walk cycles are keyed, how a guy should fall, or how brick houses should be textured.

Create, share, and integrate. Pave the way for the artists 50% there, and let the artists focus on the rest that require more of their input.

Just my 2 cents!
__________________
CGMovement:
http://CGMovement.com

Last edited by shakii : 06 June 2013 at 01:43 AM.
 
Old 06 June 2013   #7
Originally Posted by teruchan: .....Think about it. What if you never had to bother with permits to close off streets?.....


That's a good point. I remember watching "The Devils Own" (Harrison Ford and Brad Pitt) on DVD. My Dad told me the opening shootout scene (which lasts just a few minutes) was shot in Inchicore in Dublin, only down the road from where we live. He said the street was blocked off for shooting for several hours a day, for two or three months! All the shops and businesses had to be compensated for lost business, traffic including buses diverted (causing inconvenience and delays to commuters) Gardai brought in to keep an eye on things. I'm sure all that had to cost a lot in terms of money, logistics and time organizing.
Multiply that by the rest of the outdoor scenes shot in other locations, and it gives an idea of how budgets get swallowed up.
__________________
I like to learn.
 
Old 06 June 2013   #8
Originally Posted by shakii: What we need to have is an open library access to unified basic assets. Incorporate those with a game engine. Make things procedural through open source or paid-access. It will make production costs more affordable, and in turn more demand for CG productions. After this many years of CG and countless assets being built, we still see there are studios and artists creating characters / vehicles /animations from scratch - things that are so generic I think is energy not well-spent.

Imagine needing to build a model of middle-age male warrior, or an animation of casual business-man like walking and turning 90' after taking 3 steps? Just adjust the sliders. Do the performances even with real-time, then bake and modify to the production needs.

I am not talking about one software that incorporates 100's of assets, but these assets should come from a whole community. Let's be honest here, there's really no secret to how edge loops should be built, how walk cycles are keyed, how a guy should fall, or how brick houses should be textured.

Create, share, and integrate. Pave the way for the artists 50% there, and let the artists focus on the rest that require more of their input.

Just my 2 cents!


This is exactly what I was thinking when watching the debut trailer for The Crew. Very awesome stuff and totally realistic, but it's all car parts! Very real and natural looking, but tires, engine parts, transmissions and a warehouse. I wouldn't want to model that stuff. It should all exist somewhere already, and the artists can get on with making the scenes.

Even cooler would be if the real models used to make the physical parts were available. Even if they were expensive, studios would pay it. Most of the stuff being done, as cool as it is, is just generic reality stuff. There should already be a huge library of that stuff. Imagine how much more great entertainment content would be created in that case. Even smaller independent movies, or TV shows, might have larger scope than is currently possible. A TV show could never shoot the amazing car action from a trailer like The Crew, but maybe if such libraries existed, even a weekly TV show could deliver that level of action.
__________________
Terrence Walker
Studio ArtFX
Learn How to Make Your Own Animated Projects!
You don't need millions of dollars or major studio backing!!
 
Old 06 June 2013   #9
Originally Posted by shakii: What we need to have is an open library access to unified basic assets. Incorporate those with a game engine. Make things procedural through open source or paid-access. It will make production costs more affordable, and in turn more demand for CG productions. After this many years of CG and countless assets being built, we still see there are studios and artists creating characters / vehicles /animations from scratch - things that are so generic I think is energy not well-spent.

Imagine needing to build a model of middle-age male warrior, or an animation of casual business-man like walking and turning 90' after taking 3 steps? Just adjust the sliders. Do the performances even with real-time, then bake and modify to the production needs.

I am not talking about one software that incorporates 100's of assets, but these assets should come from a whole community. Let's be honest here, there's really no secret to how edge loops should be built, how walk cycles are keyed, how a guy should fall, or how brick houses should be textured.

Create, share, and integrate. Pave the way for the artists 50% there, and let the artists focus on the rest that require more of their input.

Just my 2 cents!


There are already 1000's of free models out there, and even more readily available models if you are willing to put some money down.

The big issue with this is that each model generally has to be made for a specific engine limitation, and those are pretty much customised per game.

A car for an RTS game is completely different to a model for a FPS. Crysis engine models are different from models used in the Unreal engine. Things like the ability to have detachable doors, engines in the model, glass, level of detail scalability...

The models in game engines aren't generic by a long shot.

The other issue that (and this would count for both games and other fields) is that the library assets that a company has is a competitive advantage. At my studio, we have highly detailed models of 4 major cities in South Africa (Sandton, Pretoria, Rosebank, and Cape Town). These are models that we have developed internally, and give us a HUGE advantage in being able to do work in these cities.
I would imagine that ILM reuse a lot of their New York models in their production. Giving away an asset like that, to make it easier for other companies to do the same thing doesn't make sense...
 
Old 06 June 2013   #10
D' oh!
I thought this was gonna be a Best of E3 thread...

I'm just as giddy as a schoolgirl over the fact that more and more Hollywood actors and talent are looking at videogames as a feasible venue.
__________________
My Deviations
 
Old 06 June 2013   #11
Originally Posted by Pyke: There are already 1000's of free models out there, and even more readily available models if you are willing to put some money down.

The big issue with this is that each model generally has to be made for a specific engine limitation, and those are pretty much customised per game.

A car for an RTS game is completely different to a model for a FPS. Crysis engine models are different from models used in the Unreal engine. Things like the ability to have detachable doors, engines in the model, glass, level of detail scalability...

The models in game engines aren't generic by a long shot.

The other issue that (and this would count for both games and other fields) is that the library assets that a company has is a competitive advantage. At my studio, we have highly detailed models of 4 major cities in South Africa (Sandton, Pretoria, Rosebank, and Cape Town). These are models that we have developed internally, and give us a HUGE advantage in being able to do work in these cities.
I would imagine that ILM reuse a lot of their New York models in their production. Giving away an asset like that, to make it easier for other companies to do the same thing doesn't make sense...

Not to mention QA of such assets.
-What if such assets have bugs-and say only with maya v2012 (but not other versions).
-and if they are so 'useable' a director could bypass the VFX studios and manipulate tha assets himself plug and play.
How do VFX studios compete with the easy-peasy ideal (which fortunately doesn't exist)?

As to the original post-as good as the Cinematic is there is something a bit 'unrealistic' and 'gamey' for me when a Ford F150, a police car and a Lamborghini all drive the *same*.

Last edited by circusboy : 06 June 2013 at 03:06 PM.
 
Old 06 June 2013   #12
Originally Posted by teruchan: I remember Stephen Spielberg once commented that you couldn't do Lawrence of Arabia today because it would end up costing a half billion dollars.


Speaking of SS
Spielberg foresees 'implosion' of film industry

Original news link

Last edited by kelgy : 06 June 2013 at 05:00 PM.
 
Old 06 June 2013   #13
Originally Posted by Pyke: There are already 1000's of free models out there, and even more readily available models if you are willing to put some money down.

The big issue with this is that each model generally has to be made for a specific engine limitation, and those are pretty much customised per game.

A car for an RTS game is completely different to a model for a FPS. Crysis engine models are different from models used in the Unreal engine. Things like the ability to have detachable doors, engines in the model, glass, level of detail scalability...

The models in game engines aren't generic by a long shot.

The other issue that (and this would count for both games and other fields) is that the library assets that a company has is a competitive advantage. At my studio, we have highly detailed models of 4 major cities in South Africa (Sandton, Pretoria, Rosebank, and Cape Town). These are models that we have developed internally, and give us a HUGE advantage in being able to do work in these cities.
I would imagine that ILM reuse a lot of their New York models in their production. Giving away an asset like that, to make it easier for other companies to do the same thing doesn't make sense...


I agree. There are big challenges to this, and requires a lot of effort to get people on board, and to systematize and integrate. While all these feels much like a community work rather than for profit. Another concern is some may view it as a threat to their jobs. It is definitely easier said than done, and perhaps takes a main force to push behind this, ie., Google and Google Maps or many of their pther operations.

Who can profit from doing such work for the CG community? I think maybe not for them a few years ago, but Autodesk is pretty much in a good position to be that one.

But let's come back to main reasons why this can work

1. Why should companies share their competitive advantage secrets?
They shouldn't. We're talking about the generic or one-off items, and things that many others have built countless other versions of. But, if a company decides to share its IP's, ie., main characters of a movie - there are benefits in gaining more publicity, as fan-made viral videos have always been one main area marketing departments have their eyes on. Maybe not the full version of the assets, but the down-res'ed ones. This can also serve as study purposes for others.

2. Shared assets = don't fit the software/pipeline?
Two solutions, when sharing, build from templates. When integrating, use re-targeting. Ie., No matter how your software requires a car to be modeled/textured, you can't change the look of the car. A student can follow a car blueprint just as how a industry vet would. Let's take that existing car model and then fit that into your own workflow. Consider our current workflow in modeling this car from scratch - there are probably 80% of the work done that are non-artistic and maybe repetitive - most importantly, they follow a formula. You've mentioned Crysis and Unreal, or FPS or RTS all have different ways to model. If we can name why they are different, then we can take that formula to automate solving the "why doesn't this asset work for us yet". (Things like how edges have to be booleaned, how edgeloops have to be distributed, how some faces have to be deleted, etc.) There are maybe restrictions right now with our tech to re-target (I'm no modeler), but doesn't mean it's not doable in the future!

As for re-targeting animations, that's probably even easier done with exporting the bone information without the need of sharing the rig. The technology is already there ie., Autodesk's HumanIK is across major platforms 3DsMax/Maya/MotionBuilder. I think the important thing is to identify the key animation actions and categorize with formula. Ie., A man sitting down. Identify the contact points, then we can make it procedural with the height of the chair, where he places his hand on the table, and how many steps he takes to the chair, etc. Take one step further, heck, even "how something is cartoony" may be able to be formulated. There are tons of keyframes out there that are being created in the same way (ie., double bounce, overshooting, anticipation, snappy) over and over and over again.

3. Shared assets have no QA?
For any crowd-sourcing QA, it comes from the community with volunteer moderators, editors, or small group of paid staff, and user ratings. If an asset is not good enough, members can edit and update it with improvements. For enticing the community involvement, there can be reward systems.

4. A(nother) crazy idea - large professional coalition of CG artists. Much like a worker's union, contracted by multiple studios, and we assign positions internally. Except on top of that, we have the shared assets within this coalition. Making our work more efficient/affordable with access to a larger array of quality and unified assets than most studios. This also may give more power to the artists in negotiating for our rights.

Anyway... have to go back to work, sorry if this may sound silly at some points. Thanks for reading and your responses
__________________
CGMovement:
http://CGMovement.com

Last edited by shakii : 06 June 2013 at 06:56 PM.
 
Old 06 June 2013   #14
Originally Posted by kelgy: Speaking of SS
Spielberg foresees 'implosion' of film industry

Original news link


Tom Hanks said the exact same thing on the Nerdist podcast a few months ago. It's a pretty scary state of affairs when even these people cannot get content made or distributed. The "Broadway model" is a fascinating development (though not decidedly good or bad).

I don't think the movie or VFX businesses will even be recognizable 10 years from now. But if there's anything to be extrapolated from current corporations' wet dream of cloud subscriptions for everything, it's that we are likely on our way toward this as a model for consumption of all media: movies, games, tv shows, everything. Think about it - it's a good deal for people who consume a lot of content, and a forced (but necessary) deal for those who don't. As soon as corps get their licensing shit together (something most have failed to do so far - just look at Amazon Prime's pathetic movie selections) then they can just turn everything into a minimum monthly fee, they will. Mark my words.
__________________
www.artbot.com

 
Old 06 June 2013   #15
Originally Posted by shakii: ....and requires a lot of effort to get people on board, and to systematize and integrate. While all these feels much like a community work rather than for profit.


Great - another person asking CG artists to work for free. Yeah, that's a great idea.
__________________
www.artbot.com

 
Thread Closed share thread



Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
CGSociety
Society of Digital Artists
www.cgsociety.org

Powered by vBulletin
Copyright 2000 - 2006,
Jelsoft Enterprises Ltd.
Minimize Ads
Forum Jump
Miscellaneous

All times are GMT. The time now is 04:40 AM.


Powered by vBulletin
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.