Slashdot Mirror
Report From The Land of SFX
prostoalex writes "MIT's Technology Review takes a look at the world of digital special effects, the industry worth half a billion dollars per year, according to the authors. It talks about the role of SFX in movie production nowadays and comes to the connclusion that while might not 100% computer-created in the future, we'll see more of realistic-looking special effects in future titles."
It talks about the role of SFX in movie production nowadays and comes to the connclusion that while might not 100% computer-created in the future, we'll see more of realistic-looking special effects in future titles.
More realistic effects in the future, huh? Boy, they're really going out on a limb this time.
GMD
watch this
So we can spend more and more on special effects, and less and less on those useless "plots" and "storylines"...
Common sense is what tells you the world is flat.
I have actually thought about this. It seems to me that in some movies, the plot and dialogue are actually a drag on the movie, which the action and effects are good. Thus, I think a "pure" action movie, containing little or no dialogue and only implicit plot could be great. It could be a legitimate artform of its own, untainted by cliched plots and tired one-liners. Cinematography, production design, and effects could be managed brilliantly by some directors who can't do anything with their hack actors and who themselves have no taste in scripting. Who's to say this is not legimate art or entertainment?
It's pretty close to a video game I don't have to actually play (who me, lazy?), but think about the way Half-Life worked in terms of making movies, or if you want to go back further, the old game "Another World" (aka "Out of this World". Now there's brilliance...
From the article:
But in part because digital projection does not create as unmistakable an improvement in the viewing experience as, say, the talkies did over silent films, theater chains are unwilling to foot the bill for the new projectors, which cost at least $100,000 per screen and might have to be upgraded every few years. Conventional film projectors, which last 20 years on average, cost $30,000.
I have to agree with this statement. I managed to see Attack of the Clones, projected in all digital in Phoenix, AZ. When the screen switched from analog (during the previews) to digital I noticed a difference, but not huge. After watching for a few minutes, I forgot that I was even watching it in digital. It's nothing like the difference between VCR and DVD.
Have other people who have seen movies projected in all digital had the same experience?
Forget the whales - save the babies.
Here's how I would have interpreted the post, though I'd prefer to just rewrite the whole thing and thank the original poster for providing the link:
(too bad strikeout is not a slashdot-approved HTML tag)
Thus, I think a "pure" action movie, containing little or no dialogue and only implicit plot could be great.
Hmmm. Interesting viewpoint. I don't agree with it but it's interesting. Let's draw an analogy to sex. Sex with someone you love is awesome. Sex with a complete stranger you couldn't care less about is an empty experience. But, as Dave Barry says, as far as empty experiences go, it's one of the best.
GMD
watch this
"We call them 'invisible effects,'"
And that is exactly where they should stay (barring of course the Pixar style total 3d animation stuff).
Why is this? Look at Star Wars EP2 for the obvious answer. Even with Lucas Arts churning out some of the best 3d in the movie business, there are still some amazingly obvious 3d effects (Anakin getting on and riding the creature... where Anakin becomes a 3d model getting tossed like a rag doll). And this obviousness seriously hinders the overall movie experience. Anything that pushes the viewer from the imaginary world back to realising he/she is in the theater is a bad thing, and in this case, I found myself thinking, "That doesn't look real," and bang i'm back in the theater.
So what am I say? Well, 3d should stay in the background. Until they can make it look exactly like the actor and have it move exactly like the actor, they shouldn't put it in the center of the action in a film made with real people. It is far too obvious. Though sometimes this is done pretty well.. such as Spiderman, the costumed one. The part with him in his hooded sweatshirt is also pretty obviously 3d.
But all these amazing secondary effects are just incredible, such as the backgrounds in Cast Away. Sometimes you have no idea, and that is the point of the game. Not that you can look and say, damn Lucas Arts has just made some cool 3d stuff, but to do a double take 2 months down the road when someone tells you it actually was 3d in the first place.
So basically, in the world of realistic 3d... the less recognition they get, the better they did their job.
Realistic.
The best SFX aren't the big ones, they are the little ones that touch things up and you never even suspect.
The story will always be paramount to a good film, but judicious use of CG can lend that extra touch that pushes the film over the edge of greatness.
Learning HOW to think is more important than learning WHAT to think.
Watching the Behind The Scenes featurettes isn't just the same anymore. I remember watching the making of Star Wars, how they'd string up the space ship models and rig them with firecrackers. Now its like "this is the computer where we do everything... it starts out in a wireframe like this and then we map on the textures." Whoopdeedoo. Its just not the same.
When film gets scanned into an AVID system, what resolution is that? I'm assuming its as high as possible since after its transferred to film again much will be lost.
Every time an announcement is made about movie SFX, the arthouse clique shows up with alot of high and mighty talk about how the important elements are plot and quality acting, etc etc. I guess this makes them feel like they have more culture than the average man, i dunno.. IANAP (psychologist)
/.ers are fans of the original Star Wars.. But take a moment to step back and ask yourself what made that movie a phenomenon..
The same thing happens in video games. There are those who constantly say 'i dont care how good Doom 3 looks, its still the same crap, i want better gameplay, thats what counts'..
Though these points are valid, you cant ignore the reality that eye candy and the Wow factor sell entertainment. People always want to see the next level.
Now, its safe to bet most
Was it the story? Hardly.. Simple, archetypal boy-rescues-girl plot thats been repeated since the dawn of time.. Was it the acting? That great moment when Luke lands back at base after destroying the death star, Leia goes 'Luke!', he turns to her and exclaims 'Carrie!'. Or the whining 'But Uncle Owen, I wanted to look at power converters'
Naw, what made that movie was the effects. Noone had seen anything like it before. Dont discount the audio ground lucas broke with THX, either..
Jurassic Park was another.. Dinosaurs are loose, we gotta escape. No plot, nothing to think about here. But those cool looking dinosaurs brought me in.
Most people just want to sit back, turn their minds off, and be impressed.. Always have, always will. Snooty intellectual affairs will always be the exception, never the rule.
If this wasn't true we'd still be happy with black and white film and our Commodore 64s
Now get out there and blow sum stuff up for me
I don't need no instructions to know how to rock!!!!
While Episode II was pretty much considered a bad movie, it did accomplish some stuff that is seriously cool. There was a LOT of footage in that movie where the background was a miniature standing about 18" tall. That surprised the hell out of me when I picked up the latest issue of Cinefex! They actually built this tiny miniature of a room, placed a robotically controlled camera in it, and composited blue-screened actors over it. We're not talking static background either, the camera moved through the set quite casually.
This technique isn't new, it dates back to the early 80s. (Greatest American Hero, for example..) AOTC did a wonderful job of pulling that off convincingly.
I really can't wait to see this type of advance winding it's way down to independent moviemakers. I'm really curious what happens when somebody uninhibited by mass-market considerations is able to get their imagination on screen.
"Derp de derp."
Digital effects are like markup tags. When you first find out how to use them they're really neat but over use can detract from the real content.
"I am sorry, but computer graphics do not make realistic images... Most of the graphics in AOTC were not realistic."
Wrong. Computers can do very convincing (realistic is not the right word until we get a video camera over to Coruscant) imagery that is imperceptible from real footage. If the effect is missing an element that is preventing it from being realistic, that is not the computer's fault, it is the artist that failed to produce the right elements.
The problem with AOTC's effects is that they were rushed. The movie was so overloaded with effects that several of the artists basically were forced to say "Well, this is the best I can do for now, time move on."
Computer + Good/Experienced Artist + The Right Tools - Appropriate amount of Time = Unconvincing.
"The best SFX aren't the big ones, they are the little ones that touch things up and you never even suspect."
There's the money saving ones as well. A good deal of the sets in Episode II were actually minatures filmed with blue-screened actors and robotic cameras. If this technology could bubble down to television, imagine how cheap it'd be to build new sets!
On the other hand, Rosen doubts that artists or audiences will soon want to give up the unique sensory qualities of film. "If we look decades ahead, people will come to realize that digital [photography] is another way of doing things, but film will give you a different organic look," he says. "It's like oil paint and acrylic. Digital has a different texture."
I disagree with this argument. If digital videocameras and especially the projectors continue to improve it is only a matter of time before audiences prefer digital to film. The current bottleneck is the Texas Instuments projectors which are limited to 1024 or 1280 lines. Lucas filmed Episode 2 at a higher res than that but is limited by the projectors.
While film could 'fight back' by going to 70mm or using the excellent Maxivision 48 system, I think it is a loosing battle because projectors will still have jitter, and prints will continue to wear and fade.
Larger film requires more storage space as does digital. Luckily for digital, storage capabilities continue to improve as hard drives cram ever more data into their platters.
In the long view costs will come down for digital, which is another current drawback. This is assuming digital becomes popular enough for economies of scale and competition to kick in. A complicated chicken and the egg situation indeed.
Exactly. Things you never suspected were effects, are often effects these days. In AOTC, there's a scene where Padme, Anakin, and others are talking to the new queen of Naboo. Sio Bibble (Oliver Ford Davies) was sick on the day of shooting, so they filmed his empty chair (he has one or two lines of dialogue) that day, and then later filmed him separately and composited him into the scene. I've seen the movie four times and it never occurred to me that he wasn't actually on-set!
But AOTC is a bad example; you read the above, and think, "Well, yeah, it's an effects movie, it's not surprising that there's stuff LIKE that, even if I didn't notice that particular effect". But even "non-effects" movies like Insomnia, American Beauty, A Beautiful Mind, etc. all have numerous effects shots. Making people or trees move differently in the background; continuity errors (like a modernly-dressed person wandering through the background of a shot) can be easily removed; The color of the paint on a vehicle, building, or wall can be changed; and so on.
If you want to get really pedantic, most movies are really just large, elaborate special effects. You see a guy walking down a street. The camera cuts to another view of him walking down the same street. Looks like it's continuous, right? But those shots were probably taken several minutes apart, the guy repeating his path for each shot. That is, technically, a special effect -- it's just not the kind of flashy hocus-pocus we usually think about.
"Destroy science and religion. Science would re-emerge exactly the same; but not religion." - Penn Jillette, paraphrased
Well how do you know they were rushed. They had about 18 months of production of VFX which is longer than the average. An artists are not overloaded and forced to say, too bad tat's it. It's the responsability of the supervisor and usually the directior has a say when he thinks the FX is good enough. Sometimes directors change things at the last minute and with the studios inflexibility with release dates, that might happen. Or when new FX are added at the last minute. But in general it's up to supervisors to make sure VFX shots are to a level where they aer OK. There is a cliche in VFX where people say that VFX are never finished you just stop working on them.
Effects houses have been simulating mass and inertia for many years now. If Jar Jar looked crap it was because the animators chose to ignore physics for that particular animation.
-- SIGFPE
I would take with a grain of salt that assumption. Real time CG has really taken off and some of the stuff shown at places like SIGGRAPH. But CG for VFX have other requirements and difficulties. Don't you think that the smart people inside these houses are always looking for the edge for every bit of performance and power? Even now a days we can't create in computers real time graphics that match the complexity of the work in Jurasic Park or Toy Stiry. Just look for Tom Duff's comments about the hype machine from graphic card makers. We might one day get it but I won't hold my breath. If it ever gets to that point don't you think SFX studios won't be the first to jump in?
Besides even though VFX studios charge a lot they are operating on razor thin profit margins. Many barely survive and many have gone belly up vecause of that, like Boss Films or when Warner killed Warner Digital. Second lets make a comparision. I mean in some of these movies you pay a star up to 20 million U$ plus what the director and maybe other actors might get and half the movie nbudget is spent on just a few persons above the line. Compare these to say paying U$ 30 million for a big FX show (say like Pearl Harbor or Mummy 2) on which you have to pay for maybe a couple of hundred people's salaries between 6 to 12 months. If anything studios are getting their FX work dirt cheap. Most of the VFX studios expenditures is salaries not hardware or software. You need to pay for the best artistc and technical talent.
They were rushed. I know because many of my friends worked on it. And I know they are more talented than that!
-- SIGFPE
"If effect artists want to make things realistic, they are going to have to start simulating mass and enertia on the objects they model."
If you saw how Pearl Harbor was made, you would not have made that comment. The guy that replied to your post was absolutely correct.
There is an interesting back and forth between Roger Ebert and Peter Donen (a Visual Effects Supervisor) on the effects of The Bourne Identity.
It seems that during "Ebert & Roeper at the Movies", Roeper said that "no computer effects are as good as a well-shot movie in Paris" to which Ebert responded "this movie is a convincing argument for really photographing real things happening on real locations."
Then Doren responded (as he had worked on The Bourne Identity) saying that there were actually over 150 special effect shots "includ[ing] miniatures, blue screen, wire removal, time manipulation, 3D character animation and background replacement for starters."
His take on his job: "I come from the school that says if I do my job well, my work will not be noticed by the audience."
What is music when you despise all sound?
"Well how do you know they were rushed."
Because I'm an animator. I can tell the difference between a rushed animation and a completed animation. Watch the Dex alien in the diner and then watch C3PO's adventures in the battle droid factory. There's a huge difference in refinement. Heck, I don't think you even need to be an animator to spot that.
"They had about 18 months of production of VFX which is longer than the average."
They also had an absurd amount of FX to do for nearly every shot in the whole movie. They had more to do than Toy Story. If memory serves, Toy Story had somewhere between 2.5 to 3 years to complete. And they didn't even have to match it up to real people and places!
"An artists are not overloaded and forced to say, too bad tat's it."
Yes. They are. A badly completed effect is more valuable than no effect at all.
"But in general it's up to supervisors to make sure VFX shots are to a level where they aer OK."
That's really no different from what I said. The choice is the same. Whether it's the animator or the manager that makes the call doesn't affect my point one bit. If I thought anybody was going to be zealously literal about what I said, I would have said 'animation team' instead of 'animator'. The truth is that the manager/supervisor is not going to have any idea if the schedule's going to be blown unless the animator says "Sorry, I just don't have the time to finish this without eating into the time I need to do the next shot."
"There is a cliche in VFX where people say that VFX are never finished you just stop working on them."
That's not a cliche, it's a myth. The point of an effect is to describe an event. "phaser shot must make slug like creature burn up and vaporize." Once all the requirements are fulfilled "phaser shoots, slug like creature burns, vaporized", the effect it's finished. You can add more elements if you want, but you don't make the point any clearer. As a matter of fact, overworking an effect can ruin the elements that fulfilled the requirement. Anybody remember Return of the Jedi Special Edition? There was a huge party at the end of it on Coruscant. Lotsa people were cheering. After leaving the theaters, one of my friends said "Did you see the statue of the Emperor knocked down?" There was so much crap on the screen that I missed that detail. Whoever did the effect of the falling statue had their work ruined because somebody didn't finish when they should have.
Ohh I did read it, and actually I re-read your link in case I didn't remember something right. I'm not saying that it will never happen. I just have an issue with those saying that it'll happen in a few years. Besides Carmack was pretty bold, saying they are all wrong. Do you think someone like Tom Duff or Larry Gritz are plain worng and don't know their stuff. These are also pretty smart guys who work in CG for movies, with PhDs and actual practical knowledge of what it takes. The parent post implied that a graphics card would be able to do the work of a dedicated 1000 CPU render farm. Not in a couple of years. And besides CG and VFX studios will be doing even more complex stuff also, it's not like it's a stagnant industry, a quick look around SIGGRAPH would convince anyone.
But you had quotes here are some in response, which I put in the previous Slashdot article about real time graphics, plus someone elses:
Wjat does the GSCube do
Playstation 2 and Toy Story
Real-Time RenderMan?
Toy Story Graphics
So yes one day it'll probably be true but I don't think my next computer/video card would be able to do it. The hardware papers at SIGGRAOPH doesn't seem to imply that it's almost upon us that hardware will match VFX quality graphics, which is another field in CG.
Also he brings some points but there are a little bit off. The waves for the Perfetc Storm were simulated first, Origin 2000 I think. After the data was generated, then it was rendered. Still it took hours just to render one frame each. Carmack seemed to only concentrate on the simulation step, when actually you can think of it as a 2 step process. Then again at AWGUA, Bill Buxton showed the fluid effcts from inside Maya 4.5 and in a video about Jos Stam he had a fluid simulator running in a PDA. Granted it was coarse and simple but it was really impressive.
I'm all for more realistic games but I'll just wait for it.
"He's a mulatto at best! Isn't it ironic that he's the one pushing the "skin color means nothing" propoganda in his music. Guess he doesn't buy that bullshit either."
If memory serves, he suffers from a fairly common skin disorder called 'vitiligo'. Basically, it's a light splotchiness in the skin that gets bigger and bigger... He probably lightened his skin so you couldn't see the splotches.
Nose jobs and plastic surgery are all about vanity, but making fun of a guy for his skin disorder is not cool. Celebrity or not, he is a human being and can suffer from the same ailments that you or I could. He chose a unique way of dealing with his problem, but people still trivialize it. I'll tell you something: Whether he has the illness or not (there's debate about that), he certainly didn't have the technique done in order to contradict himself.
Maya represents a major, and underreported business achievement. SGI did something that companies always talk about doing and never bring off - got synergy from a merger. They bought Alias and Wavefront, and two years later, out came a product that combined the best of both. That deserves a business school study.
Meanwhile, Softimage got bought by Microsoft, then sold off to Avid, Softimage|XSI was years late, Avid bought Motion Factory and trashed it, and in the end, Softimage moved from #1 to about #4. (As a Softimage user and plug-in developer, I found this annoying. But that's another story.)
I wonder.... and this is my point. Let's take Mr. Carmack at his word (he is, after all, smarter then me :) and say that all this power is possible within a few years. Not to take the artist out of the equation at all, but... what IS going to happen when a home machine has that much power? I really don't think we're quite grasping the first part of that statement... the power to do Toy Story (or better) graphics in real time. Imagine that. "I want the huey here, the tank there on the rolling hillside, these three missles will fire, and the explosion will go here! Roll it!" No, it will never be as good for the reasons stated (custimizability, modelling (the huey came from where, exactly?), etc), but for the 75% of people it will be close. That makes it much less of a marketable commodity, IMHO. But then, maybe I am comparing Louve pieces with a Starving Artist's sale... time will tell.
This was the most relevant quote for me:
:). It's going to be a great looking video, with a compelling story that takes advantage of the unique qualities of the medium.
"There are certain skills necessary to accomplish the shooting, making and coming out on the other end with a motion picture," Poster says. "One is cinematography. We say, if you know how to light it doesn't matter what medium you're shooting on. Likewise, if you don't know how to light it doesn't matter which medium you're shooting in."
I just graduated from college with a stack of short films behind me, and I'm gearing up for my first feature. From a technical standpoint, yes, film is still much better than digital -- I'm sure people on this thread will mention the absurdly low resolution of today's HD video. But to go to film means tripling the budget, raising tens of thousands of more dollars. And that's for 16mm, not even the 35 that we know and love in the theater.
One of my friends says, "Don't bother with video, it looks like crap. Spend the money instead to make a 35mm short that will look really professional and then people will invest in a 35mm movie." And another one of my friends actually went and did it, getting into some pretty big film fests.
But I agree with the quote -- it's not how good the format is, it's how you use it. Take two recent digital movies, Tadpole and The Fast Runner. The first is lit like the filmmakers know it's a cheap format, and treat it like a cheap format -- everything is hastily lit and handheld. Certain passages look like a home movie my dad could have shot. In the latter, the format was treated with respect and carefully lit, getting as much out of the format as possible. And it looks fantastic -- I would have no complaints is my film looked like that.
And even beyond that, what good is a great-looking format if the story isn't worth the film stock it's shot on? (I won't name any titles here.) So no, I'm going to do a feature-length movie on video before I do a short in 35 (unless, of course, I can raise the money to do a feature on 35
When it comes to SFX, "digital" does not necessarily mean "better." The models of Star Trek: TNG, with light passing over the textured, solid models in unsimulatable ways, are much more realistic to me than a Voyager frame filled with two dozen wire meshes. (I'm using TV shows for examples because the budget constraints are tighter.) My eyes have started glazing over all fake looking FX, especially digital stuntmen in features. They pull me out of the story immediately. I stop seeing them as people, and I didn't pay $10 to care about someone's digital models. I want to be like Zemeckis -- do FX that you can do well, and make sure they they serve the STORY instead of being their own attractions.
I think we are pretty much on a common ground. I'd like to apologize: Your post here is very insightful and I found it to be a very interesting read. I developed an assumption about you that I realize now was in error. When I responded to you, I was not as nice as I should have been. I'm sorry man, I had the wrong idea about ya.
Cheers
P.S. How do you know the ppl at ILM? I'm curious because I aspire to work in Visual FX, but would like to network with people. Unfortunately, I don't know anybody in the business that can guide me in the right direction. Is there any help or suggestions you can offer there?
No need to apologize, I didn't find you rude or anything, except maybe passionate. As I said on re-reading my first post I think I didn't do a good job of expressing my ideas. It's not that I don't think Carmack will be right eventually (I just disagree on the timeline) or that many FX are rushed (I sure know several examples), I just couldn't find the right words :-).
As far as people from ILM I've met plenty at SIGGRAPH (attendee since 95), usually I've spoken with a few of them for a few min. I've got an acquaintance there but we only email evry so often since he is to busy. Heck I interviewed once there but I didn't get the job, oh well. Also I've inherited a page about ILM, The Unofficial ILM website:
The Unofficial ILM Website
Because of that I now have a relationship with ILM PR and they are sometimes very gracious in providing info and scoops. They even introduced me to a fwe people where I almost had a heart attack but that's beside the point.
My best suggestion is to get involved with SIGGRAPH (if you have a local chapter). If not, start saving and go to next year's SIGGRAPH, in San Diego. That's the most important event for VFX and the biggestrecruitment event for ILM and other houses. You can certainly makes losts of contacts ther and actually talk to all these people and they can give you great advice. First time I went my eyes opened, I really didn't know how much was FX all about. In case you haven't gone to one I have a primer on the conference:
SIGGRAPH 2003
SIGGRAPH Primer
If not you can contact me at the site forums.
I really do like CGI effects. Some of my favorite shows have really good CGI, Futurama, Lexx, other Sci-Fi. The problem is all the suspense is gone. I know that the person is in front of a green screen, I know they aren't in any harms way, except Jackie Chan.
There will always be a place for real stunt men, those are the scenes that still get me excited in movies. Seeing XXX snowboard down the side of a mountain in front of an avalanche may be cool, and loud, but it's anything but suspensful.
Hopefully we'll always have Jackie Chan and others to make that kind of entertainment to amaze us.
I really hate Dan Patrick.
StarShip Troopers.
The BUGS were the finest damn meshing of CGI and live action to date. They reacted realistically, didn't 'slide' along things, blah blah blah.
Of course, it helps that insects are SUPPOSED to have smooth yet jerky movements....
Vintage computer games and RPG books available. Email me if you're interested.
Okay.. why is that idiocy? Sounds more like a cosmetic choice to me.
I appreciate that, thank you.
"See you on the other side, Ray!"
Animation, traditionally, is very preplanned. The animation director draws a storyboard, which, in traditional animation, is a series of pencil sketches pinned to a wall. All the shots, and all the timing, is worked out at the storyboard level. Production then consists of filling in the gaps; frames between the storyboard frames are drawn (this is "in-betweening") details and colors go in, backgrounds are drawn, and all the pieces are assembled. All the creative decisions were made up-front, and are seldom changed during production.
Live-action film work isn't traditionally that structured. Some directors preplan everything; some just wing it. Directors have been successful with both approaches. Alfred Hitchcock and Roger Corman represent the extremes of that spectrum.
Then came films with mixed CGI and live action. Both parts have to match. This requires more preplanning. A lot more preplanning. The newer Star Wars movies were described as "years of preproduction, a few months of principal photography, years of postproduction".
It's not as much fun for the director as it used to be. But unstructured directing runs the budget through the roof. ("Space Jam" ran into this problem. I went to a talk by the lead effects guy, who was trying really hard not to describe the director as an asshole. That project went into postproduction hell, where, every morning, the director, his cronies, and studio execs viewed the dailes from the last night's rendering and ordered changes. First shift animated, second shift rendered, third shift transferred to film and developed. This went on for months.)
One way out of this is to put the animation director in charge. Now everything synchs right, and there's less rework, but the acting may be wooden.
There's a trend towards doing the entire movie twice, first as a low-quality computer animation, and then for real. This allows studios to see what the film will look like before green-lighting the production. It's basically big-budget storyboarding.
The industry is still struggling with this.