▲ 168 points
back
97 comments
Fascinating, thanks for sharing. It's amazing how many novel techniques went into the film. I haven't seen it yet, but looking forward to it.
One of my favorite TV shows as a kid was Movie Magic. Back in the 90s, most effects were practical, rather than CG, and it was interesting to learn about the various techniques used to achieve the results. From matte paintings, foley sounds and miniature sets, to animatronics and pyrotechnics. There were entire industries dedicated to this, which were replaced—practically overnight—by computers.
While there was a rough patch of awful CGI in the 90s and 00s, and the tech has certainly come a long way since then, even today, a discerning viewer can still tell CGI apart from live action.
I feel like we lost some of that "magic" in the process.
Consider the animatronics used in Jurassic Park[1] and Free Willy 2[2]. They might be crude, stiff and surely very laborious and expensive to build, but the result on screen was much more impressive than today's state of the art CGI. Maybe we've been saturated and take VFX for granted today, but it's a shame that some of these practical effects didn't continue to improve, and are rarely used in modern films.
You mention Jurassic Park, but the fact is that movie wasn’t doable with animatronics alone. The most memorable shots in that film are CG.
I also contest your assertion that the results are more impressive than todays CGI. That’s some rose tinted nostalgia, but there’s no way to replicate some of the advanced motion and realism like the latest Planet of the Apes or Avatar with animatronics. The range of motion required is just not safe to do or cost effective.
Safety is a big one. Having an animatronic T-Rex chase a car at speed in the rain is not physically safe to execute in the same space as your main actors. The requirements are at odds. To work in rain requires materials that are hard to move, to move fast requires lighter materials and heavier motors. All of that is a giant weight hurtling at your talent. Use of lighter materials also makes them less durable or more expensive.
Furthermore, most animatronics and practical effects in film today are replaced. As much as people sell the allure of “this film is done practically”, the majority of it gets replaced with CG in post. The new star wars films tried that angle with BB-8 and Top Gun Maverick claimed the same, but they practical elements people laud were CG in the final film.
But at least the actors have something to work with, rather than having to imagine it. This makes a huge difference.
Nearly everything on a movie set is wildly fake. Imagination is required even for conventional acting. The whole job is "telling the truth under untruthful circumstances".
It can require a different toolset when it's just you and a blue screen, but in many ways it's the same craft that every actor develops anyway.
That can be handled in many other ways, including having people in suits as standins.
Often, that’s better than animatronics because they can move faster and react to the actors better.
What is shot on stage doesn’t have to be what’s used in the film, and it’s a common mistake that people who are critical of CGI make.
> The most memorable shots in that film are CG.
I know it used CGI in some shots, but I disagree that these were the most memorable. Close-ups of the T-Rex, the creeping raptors in the kitchen, the sick Triceratops, the expanding neck frill of the Dilophosaurus, etc. These were the most tense and impressive shots IMO, and all were practical (even though they were historically innacurate). CGI was used mostly for wide shots that needed to show movement, which wasn't practical or sometimes even possible with animatronics.
> I also contest your assertion that the results are more impressive than todays CGI. That’s some rose tinted nostalgia, but there’s no way to replicate some of the advanced motion and realism like the latest Planet of the Apes or Avatar with animatronics.
I don't disagree that nostalgia is a big part of it, and the fact that back then these were truly ground-breaking shots we hadn't seen before. Each film did things slightly differently, and VFX weren't as accessible as today. Still, there's something raw and, well, real about an actual physical entity that actors interact with. The lighting and physical properties are always correct, and assuming the makeup and robotics are state of the art, it really can trick our brains into believing these are real objects much more than CGI could.
For example, take the sick Triceratops scene[1], and compare it to 2015's Jurassic World[2]. Sure, the CGI versions are more life-like, but they look completely fake. The lighting and shading is off, the interactions with the real world is limited and unrealistic, etc. In contrast, the scene from 1993 still looks impressive today, even though the movement is limited, and it arguably looks a bit stiff. State of the art CGI today is much better, though[3], but still slightly off in a few places.
This can possibly be blamed on "bad" CGI, but I'd argue most usage in movies falls into this category. Even the creatures in this new Avatar, while incredibly detailed and life-like, look artificial, and like they don't belong in the world around them (which is, ironically, also CG). It looks like an animated, instead of a traditionally shot, film. Which doesn't mean it can't be enjoyable, it's just a different experience. For me, it crucially lacks the awe-inspiring moments that films with practical effects often have. I realize that most of these shots couldn't be possible with practical effects anyway, but I appreciate when directors make an effort in using mostly practical effects, instead of defaulting to doing everything with CGI.
As another example, consider All Quiet on The Western Front[4]. According to the VFX lead:
> The overall VFX mantra was to allow the footage from principal photography to serve as the primary material for compositing, as opposed to creating CG elements in VFX to augment the scenes. “I believe that adds to the visceral, immersive quality of the film,” Frank Petzold explains. “So in other words when, say, there was an explosion scene, we did not entirely rely on CG models.” Rather, the team used photographic elements as the primary foundation, implementing CG simulations to augment principal photography. The effect was naturalistic, and the integration aimed to be seamless.
I think this is the right approach to take, and the film is much more immersive and enjoyable because of it.
[1]: https://www.youtube.com/watch?v=tYm7xeg27Rw
[2]: https://www.youtube.com/watch?v=2CJL7rKJI18
[3]: https://www.youtube.com/watch?v=le0iiWwdeeE
[4]: https://www.youtube.com/watch?v=hf8EYbVxtCY
[5]: https://www.fxguide.com/quicktakes/all-quiet-on-the-western-...
The raptors that are CG in most of the shots that aren’t just their heads? Like the entire sequence where they chase the kids…
you’re also comparing things without considering how the process created them.
Avatar wouldn’t look any more real with practical work. It’s just not possible to make robotics today that can have that level of articulation and micro expression. the reason it doesn’t look “real” is because the film makers didn’t want it to look real. It’s meant to be fantastical.
You claim most CG is bad in films but I’d argue that you, like many others, glomp onto the bad CG and ignore all the fantastic CG that you don’t notice.
How many people do you think noticed that the entire environment they’re in isn’t real in these shots? https://youtu.be/V5mS7BHmZJI
Often what people consider “bad CG” is overly apparent CG. This can be down to three things:
1. Budget. Both money and time.
2. Deliberate Aesthetic choices
3. Directors often put CG more in your face than practicals
Practicals wouldn’t make a difference there because they’d fall prey to the exact same issues which is why they’re often replaced in post anyway.
Aesthetics are a big one. The visual style between the original Jurassic Park trilogy and the Jurassic World trilogy is different. The new trilogy is significantly more fantastical in style and lighting. This is why you can’t just randomly compare shots between different films. They’re created by different teams with different goals. You need to compare against the raptors etc in the same film or where similar choices were made.
When people say “I wish this was done with practicals”, what I think they mean is “I wish this was better” but ignore the why.
The Hunt For Red October never got wet. The underwater scenes were miniatures with smoke floating to give the appearance of water. Listening to the stories of the people working on these types of models and the details they put in is so much different than someone at a computer with a 3D app.
Indeed:
“Making of "The Hunt for Red October" - Behind the Scenes” https://www.youtube.com/watch?v=2_epfA20dOY&t=21m55s
“HUNT FOR RED OCTOBER miniature effects part 1” https://www.youtube.com/watch?v=V5Npc6GsUeQ
“HUNT FOR RED OCTOBER miniature effects (part 2)” https://www.youtube.com/watch?v=MGu8qcI7r-I
Yeah, it’s too bad the story was so terrible. It’s a shame when there is a mismatch with the effort that went into the tech vs the writing.
I quite liked the story. Maybe a little long. Most people I know liked it too.
A good 1/3 of the runtime was James Cameron going "look what our tech can do".
And I actually didn't mind, the CGI was gorgeous.
It really is such a shame given just how much money got poured into the project that they didn't allocate even 0.1% to writing an actively good story.
At this point, I think that while Cameron has great visual taste, he simply just doesn't have any literary taste at all. I believe he actually thinks these horribly banal, cliched stories are genuinely good.
It's gotten so bad that it makes me retroactively wonder whether he actually wrote "Aliens" and who the real author is.
I don't understand this criticism. The story was amazing.
Do you mean that you don’t agree with the criticism, or that you don’t understand why someone would dislike the story? I feel like the former is the more understandable sentiment, since the latter is just a matter of personal taste.
After a quick read of critic and user reviews through metacritic, the complaints seem pretty consistent regardless of how I personally feel about them. A ton of them say the plot was barebones and predictable. Even the glowing reviews from critics seem to mention that the plot only gets the job done.
Are you talking about the latest Avatar movie? I really liked the story.
You knew what you were doing when you wrote this you anarchist.
Did you even see the movie?
Check out the YouTube channel Corridor Crew, I also liked movie magic and I think you'll like their content
Thanks! I've seen some of their videos over the years, but hadn't subscribed yet.
It's a different "magic" in recent films for sure. The images are visually pleasing, and have a kind of muscular, wet-and-glimmering look.
The old concept of animating physical objects to make them look like they're inhabited by a spirit and behave/bleed like real animals has been absorbed into motion capture and animation rigs.
We also used to love seeing people grow into new behaviours like luke in star wars. He was awkward and unsure of how he should act, and he grew into his new role as a jedi. Another example was Alan in Jurassic Park guiding people through the park by behaving in a manner that wouldn't get everyone eaten. But it's common for people to understand how to behave in action films and real life. The demand to see people learn a new set of behaviours, isn't such a strong pull to watch movies lately.
I struggle to imagine what a new-and-improved look for animated bones and lifelike animatronics would be? An engineer could make them stronger, better, faster and move more smoothly, but the higher experience people could want.. I don't know what it could be.
> even today, a discerning viewer can still tell CGI apart from live action
You can tell BAD CGI apart from live action.
I don't disagree with the conclusion in that video. But my criticism is precisely when filmmakers rely solely on CGI for every aspect in a scene. When used in moderation, and to _augment_ practical effects, as in the original Jurassic Park, or All Quiet on the Western Front, then it can increase, rather than break, immersion.
Eventually, we'll reach a point when CGI truly becomes indistinguishable from reality, but even then I'd argue that using it for all aspects in a scene would undermine what makes the art of cinema "magic". After all, would it really be filmmaking, or animation, at that point?
The amount of work that went into those orcas is insane.
I appreciate all the work and dedication of those folks, but it feels like such a waste. Especially when the final film wasn't very good.
The craft is admirable, but it was never scalable or equitable. Very few projects could benefit from this kind of budget and vision.
The coming AI revolution will make this vastly cheaper than even CG, and it'll give more people a hand at storytelling. Without requiring fifty engineers a full year of their talents.
Of course we did. This is why otherwise ordinary "Top Gun 2" was so massively successfull: lots of live action and practical effects.
I think we will lose it at all eventually. Today's audience is not making a big difference between movies and computer games, so the current disgrace of Marvel comic movies is just okay for them (old_man_yells_at_cloud.gif)
Tangential, but reading about niche domains like this its crazy to think how different the things people do under the job of programmer
Two people could get CS degrees and one ends up building UIs/forms in React the other is implementing “Lagrangian water wave simulation”
Just totally different worlds
To be fair, the second person is more than likely pursuing more advanced education than a CS BSc (PhD)
> niche domains
I’m not sure if I would call movies and video games “niche”.
If you get a CS degree and end up doing UIs/forms in React then something along the way went very very wrong.
yeah one person makes a shit ton of money wondering if they needed a degree at all, and the other person ends up getting overworked for pennies and no equity in movies and gaming studios
You realize that there are 2.7 million software developers in the US? Outside of the HN bubble, most are “dark matter developers”.
Completely aside from Avatar, I'm hoping the tech for things like underwater motion capture and this water modelling brings about a renaissance in underwater themed games.
I'm looking at you Monster Hunter and also Blitzball >.>
When I heat the pan and throw some drops of water on it, the drops "dance" and emit a hissing sound. This effect haven't been realistically simulated and would be a good pretext to shrink those blue aliens in the next movie.
That's the Leidenfrost effect. Maybe we'll see it in a Disney movie called "Leidenfrozen".
I think that the behaviour you're describing is due to the Leidenfrost effect!
I wonder if UE5 or Blender have similar and perhaps competitive methods for the Loki water state machine in generating realistic water and mist. The article uses a lot of jargon:
Navier-Stokes simulators, Lagrangian wave packets attached to spline curves, incompressible two-phase Navier-Stokes solvers on a Eulerian grid, and air phase representation by FLIP (Fluid Implicit Particle) and APIC (Affine Particle-in-Cell) solvers.
I remember the Quake I water simulation, which I found amazing at the time. I knew water is hard to simulate, but understanding fluid dynamics and the above jargon seems hard to grok too.
The techniques that control the way electrons flow in silicon is hard to grok. But that gives way to logic gates, that let you build CPUs with branch predictors. The way that branch predictors actually work in CPUs is hard to grok but how to write code that keeps their implications in mind is more approachable. Learning how to use libraries that leverage the above is even easier to grok.
Technology is built upon a stack of useful abstractions, and the ability to understand in deep technical detail the entire stack is a super power. There's no reason why computational physics shouldn't work the same way.
Unity also bought SpeedTree, which did the vegetation in Avatar 1 (and in many video games). They bought Weta Digital in 2021 (https://blog.unity.com/news/welcome-weta-digital).
Blender has a FLIP plug-in I believe. The water in UE isn't great, but the "fluid flux" plug-in is surprisingly impressive for real-time water https://www.unrealengine.com/marketplace/en-US/product/fluid...
Computational Fluid Dynamics is a tricky area. The science (Fluid Dynamics) was worked out over 100 years ago, with the Navier-Stokes equations (3D partial differential equations). The engineering (the Computational part) is a whole bunch of complex and clever tricks, to overcome surprising problems like instability and dissipation, and often for some particular engineering trade-off (rather than being universally true).
Avatar 2 seems to use an underlying FLIP simulation for the water (which uses a (eularian) grid for pressure, and particles for movement), with cinematic tricks on top: foam, bubbles, and smaller (lagrangian) waves superimposed.
The Fluid Flux plug-in uses a simplified version of the Navier-Stokes equations, which represents the upward direction as an average - so it's 2D internally but looks 3D. It also just calculates once, instead of iterating. Both make it much faster. It uses this paper, which has several other tricks: https://matthias-research.github.io/pages/publications/hfFlu...
It’s sad that the videos are unavailable if you prevent tracking cookies.
Over each video I get “This content is hosted by a third party provider that does not allow video views without acceptance of Targeting Cookies. Please set your cookie preferences for Targeting Cookies to yes if you wish to view videos from these providers.”
Yeah, and the video elements are created dynamically so you can't just find the video trivially by opening the DOM inspector.
Here are the videos:
- “Loki: A unified multiphysics simulation framework for production,” presented at SIGGRAPH 2022 https://www.youtube.com/watch?v=_2TBvSnsLD8
- “Wave Curves: Simulating Lagrangian water waves on dynamically deforming surfaces,” presented at SIGGRAPH 2020 https://www.youtube.com/watch?v=FqCK5Gu-Op4
- Overview of “Guided bubbles and wet foam for realistic whitewater simulation” as presented at SIGGRAPH 2022. Video courtesy of Wētā FX. https://www.youtube.com/watch?v=60Dk5h_659s
Code:
for (const deck of __NEXT_DATA__.props.pageProps.entity.decks) if (deck.url) console.log(deck.caption, deck.url)Thank you!
Cameron hired all the competent, experienced CGI / animation people and the rest of the special effects industry hasn't moved ahead much since. It's even regressed in places.
I wish they’d invested a little less on the technology and a little more on the writing.
It doesn’t say in the article how this was implemented. Was all this simulation done in something like Houdini or is there a custom engine Weta uses?
Actually it says it was done in Loki, which is explained in this journal article https://dl.acm.org/doi/abs/10.1145/3528223.3530058