Hacker News new | past | comments | ask | show | jobs | submit login
The bullet effects in Terminator 2 weren’t CGI (hackaday.com)
899 points by zdw 75 days ago | hide | past | favorite | 303 comments

In the movie The War Lover, Steve McQueen (well, the actual pilot) pilots a B17 bomber "cutting the grass", i.e. flying low with the gear up. The stunt was to be done with models because it was too dangerous.

My dad was in the AF at the time, and was assigned to the movie as a consultant on AF trivia. He convinced them that the stunt could be done safely, and explained how to do it.

And so it was done. My dad got in trouble, because if the B17 had crashed doing the stunt, the AF would have looked bad. But it didn't crash, and the stunt was a big deal to the movie crowds, so he was forgiven.


It's still freakin' impressive today.

His other contribution to the movie was the mission briefing wall maps.

There were 3 B-17s in the movie. If you look close, one of them is named "Round Trip Ticket", which was the name of my dad's B-17 ride in WW2.

Sadly, the movie company burned all three to the ground after the shoot, in order to avoid paying the import duties the British tax people demanded. They didn't have the money, and didn't have the gas money to fly them back the US. They tried to give the bombers to a local museum, but the tax people still wanted the import duties, and the museum couldn't pay it.

What a tragedy.

George knew the deal:

I'll tax the street (If you try to sit, sit)

I'll tax your seat (If you get too cold, cold)

I'll tax the heat (If you take a walk, walk)

I'll tax your feet (Taxman)


> They didn't have the money, and didn't have the gas money to fly them back the US.

Knowing Hollywood accounting (really just fraud, but "its always been done that way" so I guess its cool now): maybe they just didn't want to? Not that I want to excuse the tax apparatchiks, this is horrible.

Flying a B-17 is very expensive.

This reminds me of the bison on Catalina Island.

"! Fourteen American bison were brought to Catalina in 1924 for the making of the Zane Grey film The Vanishing American. After all that effort, the footage of the Bison didn’t make the final cut, and they do not appear in the movie. When filming was completed, the production company left the animals behind! "


Tax are also a way to contribute to the public good. Not paying the tax is refusing to contribute.

So although I do understand it's a sad situation, the movie team should have first made sure it could pay the tax and then do the stunt.

How do you justify burning assets for tax purposes then? It’d be much better if they could simply give the planes to the taxman if they couldn’t pay the tax.

It's possible for a policy to result in an net positive even if it entails some individual negative examples.

it's a given for any policy, but we aren't talking about generalizations here; in the end, it's the particular that matters.

It depends on the alternative you are suggesting. No taxes? A bureaucracy so vast that can handle all imaginable edge cases? Individual discretion of the tax assessor? I think all these could fix this particular problem but might lead to worse outcomes overall.

i suggested the solution right there - forfeiture of the assets that the taxman can't get the tax for.

Ugh I must have been half asleep when writing the original comment (toddler woke up at 4am). I mistook your comment for a knee-jerk libertarian reaction, but what you propose does make a lot of sense. I stand corrected.

Some taxes are necessary, but taxes which people avoid by altering their behavior in ways that are bad for society are terrible.

That's why pollution taxes are a great idea. You want people to alter their behavior to avoid the tax.

Wouldn't any tax then be terrible, as people will always try to change their behaviour to avoid it and in general paying taxes is a public good?

You can’t avoid most taxes, it’s taken out before you see the money. Some history was destroyed here because of an inability to pay taxes. You can argue whether taxes are good or not but that’s what happened.

It's not realistic, but if the definition of a terrible tax is one "which people avoid by altering their behavior in ways that are bad for society" seems like it's catching pretty much every tax in that.

Most people don't start shopping at grey/black markets in order to avoid paying sales tax. Most people don't pay their employees under the table to avoid paying income tax. Most people don't move or de-value their own homes in order to avoid paying income tax. Those taxes make up most tax revenue, and wouldn't get caught in that definition.

So it's only those with the means to avoid the taxes are those who suffer under terrible taxes?

I'm sure the public good is served best by burning down WWII bombers.

In a paradoxical sense yes... Obviously in this case it was pointless zealotry, but one of the main purposes of import taxes is to artificially limit imports to favour internal markets, in this sense free stuff is actually worse than expensive stuff.

You mean so the domestic market for WWII era bombers is protected?

The hallmark of the bureaucrat is favouring process over outcomes. These people are the enemy.

I mean that this is the reason why import taxes exists, had it been with some generic item th UK manifactured mostly internally at the time it would have been sort of reasonable.

I am not a fan of calling people "enemies" I prefer to look at their incentives and context so that when can learn how to change them so that we will have less "enemies" in the future

In this case the airplanes had also an historical significance so it is hard to justify forcing their destruction.

> I mean that this is the reason why import taxes exists

If your point was about the generic usefulness of import taxes, then I have misunderstood you, no objections there.

> I am not a fan of calling people "enemies" I prefer to look at their incentives and context

I suspect you are in your rationalist phase right now, given the Scott Alexander link in your bio. I used to be one too. This is a good starting point, but it is just that - a starting point.

> can learn how to change them so that we will have less "enemies" in the future

What you probably have in mind is to have a civilized discussion among reasonable people and thus determine the best overall outcome for the best of all of society. Changing the system so that the resulting forces will produce different outcomes.

I'm not really qualified to elaborate on the topic of changing power structures (in essence: of the problem of governance), so I'll just stop here. But its a different problem than you imagine, and you will recoil from the things that are required to succeed.

Interested if you could elaborate further here or have some pointers to good references. I feel like I'm at a turning point in my life where I'm starting to see things a little differently. Sometimes bewildered by how little knowledge or control there really is.

> Interested if you could elaborate further here or have some pointers to good references.

Twitter is good. I don't want to recommend specific accounts. Keep an open mind, judge for yourself who has the best arguments.

> I feel like I'm at a turning point in my life where I'm starting to see things a little differently.

That number is growing.

> Sometimes bewildered by how little knowledge or control there really is.

There is a lot of knowledge, its just not evenly distributed.

Here's how it was done:

1. do it at dawn, when the air is the stillest

2. have the pilot make pass after pass, each time a bit lower

3. have a ground observer in constant contact with the pilot giving the clearance from the ground

That left bank on one of the mowing passes in the YT video is just heart stopping! A few degrees more and ....

Still, I wouldn't like to be in the plane when at point 2. That's really low !

Planes are that low twice every flight.

Yes, but with their wheels down and at a much lower speed =)

Iirc ground effect is going to help a lot during such low level flight, in fact there was a plane conceptualized (and trialed) to never be flown out of ground effect. That's why it has such stubby wings, the ground effect itself makes up for the absence of lift area:


That's not to belittle the capabilities of the pilots flying in these takes, it's a very well executed maneuver and certainly not entirely risk free (assurances aside, such stunts have also gone horribly wrong). But it's not quite as hard as you might think because the ground helps to 'push back' in case one of your wings gets too low. The bigger risk is probably from cross wind, which would affect the plan asymmetrically.

>"cutting the grass", i.e. flying low with the gear up.

Unfortunately, one of greatest stunt pilots died doing similar stunt for the original "Flight of the Phoenix"


A touch-and-go is not normally dangerous. Students do it all the time to practice landings. The error in that movie was the airplane was structurally unsound.

Kinda ironic when the plot made a big point of model airplane design being the same as big airplane design.

Sad when someone dies because of it though..

PS I'm surprised that thing was FAA certified. But this is really a case where a model would have been a lot safer.

That's neat, wow. Such a big plane at that speed and less than half a wing off the ground (so in ground effect).

This is another video in a slightly different fashion, but it gives a good idea of the feelings of the people present during a low altitude fly-over, especially an unexpected one:


If that was a movie someone would spill their coffee as a cliché :)

Ps that video doesn't do justice to how LOUD those things are :)

Ok, it's impressive, but unfortunately the YouTube video does't state the reason why this stunt was deemed necessary in the context of the movie. Hopefully not just "because it looked cool"?

> Hopefully not just "because it looked cool"?

Isn't that reason as good as any other?

Doing it with models would have been real bad I guess. Before Lucalsfilm, models were quite looking like models.

Just go look at a 1980's science fiction movie and you'll see lots of really bad model shoots, dynamics all wrong and so on (even the big ticket ones like James Bond movies got it wrong routinely). But the audience wasn't as spoiled and likely never noticed, I only realized how cringy they were on re-watching with my spoiled 2021 eyes and mindset.

I've really come around to appreciate the cringe. When you see obvious models or extremely bad painted backdrops edited in it makes me appreciate the attempt. Maybe I'll feel that way about bad CGI one of these days, but practical effects just hit different.

I think what also doesn't help is that resolution is just much higher today. A lot of model clips would have been smoothed over by the low resolution but these days we all watch the 1080p or 4k remasters.

Some stuff that stands out to me was star trek tng where you can see the wood grain in things like doorways even though they're painted. No way that would have been visible on TV.

Other stunts in the movie were done with models, and you can see how bad they are.

Even today, I doubt that buzzing sequence could be done as convincingly as the real thing. I've seen some aviation war movies done with CGI, and it just isn't quite right.

In the movie, the pilot (Steve McQueen) is a psychopath who is venting his anger at being sent on a leaflet dropping mission rather than a bombing one.

Buzzing the field is against orders.

BTW, B17 crews would sometimes buzz fields in WW2 to celebrate having completed their tour of duty and they could go home.

I don't think it is cool. Just think about the recent news about actors getting shot dead. It is a damn movie. Why risk the lives of actual real humans. People have no problems getting people killed for shooting movies but we can't even seem to take the slightest risk when creating new vaccines or drugs. Our sense of risk & probability and different reactions to human deaths in different situations are messed up

> People have no problems getting people killed for shooting movies

They did indeed have problems with people getting killed. That's why the stunt was to be done with models until it was shown it could be done safely.

BTW, a few years later when my family lived in Germany, sometimes you could see the Luftwaffe hedgehopping across the countryside. It was awesome to watch. The cold war was at its height, and flying low was how one evaded enemy radar. The Luftwaffe wanted to be well trained and ready, and the hedgehopping also let the Soviets know they were ready and able.

Your comment just reminded me of this catchy BO-105 demo from the early 80's :)


It is cool.

We're all adults and can make our own choices. Few things in life are free of risk (you drive a car I imagine). If I want to do a damn stunt I'll do it. Should we also ban most sports? Those are dangerous too. How about we stop building tall buildings? Lots of construction accidents there.

Also, the kind of accident that happened recently on set is vanishingly rare. There are a ton of safguards in place to ensure things like that don't happen and, in this case, at least one person was severely negligent.

The War Lover had inflation-adjusted production budget of 23M USD, so this had to be used somewhere too.

Beyond that, there can be many reasons; for example, in a similar situation with Mission Impossible: Rogue Nation.


1) It's fun

2) Interesting to work on for the director and the crew. If this was shot with a digital screen and a fan, that'd super be boring to produce.

3) Brings good marketing

4) Can justify paying expensive advisors / raising more production money (these stunts are difficult to financially evaluate so they can funnel out budget easily)

5) It looks more realistic (though it's not necessary a quality, as the story-telling is sometimes more important than realism).

Did you compare deaths on movie sets to deaths during medical trials recently?

It's the exact opposite situation from what you describe, which is why it has been newsworthy. But I don't understand how your takeaway became the opposite of the described reality.

Movie stunts and Drug development are apples and oranges.

I would be interested in seeing the numbers on prop guns fired in filming vs injuries. I expect it is enormous. The risks are simple, understood, easy to prevent.

Many of the risks in drug development are largely unknowable before human testing and we still use human subjects.

>I would be interested in seeing the numbers on prop guns fired in filming vs injuries.

I had a boss try to bring this up at lunch last week. "I was watching an old western and they have prop guns everywhere, how come no one died back then?" Pulling up a Wikipedia[0] listing on movie accidents brought up quite a lot of old western and war movie accidents.

Although it seems vastly more animals got hurt on set than people. A lot of early war/western movies seemed to be extremely detrimental to horses used on set.


>I had a boss try to bring this up at lunch last week.

What is the difference between bringing something up and trying to bring something up?

NPr states the last prop gun death was Brandon Lee in 1993, so ~30 years ago. I still wonder how many movie blanks are fired a year. 1 million? 10 million?


Propaganda was and is vitally important to the western power bloc.

Yes, absolutely true as it is for all power blocs. At least we can say so openly though, and question it and even argue against it in public. I think those are freedoms worth protecting and promoting.

Have you seen Chinese or Soviet propaganda? It's terrible and amateurish compared to the subtle, brilliant, cool and therefore exponentially more efficient and dangerous western one.

I think that's based on a misunderstanding of what these two approaches to propaganda are trying to do. They are both highly effective, but at different things.

The primary means western propaganda uses is to persuade. It has to be persuasive because the targets can choose whether they believe it or not. Nobody has to comply, and dissenting voices and alternative narratives are available. Also it competes in a market for ideas and so there's a degree of selective pressure to improve so you're quite right it has become quite sophisticated whereas authoritarian propaganda can rely on excluding dissenting voices.

Soviet and Chinese propaganda, at least the types you're taking about, persuade by projecting power. It's about psychological domination, with heroic poses and strong simple messages emphasising the inevitability and necessity of party control. It's blunt design is part of the point. Thousands of gymnasts and marchers in perfect lockstep. Zero individuality. It doesn't matter if you are persuaded, as long as you get the message that you have no choice but to comply, and look, everyone else is complying!

Take the poisoning of Navalny. They don't need to persuade anybody in Russia that the state didn't do it. Everybody knows the state did it, even those that are pro-state. The point is to send the message, this is what will happen to you if you push us too far.

>Take the poisoning of Navalny. They don't need to persuade anybody in Russia that the state didn't do it. Everybody knows the state did it, even those that are pro-state. The point is to send the message, this is what will happen to you if you push us too far.

I don't think Fred Hampton was persuaded as you describe.

Absolutely, that was an appalling incident. It's a tragedy not just that it happened, but that none of those responsible went to gaol. Hopefully if anything like that were to happen today they would face the full consequences of their actions. It's much more likely that those in authority would be held to account nowadays, but still not certain unfortunately. We must continuously strive to do better.

so many downvotes. I guess it only make sense to only make comments when you're sure the perspective is the most popular ones among the HN audience.

> it only make sense to only make comments when you're sure the perspective is the most popular ones among the HN audience

Or just don't worry about it. I have many unpopular opinions and posts :-)

You also have the best programming language in the world and a lot of people dispense only negativity about it.

Do you comment only for upvotes? Who cares?

Alec Baldwin was practice shooting at someone who wasn't even an actress. why? why even aim at anyone at that point? he's as much to blame as anyone else.

The most fascinating non-CGI effect from Terminator 2 is a cut scene which can be seen in the director's cut, where they use Linda Hamilton twin sister in a surprising way: to fake a "mirror" scene where Sarah is operating on the Terminator's head.

They are in front of a mirror. They wanted to actually use Arnie in the reflection to show his face; they wanted a dummy for the head closest to the camera, to show the surgery inside the head. So they used a window in the wall instead of a mirror! Linda Hamilton and her twin sister are mimicking each other's movements to make it seem it's a mirror image.

(The other use of Linda's twin sister, the scene where the T-1000 fakes being Sarah, is better known)

I agree this is super neat, but for what it's worth, the "the mirror is actually a hole in the set" technique is kind of a standard go-to in movie making.

I am very enamored with the iconic "impossible mirror" shot in Robert Zemeckis' "Contact" instead, which used subtle CGI to brilliant effect to innovate over the practical bag of tricks. It's my single favorite effects shot in movies: https://youtu.be/avRdYf78kLk (from 1:10 for the impatient)

On the topic, another neat video covering several "impossible mirror" shots in movies (including Contact): https://youtu.be/VASwKZAUVSo

Amazing. I do wonder if I'd actually notice just watching it (haven't seen the film) vs. it being pointed out to me though. I think I miss (just thinking 'nice cinematography', but not realising specifically why) a lot of clever stuff.

For what it's worth, i missed it when watching the movie. It just feels like there's something weird that just happened but you can't quite tell, like a nice and subtle spice in a well made dish.

> I do wonder if I'd actually notice just watching it

Almost certainly not; that's the point.

I don't remember whether I noticed it or not when I saw the movie, but I'd like to believe I would have. It reminds be a bit of the final scene of Hitchcock's "Birds":

https://www.youtube.com/watch?v=ULxw4UDY5g8 (up until 0:35)

1. What reminds you of Birds? I don't see the parallels.

2. The original Birds did not have that music. I cannot believe how classic media remasters completely ruin the tone of the originals. From Star Wars, to Rust in Peace, to Birds now.

The parallel is not obvious because no mirror is involved. It is the fact that camera is moving backwards in front of the actors who approach a closed door. Then Rod Taylor reaches for the handle and opens the door and the outside light shines in.

How come the camera is not in the way at that point? Did it move backwards to the closed door? The answer is quite interesting. Let me know if you'd like me to spoiler it for you.

I remember being surprised by the effect, and totally unsure how they pulled it off. Positioning the camera in exactly the right angle to match what the moving camera was shooting? I can't even picture it. Impressive work!

There is another set of twins used, that is the security guard who gets a full house on the coffee machine. He meets his 'double' that is actually the t-1000, who 'pokes' him through the head. The actor duo are in a few other movies, including Gremlins 2, where they aren't trying to fool the audience, but just are twins.

Also some of the best gags: https://www.youtube.com/watch?v=0D-n0yctyGY

This is also pretty common in video games where a mirror is just a window into another room with a duplicate player model.

It's what they did with Duke Nukem 3D, I know. But in videogames it's not a practical effect, but "CGI" by definition, so it impresses me way less ;)

That's not entirely accurate. DN3D required a large room behind the mirror (I'm saying this from the perspective of editing levels), but that room was completely invisible and did not have to look like the mirrored room at all. The mirror actually "mirrored" the scene. IIRC, if you disable collisions then you could walk through the mirror into that big room, which looked quite boring.

I always assumed that the "big room" was to avoid the culling algorithms / potentially-visible-set tripping over the mirror.

Seems like there should be a computationally less expensive way to do that.

Screen-space reflections are faster and require no new scene draws, but they won't show you "new angles" on reflected objects and they won't go around corners. Planar reflections require a new scene draw, but that's actually pretty modest: non-planar reflections require six scene-draws for a cubemap or full on raytracing.

In general, lots of things in games require re-rendering the scene multiple times per frame, so game engines are really, really good at this. Shadows, projections, transparency / translucency, motion/depth blur, indirect lighting, deferred lighting, even things like hit testing (though for dependency + latency reasons that's a prime candidate for offloading to the CPU). Anything that requires less visual fidelity will have its performance tuned accordingly: fewer pixels, simpler shaders, lower LOD geometry, lower framerates. Nobody will notice if the fuzzy bounce lights are calculated once every five frames and blended.

Ray tracing is conceptually simpler but (so far) slower at any given level of cleverness -- you'll often hear that it has asymptotic benefits, but those same benefits can be pulled out of rasterizers with effort comparable to what is required to make the raytracing practical.

Here is a writeup of the passes you might expect to see in a modern engine: https://zhangdoa.com/posts/rendering-analysis-cyberpunk-2077

Here's an overview of a modern geometry pipeline: https://www.youtube.com/watch?v=eviSykqSUUw

I don't think you could do a mirror with screen-space reflections.

Sure you can, people do it all the time, but I'll grant you that it kinda sucks (which is how I notice that people do it all the time). There are situations where everything you want to reflect is already on screen and the angular difference between the camera and reflected camera isn't enough to be noticeable. When those conditions are met, SSR is awfully tempting. Water is the usual suspect, but wall-mounted mirrors can work too if the camera will only see it from glancing angles and it's reflecting a wide-open space.

Tradeoffs, tradeoffs, everywhere.

In blender eevee render that's famous for screen-space magic you can get realtime mirrors with "reflection plane" "light probe" by placing it over the surface that's supposed to be a mirror.

Look up “instancing” with regards to geometry on GPUs. The mesh can just be rendered twice with different transforms.

That said, it often makes sense to drop some of the decorative flourishes used on the primary instance of the model due to distance, or use a different level of detail. There’s really no great way to get around needing to tell the GPU to render the triangles twice (short of rendering the view to an intermediate buffer and using that, which is still twice…)

La Haine also contains a fairly nice use of the "mirror set extension" trick: https://www.youtube.com/watch?v=okQJPUTQMqA

Interesting! In this case, was it used to avoid showing the camera?

Surely it's intended as a more deliberate (not technical) device - seeing his 'reflection' we're seeing him as he sees himself, his self-image. (I haven't seen the film though.)

Oh, I understand the cinematic effect, what I'm asking is why they used a trick instead of an actual mirror. I suppose because the actual mirror would have shown the camera behind, right?

Even for a momentary shot of a doorknob in ‘The Matrix’, the camera had to be hidden—in the jacket: https://collider.com/wp-content/uploads/2019/05/matrix-doork...

Very cool! I didn't know that. It's amazing all the kinds of shots that are done in movies that one would assume are trivial to accomplish, but that actually require tricks.

You are correct!

Without using this technique you'd have the camera in shot.

I don't think it would work with an actual mirror - either Cassel will block the mirror or the camera will block Cassel.

Somewhat on topic, the impossible flying camera in Soy Cuba: https://www.youtube.com/watch?v=rjuLkJ4m-mc

Cool! I didn't know it was more widespread. I suppose the innovation here is that the actress and her double are twin sisters.

They used that a whole bunch for the mirror scenes in Quantum Leap where Sam would look in the mirror and see himself as whoever he leaped into and there was someone else in the scene. That series was a boon for actors who had identical twins.

The other identical twin thing I remember was an episode of Buffy the Vampire Slayer where Xander was split into two, one of him all his strengths and the other all his weaknesses. Nicholas Brendon's twin brother was used in the scenes where there were two Xanders in one shot.

> (from 1:10 for the impatient)

To the impatient: The meaning of this shot depends on the first 1:09 minutes.

Amen! I remember frame-advancing the LaserDisc over and over to study how they resolved the shot at the end.

Huh? Do you have a reference for that?

This is what I remember from T2 director's cut: https://www.youtube.com/watch?v=yiBHj8Xs4sg (at 0:21).

When I saw the theatrical cut it always confused me how John could tell who was the T1000. And it also annoyed me that Arnie shot the frozen T1000, seemingly only speeding up his thawing.

In the director's cut you see that T1000 is actually damaged, but not damaged like a non-liquid robot would be. It's not losing mobility (control) of an arm, but instead losing control of its morphing functions.

It's complete genius that was (almost[1]) entirely cut from the theatrical cut. It's ridiculous that the edit it out.

[1] There's exactly one "flicker" of T1000 metalliness in the theatrical cut. A small remnant of a genius idea.

> Huh? Do you have a reference for that?

I can't find the "making of" video where they tell how they did this, but here is the cut scene: https://youtu.be/tZNE637BeEI

In scenes where you see Arnie's face in the mirror at the same time as his head being operated on, it's not a mirror, and one of the two Sarah's is not Linda Hamilton (whoever's face is shown less clearly).

Edit: lots of hits on Google saying this is how they did it, but I cannot find a single authoritative reference, like a making of video or someone involved in the production saying so.

Ok, I was being an idiot. I was thinking of the other Hamilton-twin scene.

Got it, yeah that was a good use of her.

Oh i remember as a kid it made sense, they didn't have that kind of relationship. Since his mom was a badass always fighting type, she didn't care about herself she just wanted John to live, but here she was calling him like "John, help me". While the real mom was just "Get out of the way" no BS type. And yeah like the other poster said, if it was T-1000 he wouldn't say get out of the way.

I couldn't agree more about the cutting out of the T1000's erroenous morphing. It adds what, a second and a half to the film? I can't imagine why they felt the need.

I also really like the scene where John tries to teach him to smile, but at least there they have the excuse that it's a minute or two in length.

Certain other scenes, like Kyle's visit and the extra nuclear nightmare scene, I think the movie is better without.

The one cut scene I admit without hesitation ruins the movie (and so they were right to cut it in the theatrical version) is the happy ending where Skynet is defeated.

The actor playing John Connor looks ridiculous. Linda Hamilton in aged makeup looks silly. The whole scene looks corny.

And it defeats the "no future" vibe that worked so well in the first Terminator movie.

"There's a storm coming."

I just figured that T1000 would never say “get out of the way!” instead of just shooting him.

I was too young to see T2 at the cinema and never got the big deal about T2 until I saw the theatrical cut uninterrupted in one sitting. The pacing is incredible, there literally isn’t a wasted minute. It must have been agonizing to cut and rearrange otherwise great material.

"Oh boy!"

The same technique of using an extended set to fake a mirror was frequently used on Quantum Leap.

Airplane! might be the greatest movie of all time. It's funny, but it is also suspenseful, has great acting, and is full of serious subtlety and commentary. Even with the humor removed it is a terrific film.

It was at least the third or fourth watching when I noticed that one. One of the densest movies I've ever encountered.

This is explained in one of the DVD extras on my copy of Terminator 2. I don't remeber if it's in the director's commentary of the movie or one of the extra videos and I don't have time to look it up. But it's definitely there.

From T3 (I believe this was a cut scene too) but really enjoyed the backstory for the T800 chassis and voice https://youtu.be/kayFrIR-Qfw

Ah... the brilliance of the old days.

Why not just use a half silvered mirror, that way you don't need the twin

If you could see the reflection of Linda Hamilton in the mirror, wouldn't you also see the reflection of the Terminator dummy?

Wait until you hear about how Cameron filmed the scene with a helicopter going under a highway overpass.


As an aspiring helicopter pilot, I am both awed and repulsed. Vietnam was something else for helicopter pilots. That level of skill would be hard for a perfectly-programmed autopilot to do.

With regards to Vietnam, a legend that I was told (but cannot confirm) was that Hueys had such bad tail rotor authority that they would take off straight up with the whole airframe spinning. Once high enough, the pilot would start the helicopter moving forward, which would allow the tail rotor to bite more, and the helicopter would stabilize.

Crazy if true. And after seeing a stunt like that by a bona fide insane and top-tier veteran pilot, I believe the legend.

Check out the book "Chicken Hawk" by Robert Mason, who flew Hueys in combat in Vietnam. These pilots were able to pilot in and out of some very dangerous situations using the unique qualities of the aircraft. I remember one involved taking off from a hilltop artillery battery by using gravity to drop the aircraft backwards off the steep hill after making a dropoff, or landing in a jungle using the rotors to chop away the vegetation (branches and leaves only, not trunks) to verify that a crashed U.S. aircraft had no survivors. He also talked about a stunt that another helicopter attempted to do, lifting a heavy pole or beam without the proper stabilization which resulted in uncontrollable spin and the death of all aboard.

Interview here:


I've read "Chicken Hawk", and while the stories are great, one particular scene made me not happy. Thank you, though.

How far are you along training? I'm trying to decide between helicopter or fixed wing. Or did you start from fixed wing and go to rotorcraft?

I have 15 hours, and my training is paused because I don't have a medical yet. (The FAA wanted me to be evaluated in the cockpit before issuing a medical; long story.)

I did not start with fixed wing, though I have flown fixed wing a couple of times, not logged. I wanted to do fixed wing when I was younger, but helicopter seems a better bet now.

The reason is because the fixed wing industry is mostly airlines. And when I say "mostly", I mean 90% or more. That means that the fixed wing industry is subject to the volatility of demand for airline travel. Even before COVID, I knew it was volatile; 9/11 made that clear as well. With COVID being endemic and people not relaxing restrictions, I doubt the fixed wing industry will really recover to what it was. It might, though; my neighbor trains pilots for an airline, and his schedule is full right now because they can barely keep ahead of the demand for new pilots. And there are still other jobs, so the fixed wing demand might remain pretty high.

But rotorcraft was really only affected in one area: tours. The rotorcraft industry is much less volatile, and the demand is just as high. That's why I chose it. Also, it's easier to learn to fly helicopters and then to switch to fixed wing later if necessary.

The downside is that getting your Commercial certificate for rotorcraft is far more expensive, just because even the lightest helicopters cost far more to run per hour.

I can't say which is right for you, but that should at least give you more info.

Thanks! That's more to think over, and absolutely makes sense.

Good luck with your progress, and hope the medical knot untangles itself.

literally any Vietnam documentary would prove that wrong. what does "moving forward" even mean when you're spinning? also, forward movement doesn't give the tail rotor any more "bite" ... if anything, it would get less. moving in a consistent direction would tend to orient the body of the helicopter to face into the wind because of the wind vane effect.

As a sibling comment said, forward movement does indeed give the tail rotor more authority because it has "cleaner" (less turbulent) air to work in. This makes it more effective. For a source, see [1], pages 2-21 through 2-23, subsection "Translational Lift", and Figures 2-39, 2-40, and 2-41, as well as [2], pages 11-18 through 11-21, section "Loss of Tail Rotor Effectiveness".

"Moving forward" means tilting the rotor disk. Even when spinning, this will induce a little thrust, which will help clear the "dirty" (turbulent) air from the tail rotor, and as you said, induce a wind vane effect, which would stabilize the spinning a little, which would then allow more forward airspeed, and it would compound. I could see the helicopter flying a widening spiral as it gained forward airspeed until it fully stabilizes in straight-and-level flight.

Loss of Tail Rotor Effectiveness is a big deal in most helicopters where the tail rotor is not big relative to the size of the helicopter. In Robinson helicopters, the typical trainers, the tail rotors are massive for the size of the helicopter, and you will almost never run into Loss of Tail Rotor Effectiveness.

[1]: https://www.faa.gov/regulations_policies/handbooks_manuals/a...

[2]: https://www.faa.gov/regulations_policies/handbooks_manuals/a...

While hovering, the tail rotor is pushing air into a dirty vortex. A small amount of speed, even just 20kt, will clear that vortex and give more authority. Even today helicopters will begin pitching nose down immediately on takeoff to get airflow cross both the main and tail rotor to clear those vertices.

That being said, I've talked to a few Vietnam helicopter pilots and all of their stories were ... well, exaggeration is a common theme.

pitching the nose down certainly does one thing: it pushes air backwards, meaning the helicopter starts moving forwards. the air moving across the body of the helicopter, rather than down it, is what starts the wind vane effect.

if the pilot tipped the helicopter to the right after leaving the ground, and then continuously held the control stick in the direction of the movement of the helicopter, the body of the helicopter would begin to face the direction of travel unless the pilot actively countered that turn via the tail rotor.

> if the pilot tipped the helicopter to the right after leaving the ground, and then continuously held the control stick in the direction of the movement of the helicopter, the body of the helicopter would begin to face the direction of travel unless the pilot actively countered that turn via the tail rotor.

Yes, which is why it would work to just push the stick forward, even if the helicopter was spinning.

That is an incredible feat of flying. I would be terrified of some kind of reverse ground effect from the road overhead.

It's basically a 1990s version of barnstorming.

From TFA:

"I was curious about how the overpass would affect the ground effect of the helicopter. Tragically, I have no brain for physics. But I do know a helicopter pilot, so I asked him. Apparently, at 60 kts, there wouldn’t be significant ground effect because the down flowing air off of the rotor disk would mostly be tailing off of the helicopter. Had Tamburro taken the stunt slower, the down flowing air would be more directly underneath the helicopter, creating an (obviously undesirable) effect."

That part was talking about the traditional ground effect. There isn't much study about what happens when you have a ceiling over a helicopter because that doesn't happen very often in real life. Note the discussion of the air flowing off of the blades.

I was thinking more of the low pressure zone above the helicopter being disrupted by the overpass. Apparently it wasn't too bad since the pilot managed the stunt twice without injury, but it's a possibility that would have had me crapping diamonds if I were asked to pull it off.

I feel an aircraft-on-a-conveyor-belt moment coming on.

So first of all, GE in a rotorcraft is different from GE on a fixed-wing aircraft. The fundamentals are the same, of course, but in a rotorcraft the relative velocity of the blade/wing isn't connected to the velocity of the airframe the way it is in a fixed-wing aircraft. In a fixed-wing aircraft the velocity of the wing and the velocity of the airframe are the same, of course. But in a rotorcraft things get very complicated. The upshot of all this complication is that the GE effect is reduced as the rotorcraft's airframe starts to move. The faster it goes, the less GE you have. So you might notice some differences while hovering under an overpass, but shooting through at 60 MPH the GE is effectively a no-op.

If you really want to know the gory details you'll have to read up on helicopter theory. It's way too complicated for an HN comment.

He's not talking about ground effect, he's talking about the overpass above the blades "cutting off" available air for a second.

Yes, I get that. "Reverse ground effect" was how the OP phrased it. What exactly do you think would happen in that case? Suppose the available air were "cut off" by the overpass (it wouldn't be, but let's suspend disbelief for a moment). What would be the result?

Loss of lift due to the interruption of the airflow causing the helicopter to crash into the ground at speed? Or suddenly lower pressure above the helicopter causing it to be "sucked up" into the ceiling.

I think the pilot is probably correct that the stunt could only be pulled off at a higher speed where you can outrace the low/high pressure zones the blades are creating.

And how is any of that at odds with what I said?

I think the term we are looking for would be, ceiling effect? Just recently saw footage of Fred North* flying into and out of a building, so it must be a negligible effect. * helicopter pilot for movies.

look at the underside of an overpass. it's not a flat surface, but has thick ribs; lots of room for airflow even if the blades were 1" from the underside of the bridge.

I'm in a bus under a freeway overpass as I type this. I pass under here twice a day. The underside is smooth, possibly to keep it easier to clean. I've definitely seen overpasses that are not smooth underneath, but certainly not all overpasses and tunnels have space in the ceiling.

Something a bit crazier was performed in Brazil in the 60's for a film: https://youtu.be/rSH3FVzcDSo?t=519

This reminds me the 1968 Brazilian movie "Roberto Carlos em ritmo de aventura" when a pilot flow an helicopter inside the 250m (~ 820 feet) long Túnel Novo (New Tunnel) connecting the neighborhoods of Copacabana and Botafogo.


Minute 0:57 to be more precise.

Also, you can enjoy Rio de Janeiro from above in the 60s.

What is funny you could still buy this helicopter, Schweizer 300, brand new up to 2002 as it was designed in 1955! ~Same carbureted engine with zero automation and 40L/h fuel consumption. No engine governor means pilot manually controls engine rpm, you are the ECM having to correct for air temp and pressure altitude, worry about icing and manifold pressure.

All thanks to stupid US aviation policies - recertification is so expensive nobody does it and everyone flies dangerous shitboxes despite >50 years of innovations.

There is a blog article about filming this (google translate available): http://culturaaeronautica.blogspot.com/2011/06/roberto-carlo...

Not too long after two children and an adult actor were killed on the set of the twilight zone movie.


If people survive such a crazy stunt, you congratulate them, if not, you bury them and the director. Risking people's lives for a few seconds of film is not worth it.

Take a look at what Cameron did on The Abyss:


> Ed Harris reportedly punched James Cameron in the face after he kept filming while he was nearly drowning.

The actors started to call the movie "The Abuse" while filming it. James Cameron almost died himself on another occasion. He lost his air while submerged and the guy supposed to check on him didn't do his job correctly.

There's a great documentary on YouTube: https://www.youtube.com/watch?v=3k43KHV1fDU

If I were an actor in danger and the camera man stopped filming I would be rather annoyed unless the rescue effort was assisted by stopping filming.

Do you really think you can put yourself in the headspace if an actor who narrowly escaped death from drowning?

I didn't say that he should have reacted that way, I was merely saying that I feel that I would react that way.

I would not have wanted my suffering and fear to have been wasted.

The movie industry has a long history of callously risking its own workers lives. It’s gotten somewhat better than in the 1930s, but clearly there is more work to do.

Cameron is well known as an absolute piece of shit. He brags about it, and loves his reputation.

I'm fairly sure the people doing these stunts do them because they love doing them, not because they are being forced to perform them. If people want to risk their lives for a few seconds of film, why stop them? People are into all kind of weird stuff, as long as they don't hurt others, I don't see what the problem is.

Reality is far more complex than that. Yes, stunt people tend to be a risk seeking crowd. But there’s also the risk of losing their livelihood if they refuse a director’s demands. Film business is a precarious way to make a living, and the power gulf between the director and the stunt person is massive.

History is full of “I think this is a bad idea, but I’ll do it” accounts in the after-disaster reports. People can and regularly do dangerous things against their better judgement because they were ordered to do it or otherwise pressured into it. Saying that they volunteered is to drastically over simplify the situation.

So again, they chose between doing stunts or doing something else. You don't work to do the stunts in the movie? Fine, don't work on the movie, it's not essential to life anyways so continue on with your life.

Because they chose to do stunt work as a living, they should be expected to do them in an unsafe way? Only people who are willing to do things in unnecessarily dangerous ways should be stunt people? I feel like you've got a false dichotomy here. A perfectly reasonable solution is for the director to take their staff's safety into account and not risk people's lives for a movie.

> Because they chose to do stunt work as a living, they should be expected to do them in an unsafe way?

No, did someone claim that? Stuntmen regularly decline to perform stunts they perceive to be too dangerous. It's part of their job to do risk evaluation and say yes or no.

> Only people who are willing to do things in unnecessarily dangerous ways should be stunt people?

Eh, yes, that's literally the definition of being a stuntman, you do things that are dangerous in place of someone else because you know how to manage the dangerous situation compared to the other non-educated (in dangerousness) person you're replacing for that scene.

> A perfectly reasonable solution is for the director to take their staff's safety into account and not risk people's lives for a movie.

Yes, agree! That's why there is a whole stunt team on set, not just the director and the stuntmen themselves. As a team they evaluate the stunts and stop them if they are too unsafe.

Unfortunately, every single stunt in the industry carries the risk of death, so if we want to be really safe, we simply cannot have action movies anymore at all.

The person you were replying to was speaking specifically about situations where the stunt person was uncomfortable doing the stunt and felt pressured into attempting it rather than risking their job.

Who said that it had to be 100% safe? The situations being discussed are those where someone involved felt it was being done in a way that exceeded their risk tolerance but did not feel comfortable refusing.

I guess thinking employees deserve the ability to refuse dangerous work without losing their jobs is just unreasonable coddling of people who should have to risk their lives because the director thinks doing the shot this way would be cooler.

Individual responsibility is not a thing in a nanny state.

For all their faults nanny states do generally allow professionals to do their jobs. The grandparent comment is beyond even that. They're just assuming that because they can't look at a stunt situation and assess the danger with a reasonable degree of accuracy that neither can an experienced professional.

Continue with your life, and risk losing your career. I think if you’re not accounting for this risk in how people think, then you are absolutely misunderstanding what motivates people to do what they do.

Again, people regularly do things against their best judgement because of social and economic pressure. Stunt people are not magically excluded from this process.

Tell that to the Twilight Zone victims.

What am I supposed to tell them? That they chose to be in a risky business (or their parents chose for them to be in a risky business) and in risky businesses there is a chance to get hurt?

In the accident two children were killed by a helicopter due to explosives firing too early: https://en.m.wikipedia.org/wiki/Twilight_Zone_accident

I don’t think they „chose“ to be in a risky business..

It's also noted in my previous comment "(or their parents chose for them to be in a risky business)". Since children don't have their own willpower legally or otherwise until they become adults, their parents are their guardian and ultimately responsible for them. Their parents chose to participate in something they knew was dangerous. We've always known flight in general is dangerous and when combined with explosives, extra dangerous.

From the Wikipedia link:

> All four parents testified that they were never told that there would be helicopters or explosives on set, and they had been reassured that there would be no danger, only noise.

If me and my kids were to be in a film, and the crew assured me there would be no helicopters or explosives on set, but then I arrive and they tell me to go into a helicopter and that explosives will happen next to it, don't I bear responsibility for accepting that?

You make it sound like someone forced the parents and the kids to jump into the helicopter.

They kids were not in the helicopter and the parents were told in the evening before the shooting that it will be just loud but not explosions. The directors specifically hid the kids from fire safety and wellfare officers knowing that what they were doing was illegal.

Quite honestly the entire story is sickening if you read the reports about it.

Does the director really have the expertise to evaluate how safe a stunt is? Seems like the wrong person to give that responsibility to.

I believe this evaluation is usually done by a stunt coordinator instead. This is an experienced stunt performer that finds the right person for each stunt, and communicates with the performers on any safety issues that come up.

More info here: https://www.nfi.edu/stunt-coordinator/

That’s like saying a cto should be aware and competent of the technical details of everything happening in his organization

More like the CEO.

The CEO would be the producer I believe .

How many people have died in movies?! I'd guess it is negligible. People have died during theater performances. And it doesn't get safer than a theater.

There are a surprisingly large number of them. Even more when you include serious injuries. https://en.wikipedia.org/wiki/List_of_film_and_television_ac...

A woman was shot and killed by live ammo on a movie set about a month ago.

Before turning seven or so I and my friends thought they really shoot people on set. All for the sake of the art, you know, somebody has to sacrifice to make a good movie. I remember that we argued whether it’s fake or not, and the consensus was it seemed real.

I think it is moral and ethical (and is and should remain legal) for professional adults to voluntarily consent to risking their lives in exchange for payment.

It's not really anyone's place to dictate someone else's risk appetite. Stuntmen and Stuntwomen know what they are signing up for. They are the ones choosing to risk their own lives, not anyone else.

Some people really love doing crazy shit and making wild art. I still quote Terminator 2 decades later; it is epic art regardless of whether you personally like it or not.

I think you're romanticizing stunt-people with "some people really love doing crazy shit and making wild art." Most professionals don't want to take risks with their lives, they just do it because they need to get paid.

In the vast majority of stories you hear, the actor/stunt-person dies because the director/coordinator/other staff was negligent.

It's obviously not reasonable to expect for a stunt-person's job to be as safe as office work, but "they are the ones choosing to risk their own lives" is bs.

> Stuntmen and Stuntwomen know what they are signing up for.

_Everyone_ knows what they're signing up for, so the only interpretation I can make here is that you believe there should be no protections for employees.

> Most professionals don't want to take risks with their lives, they just do it because they need to get paid.

This is making it sound like they're forced to choose their profession and take risks. If I were to guess, the truth is probably completely the opposite - the only reason they got into it was because of the thrill of stuntman work.

Seems like if you just wanted the side shot and not the straight on shot as well, you could of just pulled the helicopter on a low trailer using the median partition to block. The straight on shot would of required deleting the trailer which may have been difficult but certainly would look better than any CGI attempt

Or just CGI. But as Cameron himself puts it (copied from the article linked in parent):

> we could’ve run a CG helicopter under the freeway overpass, but it was so much more fun to do the real thing.

I'm guessing the same applies from pulling a helicopter on a trailer. It's simply not fun.

This thread fills me with horror. I admire that movie, it's incredible, especially when I know it's just safe reenactment of a scene. But it's an uneasy feeling to know that there's some serious risk involved.

Yeah that was completely obvious. Check out the scenes that actually were CGI. They are trivial to spot, with a 2021 eye and DVD level quality media or better.

The whole shot that "went through the computer" has a completely different feel to it than scenes that don't.

Compare 3:38 and 3:44 in https://www.youtube.com/watch?v=bhL8WlDHKaY.

It looks like it was shot with a completely different camera, or going into a mini-dream state.

I mean it's extremely well done. It holds up so well even... jesus christ 30 years later?!

But it's extremely obvious which scenes got the CGI treatement.

Oh, and if this is one of those "once you can see it, you can't unsee it" things: Sorry about that.

> It looks like it was shot with a completely different camera

First it's a one single still frame up until the full closure of wounds, may be it's the reason for a dream effect that you feel. It was shot on the same camera as other shots from the John's point of view in this scene, but processed differently, as it went trough film scanner and laser film recorder.

Exactly. And you can tell, is my point.

Same with other CGI scenes.

Amazing for 30 years ago, but very not possible to confuse with practical effects.

I'm not entirely sure what you're referring to. I'm looking at the still frames in 1080p.

I see the bullet holes absolutely don't match up. At 3:38 they are not as deep and wide as in 3:44. But also the actors position is different especially the mouth (closed in the first, open in the other). Continuity errors.

The morph effect is somewhat cringy but to be fair, they always were. At least in DS9 and Terminator 2 there were in-universe explanations why they existed.

There's also something weird going on with the eyes to the T-1000 after the morph effect has finished. Not sure if that was intentional.

The part that jumped out to me in that CGI was the mannequin. The bullet hole morphing is so obvious that in a way it's less objectionable.

100% of the frame has a different quality in the clip that went through the computer.

Like they smeared something over the lens.

Also check out the "goes through the bars" shot. Once you see it, you can't unsee it.

Look at things like color grading, and things that look like "upscaling" (but aren't).

Maybe it's that why I don't see it or at least it doesn't spring out to me that much.

Because it's like an upscaling effect that also could be an overuse of HD makeup.

I've seen that in contemporary movies or series (or badly configured HD TVs) and it doesn't have a dissociative effect on me seeing it in this movie because I just don't remember how movies from that era shouldn't be looking like that.

We're all on this website reading this comment because doing it all inside the computer in software is simply...

easy money.

If you're into this kind of thing, the Corridor Crew youtube channel will likely be right up your alley: https://m.youtube.com/c/corridorcrew/videos They have a long running series of "VFX Artists React" that are fun and lightly educational.

Do note that while some of them go into quite a lot of detail with good visual explanations, quite some videos have also become watered down and light on the technical content, weighing much more heavily on the "react" aspect than the "VFX Artists" aspect.

An absolutely fantastic series, so much so that they're getting attention from industry professionals, who are coming on the show.

Reminded me of this (not sure if the channel is related or not)


^ you want skynet, because this is how we get skynet. :-)

Yeah, that is their "main" channel, same group of people. The "Crew" one started as like a behind-the-scenes I believe.

> not sure if the channel is related or not

YSK that you can click on the Channels tab of both of these YouTube channels and see interlinks that they added.

It's the same guys. They're an indie vfx/short-film production house.

I recall debating this with a dude who was into makeup and masks; he pointed out the blooms moved like rubber as the character runs; seemed pretty conclusive.

I love how prosaic and simple some of their practical effects are; but the real accomplishment to me is how hard it is to pay attention to those details, to notice them; when everything flows so smoothly into captivating scenes.

Reminds me of one New Years Eve party where Aliens was playing on the TV and a friend walks past and says "Watch Bishop's severed body now as he grabs Newt to stop her getting sucked out the air lock...", and clear as day you can see the lower part of the actors body come out a hole in the ground when he grabs Newt. I had probably seen the movie a dozen times before but never noticed it.

In a similar vein, just after having watched X-Men 2 in a cinema a (different) friend who did CGI animation for a living pointed out that during the scene where Magneto moved the bridge, you can see that the shadows on the water don't move with the bridge. His theory was that the special effects team were likely completely aware, but looked at the amount of effort required and hoped no one would notice. I never noticed the problem myself until it was pointed out.

I guess back then when special effects were monumentally difficult and computers weren't as capable as they are now, directors often relied on the audience's attention being elsewhere at that moment and thus being less likely to notice.

> and clear as day you can see the lower part of the actors body come out a hole in the ground

Screencap: https://avp.fandom.com/wiki/Aliens_goofs?file=A2blactorholep...

(fixed in the bluray)

from https://avp.fandom.com/wiki/Aliens_goofs

There's a couple of those.

The "caseless" ammo assult rifle is obviously ejecting spent casing in some scenes.

The rotating light on top of the mecha suit loader is smashed when the queen drags it down into the airlock but is whole again once down in the pit.

> directors often relied on the audience's attention being elsewhere at that moment

Your comment reminds me of this Raiders Of The Lost Ark scene:


I have to admit not noticing this for decades.

I remember rewatching T2 on computer after seeing stan winston studio showing props. The movie shows a lot of these physical props for very small frame numbers. Very hard to notice if you slow things down but at normal speed it just flows through the context of the scene. It's a minimal amount of detail that your brain needs to have to accept continuation. Probably a byproduct of the time.. can't CGI all the things so you have to optimize and find the best bang for your bucks in every dimension.

Hey. It's only hard because most movies are still a blurry 24 FPS mess. Look what happened to the Hobbit movies. I had a friend complaining more FPS suck because he was able to see makeup and everything else wrong with the movie. That's why certain directors don't want any advancements in this area. It will make their job a lot harder than it is now

The Hobbit just looked bad, IMO. They seem to have lost their heads with color grading.

I remember seeing a video where Jackson was raving about how they can relight shots, showing a pretty uniformly lit shot in Hobbiton. They went on to try to add directional light, shadows etc. using masks, and the image just got worse and worse until it was an inconsistent mess. But apparently they liked that.

And don’t get me started on how they smoothed the dwarves’ skin out so they look like glamour models. I really can not fathom what they were thinking.

I read recently, that the Hobbit films have been re-colour graded to align them with the main Lord of the Rings films in the latest UHD 4k box set.

>They went on to try to add directional light, shadows etc. using masks, and the image just got worse and worse until it was an inconsistent mess. But apparently they liked that.

You'd be surprised how much the average human can enjoy fake lighting and fake elements in photography/videography.

For anyone with the slightest lighting knowledge though it looks horrendous.

It's _mostly_ due to how human vision works really. Our visual system is good at focusing on certain parts, and the rest just gets elided unless you specifically focus on it. VFX are magic tricks put to film.

24 fps can have good motion clarity with a strobed image. Theatres used to have shutters. I don't think DLP does, but I may be mistaken.

DLP response time is insanely fast (~2 µs) because they're basically using PWM to modulate the brightness, but instead of each pixel getting a specific pulse width they're essentially sending bit-planes where the per-bit-plane on-time corresponds to the bit-plane's weight.

Sample and hold makes motion clarity worse. You only help motion clarity two ways: shorter frame duty cycle and higher frame rate.

Absolutely, the bit-plane modulation implies that the duty cycle more or less corresponds to brightness, so you only get close to 100 % duty cycle from a DLP when the output is close to full brightness. Though there is some additional trickery here - at least the colorwheel DLPs actually modulate the light source in sync with the bits, so for an 8-bit output the LSB duration is not actually 1/127th the duration of the MSB. There could be more tricks - it's probably not necessary to actually have all the LSBs around every time you show the same frame, which leaves more time to reduce the duty cycle.

They are not a blurry 24 FPS mess. They are an efficient 24 FPS sweet spot

24 FPS used to be a reasonable sweet spot back when movies where slow-paced and high frame rates made for huge film reels (not to mention the lighting requirements for early color film).

Today there are no technical problems or impracticalities with high frame rates, and modern cinematography is hitting the limits of what you can do with 24 fps (and sometimes crossing them). 24 fps has become a local maximum because that's what we've done for decades, but it is also a blurry mess.

There literally are some impracticalities, though. For a Blu-ray player to support 60fps at 1080p, it's got to support level 4.2 of the AVC specification. Many (most?) players do not do this (and in fact some older models supported only 4.0), because it's more costly to include a decoding chip that can handle the bitrate needed for the uncompressed video stream.

Furthermore, much of home entertainment is streamed these days. I imagine introducing 60 fps on top of 4k would increase bandwidth requirements beyond what many consumers have access to. It may also be more than Netflix et al want to pay for, so one imagines they'll want to tack on a surcharge as they did for UHD.

The jump from HD to UHD quadruples the number of pixels, the jump from 24 fps to 60 fps only increases framerate by a factor of 2.5. Not to mention that the panels in any old HD TV will happily work with a 60 fps signal via HDMI, while the upgrade to UHD requires a new TV.

Yes, more frames means more data, but it creates fewer problems than the 4k/UHD upgrade everyone willingly goes through, while doing more to increase fidelity.

Right, but everyone agrees that UHD is an enormous improvement worth buying new hardware for. You said "there are no technical problems or impracticalities" to high frame rates, but 2.5 times the uncompressed bitrate is a technical problem, and so you have to have a situation where people are willing to pay for an upgrade. Most people seemingly don't care enough.

> the 4k/UHD upgrade everyone willingly goes through

I looked for 1080p screens when I last bought a TV and couldn't find any. 65", and wanting fairly quality, but no 4k sources so why do I need 4x the pixels? (Sadly, the week after I bought what I bought, I happened across some 65" new in box Plasma TVs at auction... Oh well)

Why is that an impracticality? TV productions are pretty much all still at 50/60Hz, they get syndicated to Netflix and streamed or sold on Blu-Ray and watched that way every day.

There's nothing new that has to be changed to support this, most of the content people are watching is already HFR.

This surprised me and so I wanted to check it out, but I think you're incorrect about this. I checked a number of shows shot in the last 15 years (The Expanse, The Wire, Schitt's Creek, Squid Game, Star Trek: Discovery, Foundation, New Girl, Community, Westworld), and every single one of them is either 24 fps or 23.976 fps in the versions available for streaming on the web.

Pretty much every TV documentary or larger production (e.g. Eurovision) is still 50/60Hz though.

It's only the narrative based shows that try to be "cinematic" which are at those low fps rates.

But the technology is there, it works just fine.

Large fast panning moves in a movie theater make my head spin, it's like a damn stroboscope. 24 FPS is not a sweet spot for everyone, it just happens to be what we got stuck with.

In many movies, I just can’t see motion and the blurry mess becomes entirely impossible to see. Which is why I end up reading plot summaries on wikipedia instead.

24fps is far too low.

The blurry mess of fast CGI is NOT down to the 24fps problem. As the action is so fast, it's faster to render complex scenes in lower detail CGI as the eye (mostly) can't see it.

However if you pause the action, you can clearly see where the rendering shortcuts have been made (The Transformers movies are a good example of this).

I'm not just talking about CGI, even e.g. taking my own a6300 and recording a moving scene has this effect (even with the 180° shutter rule).

It's just as choppy as 24Hz at 1/1600 shutter, but in traditional film, it's choppy and blurry due to the slow shutter, so I can barely see anything.

For quite se time I'd run upscalers to pre process all movies I was watching to 60Hz or 75Hz just because it was noticeably easier to watch.

The 24Hz/fps thing is a legacy from original film where it was the sweet spot that created the effect of a moving image without too much flicker, unfortunately you seem to be one of those people who are sensitive to this frame rate.

The argument is that the general populous are used to the 'film look' of 24fps, that anything else diminishes the impact of watching a film. Films that have tried high fps (The Hobbit for example) received negative reviews for picture quality; people just didn't like it.

Also, when increasing the framerate, you should try to stick to a multiple of 24 (i.e. 48, 72, 96) to prevent judder when the shot is panning a scene.

It was certainly good enough for millions of other people. What makes your eyes super sensitive to fps?

I don't know? Sometimes I also have issues where the perspective correction stops working and I see the image of each eye individually, with blind dot and perspective issues at the same time. But luckily only if I'm very tired.

Maybe something's wrong with my mind? But 24/25Hz, even with the correct shutter speed, is noticeably causing strain to process, while e.g. old school TV productions at 50Hz don't have that issue.

So you're able to interpret individual frame transitions at 24 frames a second? You've been unable to watch the vast majority of movies created in the last century as they strain your eyes? Maybe you need to get your bionic eye recalibrated?

Maybe gaming? When you play a lot of FPS in 120Hz then action scenes in the movies just look like blurred stuttering mess.

And stereoscopic 24fps look twice as bad.

Why would playing high frame rate games alter your vision? Is your brain reprogrammed when viewing image transitions at frame rates so high that they cant even be noticed.

Further research needed probably. But it doesn't seem to be far fetched to guess that what you do influences how you work.

I think you can notice high frame rates. Difference between 60 and 120 fps is striking. Even between 90 and 120. And everyone I was showing my phone that does 60, 90 or 120Hz could see it.

Conversely I don't see much difference between 120 and 240 fps.

Because it provides a different perspective, figuratively. I think that might also be a reason why HFR movies feel weird for some people, it's very much a comfort zone.

Does a lower frame rate actually mean more blur? The frame rate gives an upper bound for exposure time but not a lower bound.

Film is single-exposed for recording but has been multi-exposed for projection for >100 years, because a 24 fps projection makes everyone throw up (this is a first hint to those "eye can't see more than 24 fps anyway" folks).

To get a "natural-looking" amount of motion blur the exposure angle is generally around 180°, which is ~1/48 s exposure time for each frame. If you use a small shutter angle (short exposure time) you get less motion blur and things start to look -very- choppy (another hint for the "eye can't see more than 24 fps" fraction). Generally this translates to moving things being almost always blurry. This is also why pans are usually so slow in cinema films, because if you'd pan faster everything would just be a blurry mess. Another factor here is that film is not very sensitive - so short exposure times are kinda out of the question anyway. Look at e.g. The DaVinci Code, most of the darker scenes were shot on 500 ISO film and it's already very grainy and soft.

Apparent motion blur is also worse on sample & hold displays (LCDs, OLEDs etc.) compared to strobed displays (especially CRTs). Viewing 24 fps content on most displays at their native refresh rate (~60 fps) requires on-the-fly frame rate conversion, which is necessarily of poor quality. If you ever watched popular Youtubers and wondered why the heck their videos are so incredibly janky, it's not just because the YT web player drops frames like crazy, but also because you're viewing 24 fps content on a 60 fps display (they all shoot in 24 fps for the "professional look").

The reason why cinema is stuck on 24 fps even in the digital age is mostly because people have been culturally programmed to associate janky motion with quality film, and fluid motion with cheap TV production. That's all. It's an objectively far inferior format that was chosen at a time when film was quite expensive per meter, and so you'd wanted to use as little film as you could get away with. The only advantage a low frame rate has (today) is that you get a stop more light because the exposure is twice as long at preferred shutter angles. Digital cinema cameras already have base sensitivities of 500, 800 or 2500 ISO though - much faster than high quality film ever was

> The reason why cinema is stuck on 24 fps even in the digital age is mostly because people have been culturally programmed to associate janky motion with quality film, and fluid motion with cheap TV production. That's all.

I kinda assumed it was because it was too expensive and difficult to produce higher frame rate video that was of comparable quality. I have yet to see a high frame rate movie that didn't leave me forced to pay more attention to the bad film props and questionable acting than to the story line.

That’s like saying the only reason rock music is stuck on tube distortion is because of cultural programming, and hi res solid state amplifiers are objectively “superior” in some way.

We’re making art here. The format that people find the most compelling isn’t inferior.

In general, films are shot with a slow shutter speed because the low FPS would otherwise create a strobe-like effect. Unless that effect is being sought (famously, Saving Private Ryan), blur is preferred.

Isn’t a strobe light effect due to how long you project each image, not how long you expose it for?

Maybe strobe isn't the best word for me to use. Choppy? Whatever you want to call it, the result is kind of nauseating.

Without blur such low framerate looks very bad. Look up some gaming footage on youtube that's 24fps. It just looks like a slideshow not the motion.

> he pointed out the blooms moved like rubber as the character runs; seemed pretty conclusive.

Is it? If we could build such robots, they would be plastic. Any holes shot into them wouldn’t be rigid, and would deform when they moved.

Or are you saying “moved like rubber” obviously looks different from whatever way the stuff this robot is supposed to be made of would move?

"Moved unlike CGI" is what i meant.

OT a little, but I learned today that the scene from Live and Let Die where Bond runs across the back of a bunch of crocs wasn't faked. As a child I'd always been told that the stuff you see in movies is not real, but it turns out that sometimes it is.


Well, not sure you can call it "real" really, the crocodiles were kept in that specific place, it was attempted many times and so on. What people wanted to convey is not that it's not actually real life, but it's most likely faked real life, just like that Bond stunt.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact