How did they do the camera split in 'The green hornet' 2011?

I've been intrigued by that impressive scene from the movie since I first watch it. How did they pull it off?



Sorry for not being descriptive, and therefore not helpful to you. What baffles me is like this: near the start of the scene, 1-eyed man goes into a massage parlor and talks with the 3 women there. Right at that point, the screen very smoothly divides into 2, with 1 following him and another following the 2 bitches. That process continue until we have like 16 small split forks. Surely 1 camera can't record that sequence, so how did the filmmakers pull it off at the stage? Stacking 16 cameras?

Embedded it for you.

I don't know how it's done. But they made good effect of it in the 24 series (2001-2010), starring Kiefer Sutherland. Nice technique, as long as it's not over-used.

An army of camera men?

You can do it in a movie editing program like Premiere Pro. Just put all the pieces of footage in the aspect ratio, and then split them all together how you want, if that helps?

I haven't watched that series. Could you give a link to the mentioned effect?That sounds probable. But I can't really put my finger on how they do it in the real scene. That Green hornet scene seems smooth, which means whatever get split, was divided from a dual camera supposedly handled by 2 cameramen standing next to each other. If that assumption is true, then working backward from 16 small windows at the end of the scene to the 1st, we come to the conclusion that there are 16 cameramen crowded at that frame. Not even mention about the space for them to squeeze in, 16 parallel cameras or stacked 4x4 would produce considerable difference in the angle of shooting, therefore a skewed perspective & uneven pictures (not smooth transition)?I think if we have 16 perfect footage then editing them should not be a big problem. My question leans more on how did they pull it off on-scene...

Here is a good link with discussion: https://24.fandom.com/wiki/Split_screen

There are loads of references in web searches.

If you haven't watched any of the 24 series, you really must. There were technical innovations, and Sean Callery's music scoring was superb. The show was one of the best on TV from 2001-2010.

~Doc

I would guess that having a second camera standing on a mark waiting for the first camera to walk up and stop next to that mark would be enough to give the editor two source videos that share a similar angle on the subject to the blend them in post-production. If that was the case, then you only need two cameras at any given time. I can't imagine this shot being done as one long continuous shot. I'm sure there were cuts and digital transitions to hide those cuts.

No idea. I wasn't on set obviously, but the only reason I can think of to have a single long shot being split like this would be for a technical bragging right type of thing. It still wouldn't be practical at all and opens up all kinds of risk with so many variables.

I had to watch the clip a few times to get what you were asking. I thought at first you just meant splitting the frames and running multiple clips on one screen. I think you're talking about the effect of a single shot appearing to duplicate, but then track along two different actors along two different paths, but having started from what appear to be the same scene and frame. Nice effect actually. I missed it at first. A few times!

__________________
"My Dionne Warwick understanding of your dream indicates that you are ambivalent on how you want life to eventually screw you." - Joel

"Ever try to forcibly pin down a house cat? It's not easy." - Captain Steel

"I just can't get pass sticking a finger up a dog's butt." - John Dumbear

Oh well during shooting I would assume they would shoot it just like another scene, and then cut out the shape they want, accordingly in the editing, if that's what you mean?

I don't really know why, but 24 seems to not be able to get into IMDb's top TV list, despite outperforming some of its bottom entries.Oh, now I get it. The bad is on my part to make an OP with just a few sentences and without any description of what I was looking for. Should not make this mistake again. I'll edit the post dutifully.I highly doubt this approach would produce good footage. The 2nd, waiting cameraman can't just wait there on the spot because he'll surely be in the frame. So he'd have to hide in some corner and when the time is coming, rush out to put his camera right next to the 1st guy. This 'combining' approach will risk having shaky pictures at the joint moment because there's a high chance those men will bump into each other during the rush. Contrary to that, the 'splitting' approach is easier to have stable images, but instead will face other challenges I posted earlier.No, that scene can't be shot in a normal manner IMHO.

Oh master @Yoda, I need your help again. I tried to edit the OP & add descriptions, but can't because the "25 posts to do links" problem pops up again. So, could you please add this paragraph to my original post?

Sure.

Oh, why can't it?

Well, there is a whole set and environment behind the camera. The camera doesn't literally have to wait on a mark as long as it's where it needs to be when it needs to be there. There's no reason to believe it will be so clumsy in a worst case scenario especially considering it would have thoroughly been planned.

If not that, then maybe green screen start and end points. The first shot would be of the women around the table talking as if the guy is standing there, then follow her out to the next scene split and cut. Later, film the guy in the same room but a screen erected where the table would have been. Have him speak as if the women were with him then follow him out through his exit until the next split and cut.

Mask the green screen in post and merge the two. Two separate shots blended to look like one at first, that will actually play as two.

Repeat the set up for each split and you could probably do it all with one camera.

I've made some arguments against it in this thread. Or, you can try putting yourself in the filmmakers' shoes and explain to us how would you make that scene?I'm afraid I don't understand this paragraph fully but from what I get from it, the scene appears even harder to pull off. Why? You can see that there are several times when the camera turn 180 degrees and there's no set appearing on-scene.Ah, this one is yummy. It may explain how they did it, but I see 2 problems:
1. The camera follows the 3 women and then the 2 girls to the next split (0:27) and cut (which means no more shooting, right?). Then how can the director make a smooth transition between this mini scene and the one where those bitches are divided (0:27 onward)? Having them do it again will inevitably cause difference in body movements and positions that the audience will notice.
2. Green screen is usually used in a small closed space. Doing it in a massage room is feasible, but there are scenes where the split happens outdoor with big view. Under those conditions I suspect the effect will suffer severely with noticeable glitches.

Sorry for not going into it more. I just think that as long as there is enough space around the actor in the framing, and what the actor is doing makes sense, than you can definitely crop around the actor and have no problems. Just shoot the scene as if it was a normal 1.85:1 scene, and then crop as necessarily when creating the multi-split screen, if that makes sense?

I think that in with the two women being split at 0:27, the woman in the left shot, her arm is rotoscoped into the right shot for a brief half a second or so, and that used rotoscoping and not green screen I think.

"Was there a shot or a sequence that prevented you from sleeping?
I generally sleep fairly well, but the closest sequence to have that effect would have been the split screen sequence in which the action within the shots divides as the frame splits allowing the camera to follow the diverging paths. This continues until there are fifteen or sixteen separate cells, each peeling off a bit of the narrative. This was such a fun sequence to do, but there was a lot of room for error. The individual sequences were shot quite well but required anywhere from 2 to 13 separate plates to produce a seamless split. Additionally, there were re-edits, and growing time constraints so every element needed to be refashioned, re-timed and re-married together."

http://www.artofvfx.com/the-green-ho...cis-hollywood/

Also more ideas. I'm sure there were tons of layers playing out and overlapping along with every cut you can imagine.

I didn't find a satisfying definition on the net, but is 'rotoscoping' the act of adding unreal elements into frame(s) manually/digitally?Woa, thank you so much, @ynwtf! I tried to find behind-the-scene videos of the making of Green Hornet on YT, but ended up with only videos of Kato brawling and fancy weapon shootings. My searching skill still needs to improve. BTW, the added video about cutting cheats is great, although I can't really point out any of those 4 uses in the Green hornet clip.
So it turns out that I was interested in the most difficult technique of all in that movie. What some of you suspected is true, they followed the traditional approach of repetitive shooting many takes to find the most seamless connection among a sea of plates. On this issue, I think the last mini-scenes at the end of the video were easiest and only required 2-3 copies. We should tip our hats to the bitches because theirs are longest and got most forked, so the mentioned '13 separate plates' must have applied to their shooting.
Now that I know how they did it, I've watched the clip again, this time followed every fork from its start to the end. There are actually times when the pace in a fork is suddenly sped up and looks unnatural. I guess no matter how hard you tried, how many hours of re-timing and the insomnia it induced, pulling off a perfect scene out of this insanely difficult task is still a bit out of hands.

Search YouTube for "hidden cuts" for some good examples in film. There's a vid of some guy doing home-made cuts around his house that helps break down what all may be happening.

There's one showing clips of The Revenant and all the cuts and layers playing over each other.

Just know a LOT can be done digitally. I wouldn't be surprised if in any one of the frames from your original video question didn't have 5 or more individually shot layers (both live action and digitally produced) all stacked and superimposed on top of each other to look like a single take. Like the first two women that walk out of the room together then split off. I wouldn't be surprised they too were shot separately. Or the guy leaning against the back fender of the car talking to the guy inside the car. The guy was probably leaning on a green prop that was digitally removed to make it look like he was leaning on the car when superimposed.