Another "How is this done?" thread... (Ken Block's Gymkhana Five)

Hawk Teflon

Hawk Teflon

Veteran
Apologies for not putting this in an appropriate sub category, but I honestly don't even know which category it would fall in.

Here's the video:


A couple of times throughout (2:50, 5:20, 5:40) there are portions where time stops and the camera slides to a different position (seemingly). I know this was done in The Matrix years ago, but that was a very controlled environment (Carrie was in a pin-point location). But this has a lot more variables to it. Is this done in camera (with a slider that's hella fast), or in post?
 
If you freeze it at 2:50, then 2:51 you can see how the frame changes to a much wider angle suddenly and looks more fisheye. Looks like post not a slider.

Same thing at 5:21 and 5:22. If you freeze it and look at the frames, there is an abrupt change that would not be caused simply by sliding the camera.
 
Well why would you say this isn't a controlled environment? This isn't a race, and clearly a demonstration. They have enough money and time to close down these streets and do these many multi-cam shots. Not sure why it would be so crazy to set up 5 still cameras in a row to take a shot at the same moment then stitch it together. They don't go very far either, just a few feet at most.

With what they most likely had, I wouldn't be surprised if they didn't test it several times, took multiple takes, and/or had a motion trigger set up.

Or it could have been post, though I caught some new shadows in the frames between, which makes me suspect it wasn't 'morphed' into another frame.
 
Jon Starr said:
Well why would you say this isn't a controlled environment? This isn't a race, and clearly a demonstration. They have enough money and time to close down these streets and do these many multi-cam shots. Not sure why it would be so crazy to set up 5 still cameras in a row to take a shot at the same moment then stitch it together. They don't go very far either, just a few feet at most.

With what they most likely had, I wouldn't be surprised if they didn't test it several times, took multiple takes, and/or had a motion trigger set up.

Or it could have been post, though I caught some new shadows in the frames between, which makes me suspect it wasn't 'morphed' into another frame.
Click to expand...

I agree, looks like multiple cameras either set up at the same time, or two different takes then blended together in post.
 
Jon Starr said:
Well why would you say this isn't a controlled environment?
Click to expand...

I'm not saying this wasn't a controlled environment, I'm saying the beginning of The Matrix was a VERY controlled environment. They were able to move CAM around to pinpoint accuracy before she jumped. The car wasn't set up exactly into frame. Yes, I'm aware he's an incredible driver, but with The Matrix, I remember them talking about no room for mistakes in one of the interviews. Granted a crapload has changed since then (what was that, like 1999?), so it's very possible that's how they did it. I just didn't know if 12 years of technology had changed the approach.
 
My guess is an array of gopros. If you pause the clip at 5:49, you can see a disturbance in the force, or the curb, right below his front bumper. You can follow a line of them, going left to right, ending a little after the sign post that gets cut off by the tree.
 
Well the simplest way to do this would be what everyone else has said, an array of gopro's. You COULD do it in post, but it would be much more difficult (and probably expensive) so I doubt that's how they did.

Yay gopro's!
 
I'll "Go" with GoPros, too.

Funny thing is the "bumpy" look, as if sliding over an uneven surface, but I guess some of the GoPros were higher or lower by 1/2"?

Quite a bit of skid marks at each location- looks like lots of practices and takes.

And following "safety first" (what was the insurance rate???), you gotta love the Coast Guard approved life vests on the barge...

Love the GoPro bouncing off at 3:37!

Ahh... The Streets of San Francisco, 2012...

Here's a teaser of the car jumping with the "Matrix effect", they jump back and forth through the shot...

 
Last edited:
Hawk Teflon said:
I'm not saying this wasn't a controlled environment, I'm saying the beginning of The Matrix was a VERY controlled environment. They were able to move CAM around to pinpoint accuracy before she jumped. The car wasn't set up exactly into frame. Yes, I'm aware he's an incredible driver, but with The Matrix, I remember them talking about no room for mistakes in one of the interviews. Granted a crapload has changed since then (what was that, like 1999?), so it's very possible that's how they did it. I just didn't know if 12 years of technology had changed the approach.
Click to expand...

Which shot are you talking about exactly from the matrix? Mind you they had hundreds of cameras going off a fraction of a second after the other, so pin point accuracy is correct.

Something like what you see in this video, is less about pin point accuracy as all cameras simply go off at the same time. OR as everyone here said, it could just be gopros. You simply record video, then chose the frame that matches the other (ten times easier). The process is very simple. The only accuracy is really to get the guy in the frame, which with a wide angle lens is no problem. Plus they probably had a few takes.
 
I'm referring to the one with Carrie-Ann Moss in the basement at the beginning.


Skip to 1:30.

Ok, so let's say you've got some GoPro's (or other WA cameras) running together in a tight knit line. It looks really smooth and not how I would expect that to look (almost stop motion-y).
 
All the time slice stuff, the action stops, we rotate, then the action starts again. But in this clip the car doesn't stop moving.
 
Hawk Teflon said:
Ok, so let's say you've got some GoPro's (or other WA cameras) running together in a tight knit line. It looks really smooth and not how I would expect that to look (almost stop motion-y).
Click to expand...

The original video is not exactly very smooth and has a few bumps to it, it's just very fast (which means very few angles). It's not that hard to get a few cameras in line, especially with a rig, or just attaching them to a bar. Bring a ruler and a level. Not very hard IMO. In fact I'd say not all their cameras were perfectly level.

If you're talking about the in between frames, it's the same as they did for the matrix. You morph those missing points between the cameras and create new frames to make it smoother. If a camera is off, you do as you would for a shaky camre and stabilize it. However the video you showed goes so fast I doubt they even needed to do that.

Been seeing kids in their parents basement with after effects geting nice 'bullet time' effects for years now.
 
Last edited:
They have been doing a lot of GoPro stuff in the XGames this year including this effect. Pause this clip around :57 and I think the rail you see in the top right of the screen is the rig with multiple GoPros. I think they refer to it as "Bullet Time".



HERE is a link to an article about it.
 
Cole said:
They have been doing a lot of GoPro stuff in the XGames this year including this effect. Pause this clip around :57 and I think the rail you see in the top right of the screen is the rig with multiple GoPros. I think they refer to it as "Bullet Time".



HERE is a link to an article about it.
Click to expand...

Same thing. Motion stops, then we rotate. The original post has the car moving while we rotate.
 
You must log in or register to reply here.
Back
Top