How was this camerawork achieved?
November 6, 2014 3:46 PM Subscribe
Please watch the sequence running from 32:24 to 33:19 in this episode of the BBC's Human Universe (which may be available only in the UK). It appears to be one unbroken take. My question is this: where was the camera positioned at each stage throughout that sequence?
It appears to start with a camera bolted to the passenger door, which miraculously exits the car and the transforms itself into a steadicam operator walking alongside the presenter as he moves down the road. Finally, it seems to transform again, this time into a boom which can lower the camera smoothly through the second car’s sun roof.
Perhaps I'm underestimating what a skilled cameraman with a steadicam strapped to his chest can do, but it strains credulity for me to imagine him struggling over the stick in the first vehicle and lowering his camera through the sun roof of the second one with no visible shake at all. Equally, I find it hard to believe even the finest computer trickery could stitch together two or three different takers so invisibly in the ediiting suite.
So, please, someone who understands this stuff: enlighten me. How was it done?
It appears to start with a camera bolted to the passenger door, which miraculously exits the car and the transforms itself into a steadicam operator walking alongside the presenter as he moves down the road. Finally, it seems to transform again, this time into a boom which can lower the camera smoothly through the second car’s sun roof.
Perhaps I'm underestimating what a skilled cameraman with a steadicam strapped to his chest can do, but it strains credulity for me to imagine him struggling over the stick in the first vehicle and lowering his camera through the sun roof of the second one with no visible shake at all. Equally, I find it hard to believe even the finest computer trickery could stitch together two or three different takers so invisibly in the ediiting suite.
So, please, someone who understands this stuff: enlighten me. How was it done?
I also can't see it, but is it possibly done by drone?
posted by brainmouse at 3:54 PM on November 6, 2014
posted by brainmouse at 3:54 PM on November 6, 2014
The camera movement, with the occasional jerky lens but overall smooth movement, is similar to a lot of drone footage I've seen. Perhaps a drone passed off from the passenger seat of the first car and then recaptured through the sunroof of the second vehicle?
posted by JackBurden at 6:41 PM on November 6, 2014
posted by JackBurden at 6:41 PM on November 6, 2014
I was able to see it using the Hola Chrome extension set to the UK. It's pretty clear to me that the camera is handed a few times between cameramen. The first one is in the cab of the pickup and hands the camera to the second one just outside the pickup. The third one is sitting in the passenger set of the car and is handed the camera through the sunroof. All shooting is framed so the other cameramen are not seen
posted by ShooBoo at 9:34 PM on November 6, 2014
posted by ShooBoo at 9:34 PM on November 6, 2014
I work in vfx, and I have stitched together different filmed takes many times. It's not done in editing, but using compositing software - it's like photoshop for moving pictures. As long as the shot is planned out carefully, it is possible to move from one camera view to another. Unfortunately, I can't see the video, so I can't say if that's how the shot you are asking about was done that way. The finest computer trickery can do it though!
the biggest stitch-up I have had to do starts at 1:18 here. the shot goes from an indoor crane shot, to a helicopter flyover from colombia university, over manhattan, into a house somewhere south of brooklyn. (? I think)
There are a few obvious transitions, like when the shot moves from indoors to outdoors, but there are also at least 7 - 10 cuts in the flyover; it had to be filmed in imax, so they couldn't load enough film into the camera to last for the entire trip. I had to get all the different takes, find where they lined up the best, then reposition, colour correct and sometimes warp the ingoing and outgoing takes so I they would match up and I could dissolve between on and the other seamlessly. The image also had to be stabilized a lot to give a nice smooth motion and not show the shakiness of the helicopter.
posted by 5_13_23_42_69_666 at 9:54 PM on November 6, 2014 [2 favorites]
the biggest stitch-up I have had to do starts at 1:18 here. the shot goes from an indoor crane shot, to a helicopter flyover from colombia university, over manhattan, into a house somewhere south of brooklyn. (? I think)
There are a few obvious transitions, like when the shot moves from indoors to outdoors, but there are also at least 7 - 10 cuts in the flyover; it had to be filmed in imax, so they couldn't load enough film into the camera to last for the entire trip. I had to get all the different takes, find where they lined up the best, then reposition, colour correct and sometimes warp the ingoing and outgoing takes so I they would match up and I could dissolve between on and the other seamlessly. The image also had to be stabilized a lot to give a nice smooth motion and not show the shakiness of the helicopter.
posted by 5_13_23_42_69_666 at 9:54 PM on November 6, 2014 [2 favorites]
If you don't believe in computer trickery being able to do this, you must watch this utterly, utterly brilliant scene, which is stiched together from six different shots, filmed in four different locations to boot.
posted by Pyrogenesis at 12:56 AM on November 7, 2014 [1 favorite]
posted by Pyrogenesis at 12:56 AM on November 7, 2014 [1 favorite]
I don't think the camera is bolted to the car door at the beginning of the shot - it bounces around too much. I suspect what you're seeing is simply the result of a cameraman (person? :) ) sitting in the passenger seat, then exiting the vehicle through the drivers door to follow the presenter down the road before lifting the camera over the next vehicle to pass it down to another cameraman through the sunroof. Modern digital video cameras even for professional use can be small and light - in bright lighting situations like this shot you don't need a huge lens & sensor to capture all the available light so I bet a small rig with counterweights would be all you'd need to take this shot - no need for a full steadicam setup.
posted by pharm at 1:57 AM on November 7, 2014 [1 favorite]
posted by pharm at 1:57 AM on November 7, 2014 [1 favorite]
Response by poster: That's fascinating - thank you everyone. And thanks for the other examples of ingenious long takes too.
My own guess is that ShooBoo's explanation of the three cameramen carefully handing the camera from one to the next is probably the correct explanation as far as that Human Universe clip is concerned. That's what I (in my position of almost total ignorance) find most convincing to envisage here anyway.
posted by Paul Slade at 2:16 PM on November 7, 2014
My own guess is that ShooBoo's explanation of the three cameramen carefully handing the camera from one to the next is probably the correct explanation as far as that Human Universe clip is concerned. That's what I (in my position of almost total ignorance) find most convincing to envisage here anyway.
posted by Paul Slade at 2:16 PM on November 7, 2014
Response by poster: I've just received an e-mail note from a San Francisco cinematographer and Steadicam operator who saw this thread and contacted me via my Metafilter profile.
"The device most likley used for that shot is a MoVI, although we know that the series’ director of photography also used a Steadicam Zephyr for many other sequences," he says. "The MoVI (or a similar stabilizer) is the only device that could have been used in that scenario. Here is an example to illustrate: https://vimeo.com/108494299."
posted by Paul Slade at 1:56 AM on November 20, 2014
"The device most likley used for that shot is a MoVI, although we know that the series’ director of photography also used a Steadicam Zephyr for many other sequences," he says. "The MoVI (or a similar stabilizer) is the only device that could have been used in that scenario. Here is an example to illustrate: https://vimeo.com/108494299."
posted by Paul Slade at 1:56 AM on November 20, 2014
Response by poster: More details from that San Francisco cinematographer:
"The MoVI (and all the other similar gyro-stabilized devices now in the market) can be operated solo, whereby the camera person can control tilting and panning with some level of accuracy (though not nearly as precisely as with a Steadicam). That’s where the remote control, and the additional MoVI operator, comes in. That person can monitor the footage remotely as well and have full control over pan, tilt and even focus. That makes the MoVI able to accomplish things like seamless hand-offs between multiple operators, through windows, into and out of vehicles, and I’ve even seen it be hooked, while in motion, onto a rope that is the lowered with the MoVI, all in one take.
"So, the MoVI is invaluable for certain shots that could never be achieved with a Steadicam. On the other hand, the MoVI is limited by the amount of vibration that the three axis gimbals can neutralize, where as the Steadicam, with its isoelastic arm, can absorb a much greater range of vibration. Thus, the MoVI has a certain look that can be discerned apart from the completely smooth Steadicam look. You might even be able to tell apart the shots made with a Steadicam and with a MoVI on Human Universe. The MoVI can suffer from certain “bumps” in more extreme circumstances (again, since there is no arm to absorb that motion).
"It’s funny— we often measure the success of a cinematographic technique in terms of whether or not it was noticed by the audience. The less the audience notices the technique, the better. The arrival of the MoVI is certainly having an impact on that measure of success, though I imagine it’s just a matter of time before that, too, becomes the norm in our eyes."
posted by Paul Slade at 3:30 PM on November 20, 2014 [1 favorite]
"The MoVI (and all the other similar gyro-stabilized devices now in the market) can be operated solo, whereby the camera person can control tilting and panning with some level of accuracy (though not nearly as precisely as with a Steadicam). That’s where the remote control, and the additional MoVI operator, comes in. That person can monitor the footage remotely as well and have full control over pan, tilt and even focus. That makes the MoVI able to accomplish things like seamless hand-offs between multiple operators, through windows, into and out of vehicles, and I’ve even seen it be hooked, while in motion, onto a rope that is the lowered with the MoVI, all in one take.
"So, the MoVI is invaluable for certain shots that could never be achieved with a Steadicam. On the other hand, the MoVI is limited by the amount of vibration that the three axis gimbals can neutralize, where as the Steadicam, with its isoelastic arm, can absorb a much greater range of vibration. Thus, the MoVI has a certain look that can be discerned apart from the completely smooth Steadicam look. You might even be able to tell apart the shots made with a Steadicam and with a MoVI on Human Universe. The MoVI can suffer from certain “bumps” in more extreme circumstances (again, since there is no arm to absorb that motion).
"It’s funny— we often measure the success of a cinematographic technique in terms of whether or not it was noticed by the audience. The less the audience notices the technique, the better. The arrival of the MoVI is certainly having an impact on that measure of success, though I imagine it’s just a matter of time before that, too, becomes the norm in our eyes."
posted by Paul Slade at 3:30 PM on November 20, 2014 [1 favorite]
« Older Western Digital Black or Blue Externally for Xbox... | Could I be allergic to alcohol? Newer »
This thread is closed to new comments.
posted by idb at 3:51 PM on November 6, 2014