Home of Professional Cinematography since 1996
> What Kind of Head
>Published : 8th October 2008
>I've got a shot coming up where centre frame will be a seesaw in a children’s playground - a single child is standing on it above the fulcrum, legs akimbo, rocking it back and forth. The desire is to move the camera so that the child and seesaw stay stock still in the frame and the ground/sky/horizon line rock up and down around them.
>Shooting super 35 academy for an HD tape finish so, at a push, it could be a lock off and the camera move tracked off the movement of the seesaw in post but this will preclude a hell of a crop into the frame. Another idea is to shoot a 'dutching' plate shot and the top taken off the seesaw - then comping it in to match the camera move...
>My preference would be to try and nail it live but I'm not sure what would be the best tool for the job. A 3rd axis remote head? Some sort of mechanical link between the seesaw and the camera?
>There'll be two sizes on it. The close up will only see the kid from the waist up and the rocking horizon behind them. The wide is just the seesaw (after the kid has hopped off) as it loses momentum and slows to a halt.
>What about having two seesaws set up parallel to each other.
>The kid goes on one and the camera goes on the other, then attach them with poles so they rock in unison. Could be tricky getting enough stiffness to make the movements match perfectly. Then you'd need to do "rig removal" to eliminate the poles. This could be done actually even better with pulleys to synchronize the movement of the seesaws. As one side of the seesaw goes down, it pulls the opposite side of the parallel seesaw up. With some real money, you could create a geared connection with two "pumps" that go down into the ground, driving a rod that spins and "pumps" up the seesaw opposite it.
>I'm a post guy, so I like the idea of shooting in a high resolution film, and tracking it in post. In the end, the flexing and bouncing of the seesaw that the camera is on will probably require some tracking anyway, but if it was close, you at least wouldn't lose so much of your image to the cropping.
>>>The desire is to move the camera so that the child and seesaw stay stock still in the frame and the >>ground/sky/horizon line rock up and down around them.
>You could do this with a roll axis controlled by motion control. A 3 axis motion control head with a nodal roll would be ideal.
>I would then attach an encoder to the seesaw to track that motion, and have the encoder feed into the motion control and roll the head.
>Heck, now that I think about it, it might be possible to use a remote head with a nodal roll and somehow rig the handwheel or an encoder of the same type as that in the handwheel to track the motion, and forget the motion control aspect (yeah, I know everyone would RATHER use motion control, but sometimes you must sacrifice). You might have to work a gear ratio to make the encoder turn enough so the roll axis would work on a 1:1 ratio to the action of the seesaw. Then you would have to adjust the gain control to make the angles match.
Gear+Rose Motion Control
>Wire removal is not that expensive these days -- what about a bracket attaching the camera to the see saw itself, then do a simple removal in 2K. It would be sort of like a very long Ubangi....
>Probably cheaper than MoCon!
>"What about having two seesaws set up parallel to each other. The kid goes on one and the camera goes on the other, then attach them with poles so they rock in unison. Could be tricky getting enough stiffness to make the movements match perfectly. Then you'd need to do "rig removal" to eliminate" (etc. etc.)
>I've tried seeing this shot in my mind's eye. I don't think you would want exact synchronization of the seesaws, motion control, pulley and cables, etc. I believe a much more visually intriguing look would be to actually have the seesaw rocking back and forth in one direction while the horizon goes in the other. In essence I would split the difference between the look of a camera fixed on the ground, and a camera exactly following the plane of the seesaw. Of course this infinitely complicates any mechanical rigging solution, leaving one either the choice of using motion control gear on the camera, or inventing the look in post.
>Why don't you get a rigger and grip to tube out parallel from the see saw and rig the camera to it.
>Mount the camera on the centre of another see-saw.
class="style21">>>Why don't you get a rigger and grip to tube out parallel from the see saw and rig the camera to it.
>I think you mean perpendicular rather than parallel but yes, this was my first thought. Then the size of the shot in the storyboard put me off the idea. Even using 'doggicam' style carbon fibre rods a 235 and an 18mm lens I think it'll be impossible.
>For the tighter shot that only shows the kid from the waist up the camera would need to describe an arc (rather than pivoting about the nodal point of the seesaw) at about 4'-5' (or however tall the kid is) and the momentum it would carry will start to add all sorts of troubles
>Needless to say, you could see backs start to arch in production as soon as 'motion control' was mentioned.
>Tom Townend wrote:
class="style21"> >>I think you mean perpendicular rather than parallel but yes, this was my first thought. Then the size >>of the shot in the storyboard put me off the idea. Even using 'doggicam' style carbon fibre rods a >>235 and an 18mm lens I think it'll be impossible.
>How about a really good operator with a "dutch" head.
>I'll bet with a little practice it could be done.
IA 600 DP
class="style21">> No parallel!!
>Mmm.... Thinking about it the problem just seems to grow and grow.
>Mounting the camera directly opposite the fulcrum of the seesaw and rotating (rocking) it about the nodal point/optical axis will give a shot that's orientated too low to the ground. The horizon will be below the height of the seesaw and the both the seesaw and horizon line will be too close to the centre line of frame.
>We're shooting 4 perf 35mm for a 16:9 finish so one could extract the frame from the top portion of the neg to rebalance this - otherwise one would need the camera to describe an arc at several feet above a pivot point; a bit like the weight at the top of the arm on a metronome :-s
>*heads back to the drawing board*
>Tom Townend wrote:
>>> Some sort of mechanical link between the seesaw and the camera?
>Is it possible to build the see saw. Many playgrounds have multiple see saws next to each other sharing a common pipe as their fulcrum. If you built your own like this, Put the main fulcrum pipe on bearings so that it acts as an axle, and mount the camera to the axle. Seeing the axle/pipe in shot wouldn't bee an issue, I think as people would think it was part of a multiple see saw unit.
>Otherwise there is a see saw that is made from curved tubing and is free standing and movable, perhaps that could serve?
New York Based Cinematographer
>I tried a little test myself with a hand held digital stills camera in cine mode. As you say, the head and torso move very little in relation to the horizon - if anything as the see saw comes up, the body tends to tilt slightly in towards that motion to compensate for the change in centre of gravity. Have you tried standing on a see saw yourself and shooting some "loose" shoulder mount shots, to get a feel from the kids perspective?
>When I played back the shots where I tried to synchronously follow the movement of the see-saw itself (given the obvious hand held shake) the best result seemed to be rocking the camera from side to side, whilst traversing a slight upturned arc to compensate for being above the axis of the fulcrum.
>Have you thought about trying it hand/shoulder mount and just following the motion by eye, and then removing any excess shake later in post? A few tests with a DV cam might prove it that's possible and see if there's a large enough amount of sway and tilt in the upper torso and legs .