How to make a rig that simulate a Steady Cam

Hi,
I want to simulate a 3-axis stabilisation Steadycam (like this). I just know the basics of rigging and constraints and I don’t know if it is possible.
So is there a way to have an object (a camera) parented to another object (the first stabilisation axis) himself parented to a third object (second stabilisation axis) and so on, while having a correct behavior of the steady cam system when you rotate the last object of the “chain” (the part you hold in the hands). I want it to have the limits too (if you rotate the camera stay steady but if it’s too much the camera follow the rotation).

Anyone know how to achieve that ?

Use an armature for the rigging - then add an IK constraint to the last bone in the chain (camera end), use the camera as the IK target. When you move the camera, the mechanism will follow. You will probably need some pole targets to align the rig properly, and remember that with IK chains you should not have them dead straight or they flip around a lot. Also with IK chains, other constraints are not followed so if you want limits on the bones in the chain, these should be set in the Bone’s panel >> IK properties section. You can set limits and stiffness, etc.

Cheers, Clock.

Something alike?

- Blender 2.5 Handheld Camera Rig Tutorial

Thanks Clockmender, I will try with an IK chain.

Burnin, My purpose is not to rig a blender camera, I just want to chain objects to simulate the stabilisation of an electronic gimbal steadycam. Though your link for rigging the blender camera is really interesting ! Thank you ^^

Offhand, I’m not sure that I personally would go to the trouble of doing that … even though it “simulates reality,” I tend to think that you really don’t need to do that. You just need to plausibly simulate how the camera moves, as seen by the audience of the final film. Never mind how you get there.

You might be able to achieve “a plausible-enough effect” simply by constraining the camera to point toward an Empty, which (say…) might be parented-to the object that the camera’s supposed to track … or might be a separate thing altogether. Then, animate the camera as it points toward that Empty, to reflect the position of the cameraman as he holds his (mostly, perfect) “steady” cam.

I’ll wager that there will only be a few camera-moves that really say to the viewer, “Okay, this is steady-cam.” These, of course, are the (few) shots that are slightly imperfect. The object moves abruptly and the camera (while not entirely “losing” the object) “catches up.” That sort of thing, and probably only in a few portions of the shot.

Also: “even steady-cam footage is edited.” To you, this means that most shots might need nothing more than “the camera is perfectly tracking the empty,” while a few more actually need to be played-with to introduce a little bit of “plausible imperfection.” You’re not actually going to “shoot this thing in One Take.” Take full advantage of that.

Having watched that video, I would definitely build the steady-cam device as a series of meshes for the separate parts, then build an armature with an IK chain as it’s main structure. You can then do a “Bone” parent of the parts to the armature. By that i mean select the part, shift select the correct bone in the armature (this should be in Pose Mode) then key CTRL+P >> “Bone Parent”. this will assign all of the mesh vertices for the part to the chosen bone. For the animation, you will need a focal point object, which you can move around, and pole targets to keep the rig in line, I think.

To mimic the animation shown in the video, it would be a good idea to control the focal point object (I use Empties for this) with drivers so it moves to where you point the camera, but, it moves around slightly, using drivers, to simulate the delay in the camera tracking to the focal point. Another option here is to have say 5 focal point empties, one in the centre and four around it at 90 degree spacings, some short distance away. You can then add 5 Track To constraints to the camera object that point to each of the focal point objects, then vary the influence of each Track To constraint to simulate the camera “smoothly finding the target again” after you move the device.

I hope this all makes sense!

I am guessing you want to build the device, not just show what the photos or videos taken from the camera look like, If not I have misunderstood you - so ignore my ramblings…

Cheers, Clock.

This makes perfectly sense, I’m sorry I explained my needs so bad in the first post. As clockmender understood, I do not want to see trhough the “rig”, It must be the subject of the scene itself, so no blender camera are “stabilize” or attached to it in any ways (maybe after it’s made, just for fun ^^).

So far my experience where not very conclusive. My real problem is to set the limits of movement, I understand that “track to constraints” on empty combine with parenting are necessary here.

Thanks everyone for your answers.