[begginer] How to animate object holded in both hands?

Hello,
i wanted to ask how to animate human swinging sword, holding rifle etc.
My IK armature does not act like i need, i utilized emptys attached to desired object as reference points, attached bone of left hand to empty and second empty to right hand.
My aim was to manipulate with my object using right hand, but it did not work, on my left hand only rotation is applied to palm hand (which is the one under bone constrain), resulting in rotation of palm, but not rest of hand.
Any tips or tutorials please? Thank you.

Perhaps there is a clue here: http://cgcookie.com/blender/2011/08/22/animating-a-character-picking-up-an-object/

The general technique that I use, first of all, involves an Empty that is parented to the object that’s going to be picked up. This allows me to set the position, orientation and so-forth of “the thing that actually gets picked up” (namely, the Empty …) any way that I want.

I used to use the technique described in the above tutorial … setting a Child Of constraint and then (since you can “animate anything” in Blender now) animating the Influence in one step from 0.0 to 1.0.

But, lately, I’ve started doing it differently. I’ll now shoot the footage from the same camera in two different “Scenes.” In one scene, the influence is zero: the character goes through the entire sequence, not carrying the gun. Then, from another otherwise-identical scene, the character is carrying the gun, throughout. (All of this is being done, at this point, with “OpenGL Preview” shots.)

Typically, in fact, I will set up several cameras like this … medium shot (MS), close up (CU), extreme close-up (ECU), over-the-shoulder (OTS), reverse-angle (RA), and so on. Each with and without the gun. After all, OpenGL renders are “free.”

Now, I go to a video editor (Final Cut Pro, VSE, whatever …) and edit the strips together to produce the final sequence. I know that I can “cut” imperceptibly from “not carrying the gun” to “carrying the gun” at any point, because the camera and everything else is unchanged. I also know that I might decide not to actually use any footage at all which shows the gun actually being picked-up on camera. So, I edit together several versions of this important-to-the-story scene to decide which one works best … which one is visually tightest and strongest.

The more I learn about the importance of film-editing, and the more I do it before I “shoot” (render …), the more things like that wind-up “on the cutting-room floor.” When you do this, you might find that the tightest sequence of all goes something like this:

  • Brief shot of the gun sitting there, maybe with the good-guy’s leg in view to establish position.
  • CUT TO (CU): The good-guy has an angry snarl.
  • CUT TO (MS): The guy reaches down.
  • CUT TO (RA): The gun comes to rest on his shoulder.
  • CUT TO (MS): Carrying the gun, the guy stalks out-of-frame to go kick some bad-alien a*s.

And, notice how, in this version, you never actually see him on-camera picking up the gun, because you don’t have to. Your attention is already focused on the next shot, when you’re going to see him spotting the bad-alien, and the next shot after that, when our good-guy starts smoking aliens. :slight_smile: Therefore, I cut tightly to get straight to the point. (And shot #3 above might not “make the cut,” because if you saw where the gun is, and you see him drop it over his shoulder, you can easily guess what happened without actually seeing it happen. (“Hurry up! I wanna see him smoke some aliens!!”) So, you never have to render it at all, and even if you do render something, you never actually have to animate a change to a variable: you can make the cut anywhere you want.

The only time you’d have to animate something is if our good-guy can “Use the Force, Luke!™” to cause the gun to jump magically into his hand so that he can: use it to start “kicking alien a*s,” which is what the audience actually came to see.