Follow the joints

Hi guys,

I have some imported some (collada) 3d animations, with two different animation cycles in vvvv. My aim is to switch from the first to the second depending on certain parameters.

Now the first animation starts playing at position (0,0,0) but during its cycle it moves to different other vector locations.

If I switch to the second cycle, while the first cycle is underway, it jumps to position (0,0,0) again.

I am looking for a way to fix this. Is it possible to track the vector position of a joint (root joint) and tell vvvv to use the current position as a start position for the next cycle?

Anyone can offer any help?

I hope I am clear in my explanation, if something is unclear let me know I can try and explain again.



hmmm… if that’s impossible…

how about connecting a basic geometry to a joint (or the mesh) like a cube or a sphere and track the position of that?

Is that going to be easier?

Any hints on how to do this?

attach patch pls

Hi antokhio,

Thank for your reply.
Patch and model containing the anims attached.

Hope it works.

Doros (409.4 kB)

hmm, skeleton’s nodes somewhy did’t open
but on my taste it’s works like that…
can’t see what’s in your patch, so try attach a screenshot maybe
think if you want to blend to animations together you have to use mixpose node
check on the diploma.characteranimation
there is some tutors on that…

tests.v4p (11.5 kB)

Hi antokhio,

Thanks for the reply.
Well, my patch does pretty much what your patch does, but in a slightly different way (using the timeliner). What I am aiming to do can be seen with ur patch as well.

If u play the dance animation, and then u switch to the idle animation before the dance animation is finished, then the model jumps to the original position in the space. Instead, I want it to switch to the idle animation at the location it currently is.

The mixpose node is for models without attached (externally created) animations, is it not? So you can combine animations created with the manipulation of joints inside vvvv?

The tutorials mainly speak about creating animations inside vvvv, and not how u can access the joint transformations of already created animations, that is why I am confused.

Any help appreciated.



Here is a screenshot of my patch anw.

hello everybody,
I am new to vvvv software,
I just wanted to know it the software can give me body joint position base on a ponit clouds.


Use a Kinect for windows and a current Microsoft Kinect Studio installation, and you can access the device, including a Skeleton with all body joints. In the default vvvv installation (with addonpack) you will see Skeleton (Kinect Microsoft DX9), press F1 after highlighting the node to see its usage.

If you happened to have installed the dx11 pack, you’ll see a replacement node for that, and a Kinect2 version as well.

To actually create a useful pointcloud from the kinect, the dx11 pack and more things are necessary, see vvvv-PointCloud.

I am sensing improper english language here, but don’t want to disrespect you in any way. If you really mean to detect (human) joints from the pointcloud with your own code, you’d have to have the smarts of a genius, and will probably run into legal issues anyway. I have no better advise for you than to stick to the Skeleton data provided right now.