PDA

View Full Version : General questions about Nevron



Ryan Roye
09-19-2013, 07:56 PM
I have questions about nevron mocap animations:

1) Does every capture generate the same-sized and proportioned skeleton or does it change with every recording?

2) Does Nevron generate a T-pose at frame 0/1?

3) Does the rest rotation change at all between recordings?

I don't own Nevron; the reason I ask concerns the possibility of integrating my motion capture workflow with nevron-generated motions (so someone can use their own custom rig and apply motion capture animations to it instead of the genoma/nevron rig).

BeeVee
09-20-2013, 04:22 AM
1. If you mean bringing in an FBX or BVH motion capture file and retargeting a Nevron skeleton to it, then no. Those measurements are dependant on the performer.
2. No. Often performers assume a t-pose at the start of the mocap, if not you can (and should) create one of your own for proper retargeting.
3. Not sure of your question here. Rest rotation is applied to a rig, but Nevron moves around that.

The NevronMotion rig is required for all retargeting and Kinect capture right now.

Hope this helps,

B

Ryan Roye
09-20-2013, 02:39 PM
Thanks! One of these times I need to bug someone for some kinect-generated mocap samples and take the whole concept of applying motion capture animations to custom rigs in Lightwave even further. There's a lot of workflow ideas that need to be tested and experimented with.

basically, so long as assumptions and measurements can be made to rig template, an alternate method of retargeting allows for custom rig motion capture adaptation and I believe it will provide some excellent synergy with Nevron.

BeeVee
09-20-2013, 03:58 PM
I'm not altogether sure how you'd save Kinect mocap for use. I think you'd need to do the mocap yourself, directly in LightWave with Nevron. I'm really looking forward to seeing what you do with Delura with it though :)

B

LW_Will
09-20-2013, 06:38 PM
Ryan,

I've got Nevron and I would be happy to provide mocap data for you to use to see if this will work for you. Unfortunately, I cannot get the Kinect to work with the plugin. It is very frustrating.

As to your question of the size of the objects... its a little different that you are thinking. (As I see it, IMHO) I'm thinking of the character being divorced from the mocap. Basically, the character can be made of three things; the character mesh, the skeleton, and the kinect (fbx) skeleton. In my video, I'm worried about getting different characters set to a "universal" figure. Daz3d uses a universal character for their products, which I think lends itself to more efficiency when working with rigging. My idea is to use the Kinect skelegon to make up different walks, people chatting, whatever, and apply them to different characters in different parts of the video.

So, you can make changes, like the gorilla effect seen in the SIGGRAPH videos by changing parameters.... but you could as easily change it back or apply it to a dozen characters. As your videos have shown.

Please, contact me and we will talk!

BeeVee
09-20-2013, 07:19 PM
What problem are you having with the Kinect Will?

B

Ryan Roye
09-20-2013, 10:51 PM
Please, contact me and we will talk!

Will do! 4-6 Nevron-generated animated mocap samples would help me get a better idea of how things are structured.

LW_Will
09-21-2013, 08:41 PM
Ben,

I can get the Kinect visible in the virtual studio, I managed to get the retargeting to work, but I cannot for the life of me get the Kinect to drive either the Kinect OR the Genoma Skeleton.

WTF?

And as soon as I get that figured out, I'll send my stuff to you Ryan...! ;-)