HI All,

I just want to finish the point about work with a Translation Framework.

As we continue to examine various formulations of the Humanoid model

we will see that a, let me call it "Legacy" for now, more or less 'standard'

version emerges.

  

The skeleton is seen as the primary way to animate the character

and a more or less realistic human skeleton hierarchy of transformations

is used to produce realistic human motions.

To vizualize these motions geometry can represent surfaces for

individual body parts and be animated directly with the skeleton,

and/or a complete geometry can produce the outer 'skin' surface

and be animated by algorithm that produces a weighted movement

of skin vertices according to motions of the skeleton.

It is also common and handy to incorporate an independent means

of moving individual and sets of vertices using a scalar input.

 

So, I hope we can see that there exists a basic "Legacy" skeleton

designed from a set of widely used standardizable features. 

  

Any character authoring system somehow presents the Character,

or Avatar, or Armature, or whatever it is called in the application, 

consisting of these basic elements:

- the skeleton hierarchy,

- a way to add geometry(s) for, skin, limbs, and accessories,

- styles to animate the skeleton and thus geometry, and,

- reference locations associated with the humanoid

for functional landmarks such as effectors, sensors,

accessories, data stores, and viewpoints.

 

So, the HAnim character documents consensus

best industrial and research practices for constructing

a realistic animated humanoid with those basic features.

Thus, it is not surprizing that under the hood of various

commercial and free tools the basic "Legacy" data is

the same or derived in the same way from the same data.

 

The only real difference between the various 'Legacy'

authoring presentations is the style of the default 'humanoid'

authoring interface presented to the rigger and animator. 

That is the number of active elements of the skeleton and the

names of those elements in the more or less 'standard'

transformation hierarchy.

 

Well, there is one big difference in the basic presentation of

various tools.

Some will show the skeleton hierarchy in terms of an arrangement 

of "bone" elements, each with a pivot point, and some will show

the skeleton hierarchy in terms of a set of "Joint" elements

with a center of rotation for animation.

Possibly arguable, the tool that shows the bones-only hierarchy

is misleading because unnecessarily hiding the real

'under the hood' animation process, which relies on knowing

a joint center of rotation.

 

We see that legacy animation really needs to know

about skeleton joint center, how does that joint get

positioned in the hierarchy, and what is the current

rotation applied to that joint.

Fortunately, even though the authoring interface may

deal with bones and pivots, the joint center and animation

binding information needed to generate the 'Legacy'

standard skeleton hierarchy of joints, connecting segments,

and landmarks can be retrieved from the bone-only

presentation by a simple process. Mainly, data for a bone

is applied to the 'standard' model ancestor joint.

 

So, maybe too much, or not enough, but a main point is that

the HAnim standard is platform independent and is incidentally

implemented by the x3d authortime and runtime. I have not

seen any 3D avatar example skeleton or complete model that

could not be implemented using HAnim. Skeletons are based on

the same or similar hierarchy, basic and weighted skeleton-driven

geometry and skin animation, and even scalar-driven animation

is the same except in name and special application-dependent

structure and format.

Finally, all competent examples will resolve to the same gltf

asset formulations.

In short, whether the authoring tool shows only bones or reveals

joints the runtime gets the same data.

 

I guess I should show some diagrams, but they are in the

standard, and main idea is that of course HAnim wants help

find a common form for standards-track humanoid authoring

and delivery in in the standards-track World Wide Webiverse.

Of course HAnim is open to documenting open consensus

improvements to the "Legacy" humanoid character model.

 

Finally, of course, in my experience, HAnim and Web3D is

eager to provide a firm ISO/IEC base for facilitating

creation of standards-track documentation and implementations

for new open and free technologies and extensions aimed

at our open and free WWW.

  

Thanks and All Best Fun with HAnim,

Joe