Graphics Reference
In-Depth Information
”animation 0”
:
{
”channels”
:[
{
”sampler”
:
”animation 0 translation sampler”
,
”target”
:
{
”id”
:
”node
box”
,
”path”
:
”translation”
−
cam01
−
],
”count”
: 901,
”parameters”
:
}
{
”TIME”
:
{
”bufferView”
:
”bufferView 4509”
,
”byteOffset”
:0,
”count”
: 901,
”type”
:
”FLOAT”
}
,
”translation”
:
{
”bufferView”
:
”bufferView 4509”
,
”byteOffset”
: 3604,
”count”
: 901,
”type”
:
”FLOAT VEC3”
}
}
,
”samplers”
:
{
”animation 0 translation sampler”
:
{
”input”
:
”TIME”
,
”interpolation”
:
”LINEAR”
,
”output”
:
”translation”
}
}
}
Listing 2.1.
Animation listing.
glTF targets GL APIs by creating structures that can be sent directly to a
runtime engine built with WebGL, OpenGL ES, or OpenGL. glTF consists of
a set of files that are directly compatible with web technology—a JSON file,
binary files, GLSL shaders, and images. COLLADA2GLTF is the open-source
converter from COLLADA to glTF. As shown in Figure 2.7, it can be used as-is,
be incorporated into a content pipeline, or serve as a reference implementation
for other converters.
2.9.1 COLLADA2GLTF
COLLADA2GLTF content processing incorporates several stages that progres-
sively transform the content into the data structure accepted by GL APIs:
•
Triangulate:
Geometry is transformed into primitives accepted by GL. The
most common one is
TRIANGLES
.
•
Single index:
Only one index per vertex attribute can reference all the
attributes associated. Modelers often use one index per attribute, which
