Fbx cameras are stored in the scene graph as well as having separate camera properties.
The separate camera properties have a camera position and direction that is already fully
transformed based on its position in the scene graph so this results in the assimp
camera being transformed twice. Since the scene graph contains all the relevant
transformations the camera position and direction can just use default values. Fbx
specifies cameras along the positive x axis so the default camera values have a up and
direction set accordingly.
Fixes: #807#643
Fbx lights are stored in the scene graph as well as having separate light properties.
The separate light properties have a light position and direction that is already fully
transformed based on its position in the scene graph so this results in the assimp
light being transformed twice. Since the scene graph contains all the relevant
transformations the light position and direction can just use default values. Fbx
specifies lights along the negative y axis so the default light values have a up and
direction set accordingly.
The DecayStart value specifies the distance from the light that the light intensity
will drop to half. Using the default assimp falloff equation of f=1/c+lr+qr^2 then
the falloff coefficients can be calculated for either c,l or q accordingly by setting
r=DecayStart and f=0.5.
Fixes#851
+ change key type from uint to int (fixes bugs from negative int_t becoming a junk uint_t value )
+ detect and ignore channels with no keys in the specified take start/stop window
+ add test model with multiple animations
Another false positive uninitialized variable. Actually it's only false
if ToOutputVertexIndex does't return NULL which should only happen if
index is out of bounds. Add assertion to make sure of that.
The FBX importer now properly parses field-of-view tokens. Rebase
only the failing FBX entries to accomodate the fix, also removing
the comment noting the (now incorrect) comment about improper import
of the FOV property.
(keep the previous method for "allMaterials" option, this should be change too since the materials indexes used are not the good ones, or just remove the object / material test
Matt Oliver
Kim Kulling
Jared Mulconry
AndrzejKozik
wise86Android
roshan.c
Chris Russ
Turo Lamminen
mensinda
Kim Kulling
Otger
Olli Kallioinen
Richard
abma
George Papadopoulos
Jared Duke
Alexander Gessler
arkeon
Léo Terziman
attila-barsi
DoubleLee
Frederik Aalund