saindd
-
Posts
221 -
Joined
-
Last visited
Content Type
Profiles
Forums
Blogs
Articles
Docs
Gallery
Posts posted by saindd
-
-
A mount point is a mount point, an eye node is an eye node. Both are named joints, but there is a reason they have different names. Either way, his problem is not related to a center or anything, you're wrong.
Just check if useEyeNode is set and/or check the eyeOffset, if you don't have an eye node for each weapon model.
-
Mount points are used for third-person world view. shapeName is mounted on mountX and shapeNameFP is placed at the position of the eye node.
Mounting a first-person image to the third-person world view mount points makes no sense.
-
FP images are not mounted. You're confusing third-person world models with first-person.
-
"My skeleton, the mesh and all the animations are correctly set up in the Y-Front, Z-Up orientation."
So, yes, the arms, the weapon and the skeleton are all facing the positive Y, having Z up and are located at the center. By the way, being at the center shouldn't really matter for anything, since the FP camera is translated to the eye node in the player model.
@shoiko, you are probably not setting useEyeNode in your datablocks. Try that out.
-
It's not that simple. I'll have to re-orient all my 50 animations and the skeleton.
-
Right now, Torque uses Z-UP, Positive Y front. But this sucks, because every tool in my set in Negative Y front. Blender, Maya, whatever, you name it. Even Rigify uses a different orientation, which is Blender's default. So, how can we change the default orientation of the engine?
-
I never had heard about modules before. Guess i'm going to start to backport a few of those features to the Full template! :mrgreen:
-
What's new on this new template?
-
What's the purpose of the "ambient" sequence that a lot of the included assets define?
-
Randomly, some material, object and sequence names get upper cased by Torque 3D. I edit the CS files manually, and it renames then. For example, if i name a sequence "run", it converts to "Run".
Why does it do such an annoying thing?
-
I bought the source art pack for the Soldier and exported the FP_SoldierArms_Rigged_Master_Max file to DAE, replacing my DAE with that one. That should work, since the only thing i changed in the original datablocks was the shape path. However, the same problems i'm having with my own model happened with the original one. So, definitely not missing nodes, scale or anything.
What might be the issue here? This really sucks. :(
-
I've removed the prefix attributes from the Player/Ryder datablocks and i'm just pointing to the shape file. They are mounted to the eyeMount or eye bones, if present and if useEyeNode is true in the datablock: https://github.com/GarageGames/Torque3D/blob/2044b2691e1a29fa65d1bdd163f0d834995433ce/Engine/source/T3D/shapeImage.cpp#L1876
Talk about hard-coded stuff: https://github.com/GarageGames/Torque3D/blob/2044b2691e1a29fa65d1bdd163f0d834995433ce/Engine/source/T3D/player.cpp#L514
-
-
Well, this is being more difficult than what i initially thought. I'm somewhat finished with my first project in terms of code and logic, so i started replacing the stock T3D art with my own. Unfortunately, it's being a nightmare! In the shape editor, everything is OK: the sequences are setup fine, the scale is correct ("correct" as in the same size of the original player), the bounds are set, the eyeNode/eye/camera bones are all set. It should be just a matter of replacing the art files in the datablocks. But that's not the case.
The first-person arms model is not appearing at all, or playing any sequence. As for the first-person weapon model, it's appearing, but in the wrong position and playing the old Ryder animations (that's why i'm assuming it's in a wrong position, because the model is being bound to another skeleton instead of mine). I don't see how it's possible to do that, however, because the model i have specified for my weapon in the datablock contains the same skeleton as my first-person arms one. They just differ on mesh data, the skeleton and animation are all the same.
Since it's not the first time i find hard-coded stuff in T3D, would anyone be interested in sharing the caveats, tips and tricks involved in replacing stock T3D art?
-
So, in theory, i just have to make sure my movement animations are not in-place, so the engine can calculate the difference between the root node and the bounds? What if i don't have a bounds object, using the one specified by the engine?
-
I don't want to disable shadows being casted form the FP model, but casted from the TP model into the FP one. Open a sample scene and note that the shadows from the upperbody and head of the third-person model are casted into the FP arms and weapon. You can notice this if you tweak with your PSSM configuration, or making the shadow volume more "hard".
-
Yes, but that was exactly what i was trying to avoid. Checkmy first post. I didn't want to have thousands of animation files pre-baked for something that could be done in realtime.
BTW, here's how a model looks with a single-frame animation blended with a walk animation:
http://i.imgur.com/AJiqONa.jpg
It seems to be blending the root pose in there, due to the position of the arms, but it's not doing any subtraction or addition between the selected animations.
-
Torque prevents the weapon from being rendered through geometry by pushing the model towards the camera. As described here:
http://www.garagegames.com/community/forums/viewthread/121167
https://www.garagegames.com/community/forums/viewthread/131993
But there's another problem, other than that. The 3rd person player model casts shadows into the 1st person model, resulting in a terrible effect. It might be realistic, but it's not what other games and engines have been doing. Is there a way to disable this?
-
What if your walk animation animates other body parts, because it's used in the no-weapon sequence?
-
What do you mean by automatically blended? Is this hardcoded somewhere?
-
I don't think you got the issue. I don't have an animation for each weapon type, just a pose for the aim position. I want to blend the positions defined in that pose with the rest of the keyframes in the walk animation, dynamically creating the walking animations for each weapon type.
https://docs.unrealengine.com/latest/INT/Engine/Animation/AnimHowTo/AdditiveAnimations/
http://udn.epicgames.com/Three/AdditiveAnimations.html
http://www.gamasutra.com/view/feature/3456/animation_blending_achieving_.php
-
According to this: http://docs.garagegames.com/torque-3d/official/content/documentation/Artist%20Guide/Primer/torque_art_primer.html
The export of ground transforms depends on the modeling package, which i'm assuming depends on a specific DSQ plugin, like the one for Milkshape. Is this information still valid? How can i export a ground transform from Maya to COLLADA and, then, Torque?
-
I'm trying to replicate some behavior from other engines in T3D's animation system, specifically threads and blends. I have a walking animation set, in 4 directions, and the player has no pose for the arms. However, i also have a single keyframe of the character holding a sword and another a bow. How can i blend that single-frame animation (pose) of the character aiming a weapon with the movement animations? I really don't want to create the same walking animations over and over again, for each kind of weapon i end up creating (since it's a medieval game, i have a shitload of poses to work on: swords, bows, crossbows, magic spells, etc.
-
Is there any fix planned to support more layers being imported per heightmap? What if, instead of loading RGBA, i only import 8-bit grayscale masks? Instead of combining channels in World Machine, i can export them individually as heightmaps. Would Torque support that?
Change orientation
in Rendering
Posted
COLLADA doesn't seem to support horizontal axis definition, only up_axis in the asset tag. It would have to be something like the "ground model" option.