as a few of you maybe already know, we’re currently working on a plugin which allows vvvv to load collada files (*.dae).
so far geometries, textures, materials and animations are loaded correctly.
a troublemaker is skinning, because of the lack of available test data and especially the lack of available collada viewers which support skinning. at this time it’s often a little hard do distinguish between a broken viewer or a broken collada file.
if you’re familiar with collada and know how to model a skinned mesh, it would be very nice of you to support us in this matter in form of contributing some collada files with skinning information (or better, animated skinning).
if you do so, please add a screenshot and the program name which exported the file.
thanks,
elias
ps: for debugging reasons small files are prefered of course…
That’s fantastic, I’m looking forwards to being able to use collada files.
Here’s a little tip I learned from the Papervision community:
The google earth file format (kmz) is actually a collada file zipped up with textures. So all you have to do is change the file’s extension to zip, and run it through your unzip program.
Another tip:
Using the free version of sketchup, you can export any of the files in the Google 3D warehouse to kmz (and thus DAE).
loading .dae files would be great! Would love to have some animated 3d meshes in vvvv.
Here a few .dae files exported with the feeling software plugin for 3dsmax 9, ive used them in papervision 3d,Flash.
Ive not been able to export animated .dae’s to flash so these ones are not animated.
(or better, animated skinning)
What does this mean? Never tried this, could i take .gif’s as texture?
If you want to check if these files are displayed correctly
(just changed the paths to be relativ hope it works…)
-> http://vimeo.com/3780791
hi gegenlicht,
thanks for your input, i’ll try your provided files later.
to answer your question:
no, it has nothing to do with .gif’s as texture. skinning is a technique to describe how a few bones of a skeleton influence a vertex’s position in 3d space.
if i remember correctly there is a special tool in 3dmax for this purpose. for example if you want to animate a character, you first have to create a mesh (you could call this mesh the character’s skin), then you choose a skeleton which will be connected with your skin. if you’ve done that all future animations are done by setting keyframes and moving the bones of the skeleton. skinning will make sure that all vertices move along with your animated skeleton.
3dmax has a good help topic about this. a few examples are also included.
i’m not in possession of a current version of 3dmax, so all i can tell you here are my memories going back about eight years when i played around a little with 3dmax.
i’m not in possession of a current version of 3dmax, so all i can tell you here are my memories going back about eight years when i played around a little with 3dmax.
if you just want to try this plugin out i recommend downloading the Collada Asset Manager.
They have a nice example included called astroBoy_walk.dae, you find it under C:\Programme\ColladaAssetManager\Data\astroBoy_walk.dae if you just install it under the default location.
Open the help patch of Mesh (EX9.Geometry Collada) and use it in the skinning example.
So i experienced some troubles with the collada export from max, I still can grab animation only on some exported models, and did’t get a clue how i actually can do skining on the model.
Would like to know for some software with native collada suport…
Anyway gonna try maya and XSI everything better then blender ;]
Ok finnaly i’ve been able to make skinning in max and export with it animation. Schem is buggy, but here is files… Dos’t work from first time, reopen your DAE in max and export it again.
HI! Elias, hope you will ever look in this post! I played a lot with collada, i can’t understand, maybe that is a bug in expot! But if i export some sort of Complex movement (rotation scalling and movement) in all ZYX Axis, it won’t ever move in vvvv, but if it will be some sort of simple movement (2 Axis Max) it’s ok ;(.
(I thoght the problem is in number of keyframes or objects, but no…)
T-morrow gonna check out skinning more closely for that…
Anyways you can find me on skype: Antokhio anytime. Can share you more detailed results, and might be some test we can perform with that!
Documentation for ColladaLoader plugin available here.
New version (beta2) available for download. Has better support for COLLADA files loaded from 3D Studio MAX and adds various config pins… see COLLADA wiki page.
this is a nice one. It´s working like charm with my Cinema4d- Exports. Now here comes my little feature request:
Viewing my Test-DAE in an XML- Editor i see the complete camera- information of the Scene integrated (tagged “library_cameras”) into the code.
Would it be possible to get that Information, too?
This would open up a whole new universe for me: being able to animate the cam in C4D (or any other 3d- Editor), use it in vvvv and make Cinema a comfortable Editor for vvvv- Scenes.
I was thinking about getting that information out via regexp and vrml- files (which pass the most comfortable formatting for camera-animation-data coming from c4d in texteditor-readable format) but the collada- way would be absolutly my favorite over that.
Another more easier thing that would be great would be the the length of the animation in the .dae (there is a parameter called TIME somewhere at the beginning of the “library_animations”-tree, which seems to inherit this info).
I attached a Test- DAE coming from c4d for checking back to other exporters… no skinning data for now though, sorry!
i hope you like my ideas, cheers and thanx for now, c
thank you very much for your feedback, i really appreciate it.
nice to hear the export/import from c4d to vvvv works. did it work out of the box? if not, could you be so kind and add a few notes to the COLLADA wiki page?
to your feature requests:
the camera thing should be easy to implement. thinking about a new node (Camera (EX9.Geometry Collada) or something like that) which takes the Collada Model and an index as inputs and outputs the view/projection matrices and those other details, like field of view etc…
the TIME thing:
hm, well i could output the maximum animation time of all the animations affected by your selection, but somehow i don’t know where to put this new animationTime pin…
you have to know, the goal of this plugin was to load meshes from a collada file into vvvv and then have a look if it is even used by the community. well thanks to the feedback here and in other threads we see that people actually use it (and hopefully like it) and so i think we’ll have further discussions about this topic and especially how the whole collada part should work inside of vvvv (what nodes/pins etc.).
so please, don’t be mad, if i’ll not rush for that animation time pin right now, because i have that feeling the whole animation part will become something bigger eventually.
the export/import- workflow worked very well and for now it seems like everything showed up correct in vvvv. I will have some more tryout concerning textures, performance with bigger an more complex meshes and let you now about my experiences.
A camera- node would be definetly great and i´ll start looking forward to that right away (when will we see it…oh…i can´t wait… ;-) )
Concerning the TIME- Info: I´ll have a try to get it out via regExp or such (or perhaps code my first own node…) and build me a module for now. Actually all i need is exactly what you mentioned: the length of all animations exported to the collada- file, as i want to be able to use the whole exported collada- scene as something like a 3D- Videosequence, which has the same length as my c4d- scene.
btw.: should i keep posting in this thread or switch over to another more general one (e.g. the COLLADA wiki page you mentioned… did you mean the vvvv-wiki-COLLADA- page by the way?)?
hey guys, i´m not sure if this is where I ask for help with collada, but i’m having problems with one .dae file.
the mesh seems to be loading fine because the spreaded transform seems to be ok, but I can’t see it. It’s exported from 3dmax and it’s going to be used for a mapping projection.