The Orange Project

Game Engine, Players & Web Plug-in, Virtual Reality, support for other engines

Moderators: jesterKing, stiv

CG_Tiger
Posts: 0
Joined: Thu May 12, 2005 11:38 am

The Orange Project

Post by CG_Tiger » Tue Oct 18, 2005 10:59 am

The Orange project has done wonders to the animation side of Blender, and there are talks of an Orange 2 project.:D

What do you think of a Game Orange Project?
This will give a massive boost to the Game Engine part of Blender.:D

LetterRip
Posts: 0
Joined: Thu Mar 25, 2004 7:03 am

Post by LetterRip » Tue Oct 18, 2005 9:15 pm

CG_Tiger,

I briefly discussed the idea of a 'Game Orange Project' with Ton.

If one were done, it would almost certainly not use Blenders internal game engine. Instead it would focus on improving the work flow between Blender and external engines such as CrystalSpace or Ogre3D and improving Blenders modeling tools for game development.

LetterRip

malCanDo
Posts: 1
Joined: Mon Oct 21, 2002 1:44 pm
Location: Ireland
Contact:

Post by malCanDo » Wed Oct 19, 2005 1:37 pm

I think that would be a great idea!

With the Orange movie project, a number of new ( and necessary ) related features have been added to Blender.

For a games project, one can only imagine that development would focus on adding necessary new features to allow for a semi-professional game to be developed.

With VERSE, it might even be interesting for interested artists from the community to get together to build some test environments, which then could be used with an existing game template ( eg a fly-around / walk-around ), to see if simple environments could be created with a minimum of cost.

Mal

poutsa
Posts: 0
Joined: Sat Jan 03, 2004 4:49 pm
Location: Munich (Germany)

Post by poutsa » Wed Oct 19, 2005 5:02 pm

I also think that would be a Great Idea!

I also believe......
Blender needs after the Orange Film Project a New Special Site "Blender Animation Studios" like Pixars and the Next Steps to make a Full Lenght High Quality Animated 3D Film like Pixars Studios ...as Example "Toy Story" "Nemo" "Shrek" and others..!
So we can see in Cinemas Worldwide a Complete Animated 3D Film made with Blender! That would be sooo Cool!!
With a biger Team with People who use Blender and other CG Programms very Good! I also believe this People must be Paid for this.....like in Pixars Studios!
So the Blender Foundation can be one Day like Pixar maybe!!Why not!
I know thats very Difficult to do bud nothing is inpossible!

You see Manosdigitales will Produce a FullLenght 3D Animated Film with Plumiferos as Example!!

http://www.manosdigitales.com/produccion.html


sorry for my English!

reimpell
Posts: 0
Joined: Wed Oct 01, 2003 5:09 pm

Post by reimpell » Wed Oct 19, 2005 5:25 pm

LetterRip wrote:Instead it would focus on improving the work flow between Blender and external engines such as CrystalSpace or Ogre3D and improving Blenders modeling tools for game development.
Blender can be used to create and export static, armature and vertex animated meshes as well as complete scenes to Ogre. There is also an import script for static meshes from Ogre to Blender. For scene export, Blender's logic properties have proven to be very useful to add game specific data to the scene.

At the moment, the main drawbacks using Blender as DCC tool for Ogre3D are:
  • No internal lightmap solution, together with the restriction to a single uv set (i.e. no external lightmaps can be imported that have a different resolution than the diffuse texture and hence use a different uv layout).
  • No custom shader support (preferrably something render api independent like Cg).
  • Billboards and particles don't map well.

LetterRip
Posts: 0
Joined: Thu Mar 25, 2004 7:03 am

Post by LetterRip » Wed Oct 19, 2005 9:16 pm

I think it would be valuable to come up with a well defined list of what would be disireable for such a project to accomplish.

Also it would be good if we could do a preliminary assessment of the capacity for funding. If a cosponsor were found (as is the case with project Orange) then a game project could be done with likely similar funding amount - ie 1000 DVDs at 40 Euro each. If Blender Foundation were to provide all funding and all talent then possibly triple that.

So what key deliverables are desireable, reimpells list is a good start, (could you be more specific on what could be done to make billboards and particles 'map well'?)

In addition to his list, I would suggest smoothing groups functionality, and the ability to generate tangent space normal maps given two meshes (and the ability to sculpt a high resolution). Also improved mesh reduction tools (Cambos script is quite excellent at this, although something like Modos might be better), and perhaps general polygon tool improvements (ie perhaps aim for Silos functionality), and texture painting tools (3d view with storage of the view matrix and pose, the 3d view is saved as a high resolution image, can be painted on with the gimp, and then once finished with the gimp reprojected onto the model).

LetterRip

LetterRip
Posts: 0
Joined: Thu Mar 25, 2004 7:03 am

Post by LetterRip » Wed Oct 19, 2005 9:30 pm

A quick note,

instead of a full game project - it might be possible to sponsor a capable developer who is already familiar with the code base to address these issues.

zr, theeth, brecht and jiri are all very capable developers familiar with the code base and are students. If Ton is not interested in such a project, one of them might be willing to be sponsored to do the work. Or alternatively see if it would be possible to have these as suggested summer of code items.

LetterRip

kakapo
Posts: 0
Joined: Sat Sep 04, 2004 2:32 pm

Post by kakapo » Wed Oct 19, 2005 10:02 pm

btw. what happened to the half-edge mesh project? i think it was very promising and it's one of the most important things blender needs.
Last edited by kakapo on Wed Oct 19, 2005 10:15 pm, edited 1 time in total.

LetterRip
Posts: 0
Joined: Thu Mar 25, 2004 7:03 am

Post by LetterRip » Wed Oct 19, 2005 10:12 pm

I just emailed with Joeedh about two days ago, he is just really busy with school and doesn't have time for it right now.

LetterRip

reimpell
Posts: 0
Joined: Wed Oct 01, 2003 5:09 pm

Post by reimpell » Wed Oct 19, 2005 11:27 pm

LetterRip wrote:could you be more specific on what could be done to make billboards and particles 'map well'?
Billboards are just rectangles that always face the camera while fixing a point or an axis. They usually display a texture, e.g. the light flare of a lantern. The concept in Blender that comes close to this are HaloTex Halos with image texture. Unfortunately, there is neither a way to prevent Blender from radial fading this image on display nor is one able to fix a rotation axis.

Ogre's particle system emits billboards according to emitters. There are predefined emitters like point, box, cylinder, ellipsoid, hollow ellipsoid, ring. Blender's particles are emitted from vertices/faces instead. Exporting them would mean to create a point emitter for every such coordinate which isn't very efficient.

LetterRip
Posts: 0
Joined: Thu Mar 25, 2004 7:03 am

Post by LetterRip » Thu Oct 20, 2005 12:01 am

Heh,

I knew what billboards were, I just didn't realize Blender didn't directly support them.
Unfortunately, there is neither a way to prevent Blender from radial fading this image on display nor is one able to fix a rotation axis.
I could have sworn I saw a patch for an always face camera for billboard equivalents from harkyman.
Ogre's particle system emits billboards according to emitters. There are predefined emitters like point, box, cylinder, ellipsoid, hollow ellipsoid, ring. Blender's particles are emitted from vertices/faces instead. Exporting them would mean to create a point emitter for every such coordinate which isn't very efficient.
Jahka is doing cool work on particles and was looking for suggestions of good ideas to implement, you might want to contact him regarding this request - I'll be sure to mention it on IRC next time I see him, I'm almost positive no one was aware that this was an issue.

LetterRip

reimpell
Posts: 0
Joined: Wed Oct 01, 2003 5:09 pm

Post by reimpell » Thu Oct 20, 2005 12:31 am

I forgot to mention that some people asked for a "snap on grid" option for level/mesh editing. Anyway, the most severe issue is the restriction to a single uv coordinate set and the lack of an internal lightmap solution. The other things are just nice to have but not absolutely necessary. E.g. the combination of diffuse and normal texture is already exported to a corresponding Cg shader, billboards and particles could be exported via some logic property conventions and BPy creation wizards.

LetterRip
Posts: 0
Joined: Thu Mar 25, 2004 7:03 am

Post by LetterRip » Thu Oct 20, 2005 12:57 am

for the multiple uv maps - Brecht is already working on it I think; for the baking of lightmaps - I believe that theeth already did such but not sure if it has been committed (or the patch might not do what I think it does.)

Hmm can't find the theeth patch, so I could be wrong - here is a script called Blender RayTrace Baker which I think does what you want.

http://www.alienhelpdesk.com/files/BRay ... beta.16.py

I'll contact the author and see if he'd be willing to allow it to be bundled in the next release.

For the grid snapping - hmm I thought that already existed...

LetterRip

LetterRip
Posts: 0
Joined: Thu Mar 25, 2004 7:03 am

Post by LetterRip » Thu Oct 20, 2005 1:03 am

See these threads for a discussion that includes the above script

http://www.elysiun.com/forum/viewtopic.php?t=51737
http://www.elysiun.com/forum/viewtopic.php?p=487380

LetterRip

reimpell
Posts: 0
Joined: Wed Oct 01, 2003 5:09 pm

Post by reimpell » Thu Oct 20, 2005 1:40 am

LetterRip wrote:Blender RayTrace Baker which I think does what you want
It doesn't. This script bakes the lighting into the diffuse texture. This prevents you from using a detailed texture for more than a single face which is unacceptable.

For example take a tileable grass texture that you use for several terrain meshes. If you would use this script you had to either sacrifice a lot of texture space as you had to copy the texture for every face, or you had to use a low resolution version of that texture for every face (i.e. you had to sacrifice image quality). In the first case you would also lose a lot of speed as you can't render the terrain in a single pass and you need a lot of texture swapping operations. The whole point of lightmaps (i.e. precalculated lighting) is to increase image quality without sacrifying rendering time due to per-pixel lighting. This can be accomplished using a lightmap with different resolution than the diffuse texture, and hence a different uv layout. If you use the tileable grass texture for the terrain faces together with a single lightmap (arranging lightmap portions according to the variation of the light on the actual face), the terrain can still be rendered within a single pass, while looking almost as with per-pixel lighting.

A lightmapper that has been used to postprocess Blender scenes to use them in Ogre is Gile[s].

Post Reply