Ogre Forums

I have seen for a while the general feelings that good artist works are hard to come by (Unreal 3 demos anyone?) But then I also see some of the work done by many people on the www.elysiun.com and other boards, and wonder if the two projects could perhaps benefit each other.

Perhaps through some advertising on such boards, we could attract some artists to come and make public domain art that could then be shown as Ogre Glamour shots (imagine Unreal engine with coder art )

As an example of what I mean, look at some of the things that these boards can produce:
http://blender3d.org/cms/typo3temp/80a6b3d958.jpg
http://blender3d.org/cms/typo3temp/35715d7d3f.jpg
http://www.thirdwave.de/pics/3wcruiser.jpg
http://www.elysiun.com/forum/viewtopic.php?t=36491

I think we can all agree that this has more 'glitz' than perhaps the ninja and ogrehead models

I think if we drew up an enticing proposal (IE: we would showcase their work and plaster their name all over the overlays and everything) if they would donate their model to the public domain. Another key would be making sure and explaining what Public Domain actually meant and the ability to which they could retain credit of their work while still allowing use in Ogre and other projects. (not just throwing it out willy-nilly on the internet)

www.creativecommons.org should come in handy in that respect.

Good idea; but I have a few concerns.
Artists/modelers are generally perfectionists. They see to it that their work goes from design through production and out the door. However (I might be wrong on this, so please correct me if I am ) Ogre ATM doesn't have steamlined tools which can facilitate beginning to end work flow, but rather different components and tools which one must learn to use to go from beginning to end. I think most Artists/Modelers look for tool-sets or a single utility which can retain the same look and feel to assist in a smooth-as-possible work flow. So IMO I think what must happen before such advertisements is merging and fabrication of utilities which can go from the beginning to the end, start up ogre and let the artist view his/her creation without the artist ever having to tweak the configurations of Ogre (materials, resource, particle, shader..ect files) or even having to worry about moving it to the right folder. If you look at the commercial Game SDKs, they provide a transparent work flow, meaning that the application handles everything. So the artist can concentrate on their art rather than muck around with the configs. Such utilises can be designed on the same fundamental concepts that Ogre is built-on.. modularity. So various packages for various applications can be distributed without the utilitises having to be rewritten.

Next is a silly concern but what the heck; those links you provided to those models, wow.. they look terrific. But I have doubt that any "decent" hardware can render those while maintaining decent FPS Thats also a major question from Modelers "Whats the maximum polygon limit?". As we know each particular hardware configuration can pull a different amount of polies; but I think there should be a benchmark on various "decent" configurations. An Ogre Stress Test if you will, which would test all features on the hardware, and barf a simple summary which the artists and 3D modelers can feed on so they get an idea as to how much polygons to limit a certain model to.

Good point, The .material and possibly shading scripts are a pretty big turnoff to most artists that I know. I see the Mesh Viewer as a pretty important tool that is luckily there, but I think a materials editor is what is pretty obviously missing here.

Perhaps if we could do a partnership, wherein coders could help the artists create their desired effects through .material and shader scripts and start a back and forth dialogue with a strong cooperation for a finished product.

As for the polygon numbers, yes those are huge (probably) although in a scene alone I can see it possible to model 200k tris at a time which should be doable for most modelers (with normal mapping esp.) Something like the characters in Unreal 3 screenshots I have seen (showcase of one entity with large poly counts and texture quality).

Now, I like the idea, in general, but there are going to be one or two specific problems.

One of the problems we've had is artists making great content.. that is entirely impractical. No textures, all vertex colors. Millions of polys. You get the idea.
Something you need is not only good artists, but experienced artists. For example, our main modeller here now (RyanN on this board ) can not only model something that looks great, but can do it in such a way that it looks better then it should, if you understand what I mean.

Some-one said to me the other day (month?) "Oh, you can poly-crunch your models by up to 80% before you even see it, and they will be much smaller." We were, WTF, that's not the way we work here. If you have that much spare detail, then it shouldn't be there in the first place.

As for if .material is too hard for modellers, again, not in my experience, a.k.a. RyanN. I think some-times we under-estimate their abilities. Of course, they wont probably wont write their own shaders.

And (cc) is a Good Thing, especially with the sub-variations. We (our company) has done some stuff relating to the Creative Commons, in the past.

So, to summerise, good idea, but be warned, it won't be good render abilities == good modelling abilities.

Good point, perhaps we should contact elysiun (or even blender.org) and try to arrange a contest with set rules polys < 10,000, textures must be exported to .material files, final screenshot must be made from the Ogre meshviewer etc..

Then we could use the laws of competition to determine who has the best modelling skills for us I'd imagine that we would need a prize of some sort though (unless the sponsorship would be enough?).

I'm willing to sponsor a series of contests ($) to have Blender created mesh/animation for Ogre.

I think some stunning quality art work shown in Ogre demo program is sorely needed

What I can't seem to find anywhere is a GOOD looking sample program where:

1. The Blender files are released as part of the Ogre set. Well organized so us programmer artists can understand what we are working with.
2. High poly bump/normal mapping done at various levels. Maybe a demo of the same model done at 3 different poly counts so we can actually see how they perform in Ogre?
3. Basic animation... such as a biped or quadraped walking.
4. Realistic scale and shape. If you model something, model it to match the size/shape/proporitions of the real world object.

Three things I can think of for good objects: a) a stock car that everyone can recognize (Porsche?). b) A biped or quadraped. c) A set of 6 low-poly trees that don't look low-poly. Pine, Palm, Oak, etc.

A series of contests?

Yep, doing the equivalent of crystalcore (flagship project, sort of techdemo for crystalspace) for ogre should be a nice idea

RoundSparrow: Really, you would be willing to pay the contest award? Perhaps this could become a reality then

I agree with the categories although the tree one I think should probably be omited as Sputnick made an excellent auto generator here: http://www.ogre3d.org/phpBB2/viewtopic. ... +generator and as for the organic, something like would be 'nice' http://www.unrealtechnology.com/screens/p_bezerker.jpg IE: a serious character that would display normal mapping etc.. to a good end.

As for a car model, I guess this could be put in if desired although it would be somewhat reliant on pixel/vertex shaders for the car paint, so perhaps might be aided by programmers?

What I would imagine for the demo is a series of keys (or buttons) that would allow you to turn on and off: normal mapping, stencil shadows etc.. so as to see the visual/performance boost or detriments each feature adds.

I don't know about a CrystalCore type demo, I think that once ppl see that Ogre can render individual items such as those above, we shouldn't need to show ppl any more capabilities of the engine (beside's its only REALLY a graphics library, everything else would be other libaries, Yake on the other hand, might need a demo like this ).

Well if anyone has a good idea for an Ogre demo and is good at art (I mean, there's enough ideas already just noone to actually bring them into practice) then I'd love to do the programming aspects.

Maybe someone can port the demo below to Ogre (if we are allowed to use the graphics)

http://www.flipcode.org/cgi-bin/fcartic ... 3&id=64208

Nahh, we want an original demo!

Back in the days when the Genesis3D engine was still around they had a demo showing off all the capailities. It had multiple scenes and was completely non-interactive. I think we could do the same with Ogre.
You know, like all these little programs that are spawned by the demo scene. Little non-interactive, realtime rendered movies, showing off all their l33t techniques and stuff
3DMark is another good example for the type of demo I have in mind...

Having one demo that shows in an aestheticaly pleasing and flashy way most of the capabilities of Ogre.
Coming up with guidelines for the contest is hard, however, since most of the flashy things that can be done with ogre aren't ogre core features but the creative use of those features (shaders and stuff).

Some things I'd like to see in the demo:
lots of nice shader effects (normal mapping, displacement mapping, other mapping stuff)
dynamic textures
mesh deformation (like in the water demo)
mulitple viewports and render targets (render to texture, etc)
terrain

And all of that with high level artwork of course.

I already have some nice ideas on how to incorporate those features but I'm lacking the time and experience to actually code all of this. I'd probably need about half a year because all of the learning. I could, however, probably come up with a script (script like in storypoard, not like in Python)...

Come to think of it... I've been looking for a doable project to get me started with 3D programming again, this might be a good start. Only thing I'm lacking is (again) the artwork. I'd need several well done models and textures.

Oh well, I'll get to the programming anyways, jsut for the heck of it!

bana wrote:RoundSparrow: Really, you would be willing to pay the contest award? Perhaps this could become a reality then

Yes, I think start small to build interest. Spread the $ out over a series of contests (and a few free ones to boot).

I specifically want them to be Blender artwork. I'm thinking US$25.00 for the smallest contest... I'm willing to put up as much as $250.00 in total...

I will want some influence on the topics of some. But not a big issue, I just want to see good artwork done in Blender and used in Ogre.

My desire is that the artwork created be entirely BSD style license... free to use commercial or otherwise (Creative Commons, "Attribution"). Since we would actually be paying the winner, this seems fair? Maybe a less restrictive terms for the other entrants (non-selected winners).

RoundSparrow wrote:Yes, I think start small to build interest. Spread the $ out over a series of contests (and a few free ones to boot).

I specifically want them to be Blender artwork. I'm thinking US$25.00 for the smallest contest... I'm willing to put up as much as $250.00 in total...

I will want some influence on the topics of some. But not a big issue, I just want to see good artwork done in Blender and used in Ogre.

My desire is that the artwork created be entirely BSD style license... free to use commercial or otherwise (Creative Commons, "Attribution"). Since we would actually be paying the winner, this seems fair? Maybe a less restrictive terms for the other entrants (non-selected winners).

Sounds good to me, to get started, I think a elysiun centric contest would be good (IE: thread where submissions are placed) with news posts on blender.org and maybe blender3d.org, if you want guidance as to a good framework for contest rules and design, check out Cgtalks forums as they have some contests with good ideas (IE every submission must have a WIP thread to show the progression of the model (and to prevent someone just ripping it off another project).