texture my model please? [IRRLICHT FAMILIAR ARTIST NEEDED]

xDan · Post by **xDan** » Fri Apr 22, 2011 9:21 am

your physics gameplay seems an interesting concept. I'll be intrigued to see what the final game looks like

ChaiRuiPeng · Post by **ChaiRuiPeng** » Fri Apr 22, 2011 2:02 pm

thank you xDan, yes i hope when i start testing the concept v3.0 t turns out as i want it to.

in about one cycle (5 days) i should be ready to "officially" announce this on the projects forum

EDIT: okay another day started...

i hardly got as much work done yesterday as i would have liked... people in my building installing new air ducts.. lot of noise and i had to stay out of my room for the whole evening until they were done.

christianclavet · Post by **christianclavet** » Fri Apr 22, 2011 11:35 pm

Ok. Here is the model with the texture. This was a good practice as I was feeling rusty in UV mapping... I think I even improved a little...

Here what the model look now with the texture (other tweaks done to the model)

And here the model for you to download from my site: (OBJ format)
http://www.clavet.org/files/models/fluttershy.zip 649.03Kb

The model contain 3 textures.
First texture is 512x512 and is currently assigned to the model
second texture is 1024x1024
third texture is 2048x2048

I figured that 512x512 would be ok, considering the number of tris in the model (around 850 now). If you need to use a higer res texture, simply rename the needed file.

ChaiRuiPeng · Post by **ChaiRuiPeng** » Sat Apr 23, 2011 2:34 am

okay thanks it looks great

EDIT: i will try and get it into my game as soon as possible and show some screenies

christianclavet · Post by **christianclavet** » Sat Apr 23, 2011 6:06 am

Thanks! I decided to work a little bit on it, added more geometry and trying to update the geometry and the UV for testing.

The face area was too flat and the back for a women had really no volume. Tweaked thoses (adding about 150 more tris to the model) but I really like this result more than the one I sent previously. If your project can handle a character with about 1000 tris then this one is good.

I improved the hair a little by moving some vertices, she looked like a very old women, I think she should have about 30-35..

Check the face now, we have volume, even from the sides:

OUps! An ugly seam here... I will have to fix this tomorrow... EDIT: (Fixed now)

Here the new "package" with the same textures, but with the updated version:
http://www.clavet.org/files/models/fluttershy2.zip (600Kb)

fmx · Post by **fmx** » Sat Apr 23, 2011 5:43 pm

@Christian:
I know you are making this character model as a learning exercise and I dont want to deny you the opportunity, so I will give you some advice

Add more polygons to it, please!

there is no reason why you should restrict polys to 1000, you should go for double or triple that because there would only ever be ONE of those characters onscreen at a given time, so it would be worth the performance hit for the extra detail

creating lowpoly character models is one of the most difficult tasks imaginable, so dont make things harder for yourself.

Dont worry too much about her shape and physique, focus on defining important volumes such as forearms and clothing, it will ease the UV-mapping proccess by giving you more vertices to play with as well as improving the general look of the model

Please change her hairstyle!

Or just add a ponytail to the back of her head and edit the texture to show strands of hair instead of a curly bunch.

I'm sure you will get around to fixing her face in the future so I wont comment on that.

@ChaiRuiPeng:
I dont think you are doing yourself any favours by rushing to complete your game by May 3rd.

If you want your game to be good enough for people to consider buying it, you need to focus on getting nice artwork and models finished a long time before the deadline so you have time to tinker and tweak.

Having lots of nifty graphics effects and eye-candy is no good unless you also have nice looking models underneath

Assuming Christian can complete this character model in good time, what is happening with all the other art and models needed in your game?

ChaiRuiPeng · Post by **ChaiRuiPeng** » Sat Apr 23, 2011 6:22 pm

fmx wrote: Having lots of nifty graphics effects and eye-candy is no good unless you also have nice looking models underneath

haha

the thing is i am not focusing on graphics at all. if you have seen some of my screenshots i think that should be fairly obvious. i dont know how that you came to conclusion.

i intend to deliver the lowest grade graphics possible. why? because it annoys me how almost all the game devs i meet are so obsessed with all the nice effects they are able to do put in. things like SSAO, post-process effects, it seems like they are just adding whipped cream to an empty platter.

(

YUUCCKK )

the current poly count on the model is fine. in fact the only graphics that are going to be visible for now on the character model are the debug graphics cubes for a while. that show different transforms, but that is all i need for the next cycle.

i know it sounds dumb that i want to push out by may 3rd, but that is the only way i can get any work done. i am one of those tragically lazy people, unless i am constantly pushing my limits i am going to fall back and slow down.

i am pretty happy where i have gotten with my game architecture as a whole. i know the screenshots are limited but speaking for whats under the hood if you had told me two hextets ago that i would have succesfully implemented some of the stuff i had i would have trouble believing.

the may 3rd release is not just about the money, which i probably wont get, but just trying to set benchmarks for myself. just an exercise

to stretch my mind. you can't learn to win until you have lost a few times.

EDIT: as far as other models go, there are only dynamically generated meshes, and simple triangles meshes that will rarely be used.
the only effect i plan on putting that can be considered "advanced" is a spherical harmonics shader to provide all lighting

christianclavet · Post by **christianclavet** » Sat Apr 23, 2011 6:51 pm

Hi FMX! Thanks for the comments!

That exercise was good for me because it's the first time I can "pelt" something correctly. Before, I had to use planar mapping and it was really time consuming. (one of the reason, I was staying away from characters before, as ships are easier to UV maps since lots of areas are flat)

I do like the look of the hair now since I fixed the bottom back that is more what I was looking for. Doing a polytail, would ask ChaiRuiPeng to do some physic affected bones that would move the ponytail... I don't want to impose that... From what I've read so far, this character will do some "ninja" moves!

For the polycount, the reference model that I was presented with contained only 200 tris that was a bare minimum for a non-moving object (statue like). I asked for more polys (and it was accepted). Doing a 20,000 tris object (UDK quality) mesh could be dangerous if the model only goes to an old IPhone... Already I'm about 5 times the tris count as the reference model!

Also If I'm doing a model with that high polycount, i'm also assuming, we're going for ultra-realistic model (ears, all fingers, etc.) and that model would need a normal map. This is a totally different ball game here. Were talking about a week or more of modeling work! And ChaiRuiPeng only asked for a "representation" of an human form.

This is an exercise for me and a opportunity to help a forum user... Not a paid job...

Anyway thanks for the comments, I hope that explaining the context of the work did not offended you...

ChaiRuiPeng, I updated the model (subtle changes made to some vertices), that maps have all been fixed for seams (I did not see any visible seams this time). Just download from the same link as before, I changed the archive in my site...

If you have problems when the character moves then just tell me where and I will insert more edge loops to facilitate the bending.

ent1ty · Post by **ent1ty** » Sat Apr 23, 2011 8:44 pm

ChaiRuiPeng wrote:...it annoys me how almost all the game devs i meet are so obsessed with all the nice effects they are able to do put in. things like SSAO, post-process effects, it seems like they are just adding whipped cream to an empty platter.

Writing a vast game where only the structure is a real challenge - boring
Writing a 150 line shader with some (relatively)complicated math - challenge + eye candy

At least that's the way i see it

Other part of game developing i find interesting is AI

fmx · Post by **fmx** » Sat Apr 23, 2011 8:50 pm

hey its cool, I was only offering advice, both of you can decide what to do with it for yourselves

@Christian
no reason why I would be offended

I totally understand the requirements and your intentions, do the best you can and use the experience to do better stuff in the future

@Chai
understandable too, something is better is nothing.
Yes I thought from your spherical-harmonics topic you might be going all out on shader overload, but obviously not

xDan's Puzzle Moppet is a really good example of what can happen when simple shaders and effects are applied to an already nice and clean scene in irrlicht:
the visual impact gets boosted tremendously, almost like multiplying a multiplier

If not this time then maybe for your next project you can try to focus on entering the screenshot competition or something

Good luck with the project anyway, I feel tempted to help but I already have a lot on my plate sorry

ChaiRuiPeng · Post by **ChaiRuiPeng** » Sat Apr 23, 2011 10:15 pm

thank you for taking no offense

i never intended to inflict any, i was just stating my preferences

@entity

i can see why that would be your preference, as this is a more graphics oriented forum.

everyone thank you for opinions, views and wishes of good luck.

i can't guarantee results you will like but i can guarantee i will put my best efforts into making a good _game_. if i cant make the deadline then i will extend it and make sure only my top notch quality work gets through.

christian thanks for all your help.

now.... back to work

EDIT:

i revised the to-do list
to-do before formal announcing:

finish prototype of player v3.0
fix a nasty insert keyframe bug, that occurs when you insert a keyframe in the bottom half of between two already existing keyframes.

btw i got rotations working

to-do post-announcing:

XML level saving/loading
more platform primitive shapes
construct platfom from set of 3d point (compound/convex hull)
texture applying

christianclavet · Post by **christianclavet** » Sun Apr 24, 2011 12:05 am

Hi, Done a quick test by using a Max Biped and skinned a little the character.

It seem to work, but you need to be precise when you skin the mesh. Each vertice count! Bending is looking like the character in a Quake game, so it's about the same quality (quake). If you want to improve the bending more, we would have to increase the tris count by adding more edge loops...

Here a screenshot of the mesh posed in a running stance:

shadowslair · Post by **shadowslair** » Sun Apr 24, 2011 12:26 am

Just to say that no normal person is running this way- usually when the left hand is forward the right knee is forward and the opposite. It`s a matter of basic biomechanics, because this is the way we generate the impulse to move and keep balance. And this is something every beginner animator has to be able to imagine and have a flair for.

serengeor · Post by **serengeor** » Sun Apr 24, 2011 8:41 am

ent1ty wrote:
ChaiRuiPeng wrote:...it annoys me how almost all the game devs i meet are so obsessed with all the nice effects they are able to do put in. things like SSAO, post-process effects, it seems like they are just adding whipped cream to an empty platter.
Writing a vast game where only the structure is a real challenge - boring
Writing a 150 line shader with some (relatively)complicated math - challenge + eye candy

At least that's the way i see it
Other part of game developing i find interesting is AI

Applying shaders to a good structured game is a lot better than applying them to nothing. Game with neat effects and sucky gameplay is just a sucky game, while a game with good gameplay and bad graphics is still a good game.
Well if you'd combine both then of course the game can become really neat.

ent1ty · Post by **ent1ty** » Sun Apr 24, 2011 9:21 am

Off course... that's my problem actually

I was just saying, i like messing with gfx effect more than coding gameplay