Warai otoko

Code and custom binding

2011-06-26T13:19:00.000+01:00

I was supposed to post some environment sketches but soemthing big happened during this week and I had to start working on my game.

So I'm using a azerty keyboard when the rest fo the world use a qwerty one. So look into key binding in-game to remap the key. After looking for hours I finally found out how to do so;

exec function GoQwerty()
{
local name key01, key02, key03, key04;
key01='a';
key02='w';
key03='q';
key04='z';
PlayerInput.SetBind(key01, "GBA_StrafeLeft");
PlayerInput.SetBind(key02, "GBA_MoveForward");
PlayerInput.SetBind(key03, "GBA_ToggleTranslocator");
PlayerInput.SetBind(key04, "GBA_ShowCommandMenu");
`log("Qwerty keyboard");
}

exec function GoAzerty()
{
local name key01, key02, key03, key04;
key01='q';
key02='z';
key03='a';
key04='w';
PlayerInput.SetBind(key01, "GBA_StrafeLeft");
PlayerInput.SetBind(key02, "GBA_MoveForward");
PlayerInput.SetBind(key03, "GBA_ToggleTranslocator");
PlayerInput.SetBind(key04, "GBA_ShowCommandMenu");
`log("Azerty keyboard");
}

that's a easy way to do then you just nneed to press tab in game and type one of the command GoQwerty or GoAzerty.

I put that code in my playerController.

t's quite easy to understand and that's the basic to key bind with unrealscript.
Careful when ever you change those key bind it's save in the config file.

value sketch

2011-06-22T02:43:00.000+01:00

Agian some value sketch.

Try to get some job done but I quite busy at the moment. I started some env painting I'll update them tomorrow. Nothing crazy just some of my experiements.

painting

2011-06-17T14:05:00.001+01:00

I getting more and more comfortable with my tablet and I just finish a new value painting.

little sketch

2011-06-15T15:10:00.001+01:00

I did acquire a trust 7300 a few months ago and since then I always regret it. My bamboo pen was way better in term of support and ease of use. I had proble with my dual screen. I had problem with windows tablet support. I had problem with the stylus too and I just felt like giving up. After I got a new stylus change my dual screen configuration and look for new drivers. But I manage to make it work. I was on my way to by a motion computing le1600 or le1700 but the fact I did not know if the screen could be use it as a screen on my desktop pc so I look into other option like diy cintiq and such.

But I manage to make it work. It was painful and I would not recommend it to anybody. some intuos3 are available on ebay and maybe that's better than a faulty tool.

So this is a value study I did this morning.

Canvas, Scaleform ? Dochi da ?

2011-06-10T21:45:00.000+01:00

So I'm working on new project (yeah a new one) and I was kind experimenting the Hud which is really important for that game or rather the feel of the game.

So canvas or scaleform. Scaleform sounded like a nightmare. Having to learn flash ... Ouh ... Right ? And Actionscript 2!!!

Where to find a simple video tutorial of action script 2 ?

Exactly, nowhere. So was just left alone with the silly adobe page learn as2. What a joke.

Canvas in another was a good surprise. Fast, really fast to pick up and well integrated with unreal sounded like the perfect pick so I was wondering with Scaleform is getting more attention ? I still don't know since I haven't finish exploring. I'm still learning canvas and I haven't built anything near a decent hud or menu but I just learn on the forum, scaleform 4 is coming to udk really soon. that's good news.

I mean why bother learning an outdated language. Anyway that's the good news.

Scaleform 4 also bring optimisation and such.

The thing I don't like about scaleform is how clumsy it is to get anything working. SWF files, kismet, actionscript, unrealscript. That's a bit too much if you want to work efficiently. I still want to try both to see what works the best.

But from what I can tell canvas may be a little bit complicated for a complicated menu. My new project doesn't really have a one so that's a good thing.

outline without post process

2011-06-09T10:31:00.000+01:00

I did a sort stencil outline without using buffer (you can't access it in udk) and it turn out ok I guess. It's object based and that's pretty cool.

Here the image. I just need to make it glow a little bit more but that goo for now.

Not dead massive update

2011-06-09T10:02:00.000+01:00

I've been active those last six month but I couldn't get any time to update or post on my blog. Also I was so caught up in several project that I could not create anything personal.
so some pics of my latest models. Really low poly this time.

The Hurling player

2010-10-31T10:29:00.000Z

busy busy busy !

2010-09-23T11:53:00.000+01:00

I've been busy but I still manage to work on zbrush.

So this si one of my last work. It's not finish, just a rough version of a sportsman but I4m pretty happy with the initial result.

Basically my approach was to stay as low poly as I could. Even now the model still quite low poly ( 19k) of a zbrush model.

asset update

2010-09-05T00:27:00.002+01:00

So I've been busy working on zbrush-maya-udk workflow and this is the result of my experimentation.

The tree was made inside zbrush using zsphere and zsketch. A really good workflow and I might come up with the way I did it in few days.

those stairs were a little bit difficult to bake on a plane so I decided to use topogun.
Zbrush still has some flow for game artist like the fact you can't bake normal map from different models.

but still I'm really please by the result. I may keep working on little stuff for a while just to get a really good understanding on what can or cannot do within zbrush.

divin' into zbrush

2010-08-30T06:33:00.000+01:00

so this mesh is not impressive by itself but the technique I use is quite new to me. I use projection master to create the detail on the wall then I texture it, apply a mask using the texture intensity and inflate the mesh to get that look.

I did the texturing using spotlight.

this is all testing at moment but the main goal is to get more use in working in zbrush for some stuff that actually aren't real sculpting.

Playing with Zbrush again

2010-08-29T20:23:00.000+01:00

I finally got my ai kind of working and I started working on some art, but first I need to sharpen my skills. I've been lazy and obviously I became a little bit rusted so I working on zbrush new tools and in the process trying out the different way to creat art for my game. So there is a rock with 256 poly. I create a rock and then using the zphere I created a low poly version of it then project on a cleaner mesh.

Tiled floor

2010-08-29T06:58:00.001+01:00

I just try to create a tillable texture floor in zbrush and this is the result in maya viewport.

Kakurembo Update

2010-08-25T01:12:00.000+01:00

Since I'm not participating in any contest I decided to take more time to create the game and add several mode.

So basically the game will have a solo and coop campaign. The game will have a storyline with main characters and some rpg aspect. But I'll talk about it when I start implementing more some of those mode. For the time being I might focus more on the hide and seek multiplayer mode. I decided to create 5 differents map with each one an identity. Maybe set different time of day or climatic conditions. I still need to finish a lot of stuff.

But basically I've planning a lot lately and I might update the overall design and such. So now since I have an AI working I'll start working on the main game mode and the highscore and such to get some game working and then from there start doing the game itself.

Not really happy about that

2010-08-22T19:47:00.001+01:00

My game didn't get selected for the intel game demo contest. I really think the contest itself was bizarre since it's based on game ip and not on content.

But anyway I don't know if I'll continue Kakubo till the completion but I might continue to work on it for the time being.

AI bot working

2010-08-15T08:02:00.000+01:00

I just finished my AI layout code and I'm pretty happy with it. I did a lot of progress with the ai and the unrealscript in general. So basically my current bot is able to spot other bot in it surrounding while running, spotting item and cover volume. That ai is pretty smart so now I just need to make it interact with its surrounding during the game. But it already pickup item and stay idle at hidden.

the final Ai should be able to travel across game area, to spot the seeker if he's close enough, look for hide out and wait until he's gone to start moving around again, collect item and use them again foes or enemies.

I'm pretty confident about accomplishing that task which wasn't that important to that game because the final product will be aimed to work online or splitscreen. But still a really good practice for my futures games Krysalid and skullz this is a really good exercises.

Intel is finally announcing the finalists of his game demo contest. I hope I'll be in it because I didn't get into any this year so I would be glad to just get a shot at that contest because one of the prize is actually a trip to japan during the tokyo game show.

taking too much time solving code

2010-08-13T03:00:00.002+01:00

So I've finished last week setting up the nav mesh in udk thank to dungeon defense source code available on udk forum but I still having trouble integrating a dynamic choice maker. For example while the character is doing the pathfinding check top surroundings to see if there any cover or item.

Still far from being finish but I got the new zbrush so I'm distracted. :)

Dealing with navmesh

2010-08-08T03:28:00.000+01:00

It's a hell. I manage to make it work but there still some bug on few maps I still don't get it all.

But the game is moving forward with more solid code.

I'm using the layout package to get as a placeholder pretty useful.

I started to work on shader pretty handy. I might work on sort of sky shader to make the light change during the game.

Maybe create some thunder or something.

unreal script progress

2010-08-07T00:09:00.000+01:00

I've been working my code those this last week and it hasn't been easy but I prefer coding than using kismet. Kismet sound like a great tool but not so much for complicated task as gameplay.

So far I've been able to create a simple ai for the bot and the basic layout of the gameplay.

I start working on vertex painting and I'll post update in the following days.

Looking forward zbrush 4 release on monday.

cheers.

Giving up Kismet and working on the code

2010-07-31T18:50:00.001+01:00

Baiscally Kismet is not for me. Of course simple task can be done thourgh kismet but I just feel like I need to practice more to get my head around it. Basically I feel like it's easier to do it though code that using kismet.

Anyway I won't be on time I guess but I'll try polish my result to get ready for the intel game demo contest.

cheers

Media Update

2010-07-31T18:48:00.002+01:00

I've been quite active lately but not posting anything.

So apart from not being ready to submit my game on time for the indiepub game contest I have few model ready. So this is the update.

I'll update more later. I've a lot of picture of the asset I created for the game.

cheers

Getting back on Kismet

2010-07-29T21:43:00.000+01:00

I never got my head around kismet so it's a bit challenging to create a gameplay with it. But is just what the new version of kakubo will be about. I can't take the time to sort it out in unrealscript but I can try to do it in kismet at least I hope.

I'll keep and update and some screen soon enough.

cheers

Back on udk

2010-07-28T03:42:00.000+01:00

Because I Didn't have the energy to learn a new tool and because I've been playing splinter cell lately I decided to drop unity for the moment. I don't have that much time left for the indie pub game contest and It would be better to have a proof of concept ready on udk than a half finished one on unity.

Unity is indeed easier and everything but I do miss a lot a feature. I did learn a lot of stuff on networking and everything and that basically the game might but I'll come with all my update in the incoming days.

roller coaster game

2010-06-24T22:09:00.000+01:00

A client ask me to create a model for his really good roller coster game using iphone and unity. Few screens.

Kakurembo early screenshots

2010-06-22T20:57:00.000+01:00

So this is the first shot of kakurembo so far nothing really exiting but its just the beginning. I have been busy with few jobs but I hope now I get the time to work on my thing.

low poly bench

2010-06-10T03:06:00.001+01:00

Ive been ask by few people to work on low poly so here a first try with a bench

390 polys.

Animation

2010-06-08T04:27:00.003+01:00

New animation test with maya. Just learning few new tricks. I my start to come up with new stuff in animation.

New tools

2010-05-29T15:26:00.000+01:00

I came across two new tools which are great and free. So the first one is sculptris. It's a good alternative for sculpting if you can't afford mudbox or zbrush, but it still in alpha version. The second one is Alchemy
http://al.chemy.org/download/

Really good one for concept artist.

This is an extract of the new blender foundation tutorial.

Scaleform supports on UDK

2010-05-26T03:19:00.002+01:00

I haven't use the ui that much on udk so I really don't know if it was that bad, but from the post on udk form I would say that a lot of people were complaining. Anyway now scaleform is integrated with udk which is a great new.

The May release just came out with scaleform and steam support.

starting a demo game

2010-05-24T07:08:00.001+01:00

So I've spent a the last week following tutorial on how to create game using unity and now that I feel a little bit more confortale with I'll start to create a really simple game demo. So the basic concept behind it is to create a shooter game like gigawing in a week. It's really a concept than a game but I need something challenging but still abordable for a beginner to understand better the api of an engine.

few days away and now I'm back to business.

2010-05-23T21:02:00.002+01:00

I was a litte bit annoyed by all those problem with my new mac. I mean most of software don't work at all. Maya is so terrible with the new leopard I'm considering of learning blender to do small task when I'm not at my place and I need to use my laptop.
But what was really bothering me was the fact visual studio was only on pc. But I found something really interesting. Using mono develop you can sort of plug unity into the IDE and use the intellisense.

can find some information here

http://forum.unity3d.com/viewtopic.php?p=219018

This is really good

2010-05-19T20:09:00.000+01:00

So I just found out about a game contest which you can win 100 k $. It's the independent developer game contest.

This is a lot more than the Dreambuild on and even if the contestant will be though using unity can make the job really easier for me. I was planning on participating on Level up Game conest and try to get the trip off to tokyo, so it's cool to see more than one contest to participate. With unity I can really focus on the content and not the programming. I'm planning to get into each with kakurembo which is the hide and seek game. There is a lot more going but I'd rather keep for the last moment.

I hope I'll get some art done very soon. At the moment I'm focusing on the networking and the api of Unity.

Lazy me

2010-05-18T17:45:00.000+01:00

So by lack of time and also because it was more funnier to just look for fancy effect in unity indie I started to look for an animated shadow. So the main option that come to mind were either stencil shadow without using a buffer (unity indie cannot access it ) or animated decal texture. I though the animated texture would be the easy way but then I discover the shadow volume on the wiki which is greta but just can't work with what I had in mind. I try stuff work around and modifying the code but nothing worked properly so I switch back to animated projector. I found a fairly good example of it in the invader demo. Even though the technique they use is limited (the code actually set manually every texture one by one), it gave me a fairly good understanding of what I need to create my own code.

So I didn't gave up on getting some kind shadow using unity. I was looking on doing realistic shadow but maybe I'll stick to more stylish and less accurate one.

So I didn't really made any progress on understanding unity api or so but I still I know what it's possible r not.

Switching from udk where the shadow just got the updated to unity is really annoying because I kind miss all the awesome features but because of one (Networking and web deployment) I might stay with unity for kakurembo which is multiplayer oriented.

Day 2 Unity

2010-05-17T19:21:00.000+01:00

After just one day working with unity indie I have to admit that the application is really good. I really choose it because of the networking option and it fits quite well to my need but the ease of use is really awesome. I mean creating small game and prototyping really fast is a blessing. There is not such a thing using udk. Creating a menu is an hassle even with the incoming scaleform I still think unity will have the advantage over epic's product. The other engine I'm waiting for is the cryengine 3 sort of indie version. They are going to release the application and you'll be able to create game with it. I really love Sandbox cryengine for the powerful toolset and graphics. But UDK so much progress in the last month I don't really know which one will be suited to create Skullz. I'll wait and see meanwhile working on Kakurembo.

I'm really thinking the concept through and I believe people will enjoy playing it.

I still have a project in mind called Armance, which is a great scale game and of course the cryengine 3 will be my favorite choice for that one.

the indie market has really change the last four year, and I think it's a really good thing.

Working on a new project using Unity Kakurembo

2010-05-17T01:00:00.000+01:00

I haven't done much lately. Why ? Too busy with film school, moving back to london and a lot of stuff who have been happening in my life. But then it was already mid may.

So I realize it was almost june and I was far from finish Skullz. I was working on the props and I didn't work that much on the character and the script so I just figured out I should start a small for the incoming competition I'm entering by the end of june.

So basically the game itself is called Kakurembo and it's a multiplayer hide and seek.

Unity has a great networking solution so I figure out that maybe I should just take advantage of it.

I don't have much picture at the moment but I'll start sketching in the next day. Right now I'm leaning unity and I have to admit even if the tool set is not comparable to what udk can offer the ease of use and the scripting are far easier. I start programming in c± so it's quite easy to just understand and also javascript is quite close to .net language.

Anyway I hope you'll like my new project

Shooting range

2010-03-20T13:29:00.000Z

So in order to progress and keep doing experiment I just start a shooting range level. This is a ancutal game sequence it'll exist in the final game but this level is temporary. I just want to start working on kismet and I did this level as an exercise. the final version may look a more different.

Digital Painting

2010-03-16T15:40:00.002Z

Just a sketch I've done today

Not much lately

2010-03-16T09:59:00.002Z

I've been working on some code so I didn't do any progress. I'm also trying to improve my skill at photoshop, to create some nice artwork and design without having the trouble to scan my drawing and then start working on them.

So this is my latest work on photoshop. Still coming from a tutorial but I'm pretty happy with it.
I really need to focus on photoshop for a bit to create sketch before I even start modeling my meshes.

Ruins

2010-03-03T02:02:00.000Z

So lately I've been investigating what are zbrush possibilities in terms of environnement modeling. So I Created this scene. I took inspiration of Gnomon masterclass. But most of the ryan kingley didn't fits my need. ( I'm already too use to my own sculpting method)
So this what I came up with. Not much modified version or the original scene but still did it my wa. It was a good exercise. All the asset have been created within Zbrush. So the conclusion I came up is zbrush is not fit to created level from scratch. I mean level assets. I still need to create the model inside maya first.

Head Texturing

2010-03-02T03:33:00.001Z

I'm slowly starting to understand the process for realistic texturing using polypainting.

So this is the 4th attempt.

Still not impressive as using image projection but I really think it will be more efficient on different aspect.

progess on texturing.

2010-03-01T09:06:00.000Z

So this is my third attempt to get realistic texturing with polypainting. I still far from what I'm aiming. Still learning the process.

Polypainting

2010-02-27T06:57:00.000Z

So here some poly painting I try tonight. Just an hour of work. Maybe less. I was fooling around. But the main point is how to get realistic tone, using temperature zone technique. I really like the way it give a realistic skin. I still need to improve my skills. Combining it with ambient occlusion make a really great and clean texture.

I may come with more character in the couple of days.

Two post in a Row : little Update.

2010-02-27T05:31:00.002Z

So I did this one a while ago. But I didn't finish it. I was trying some polypainting technique. From what I can say is the detail aren't obvious and it's less precise than image reference but the color and the tone aren't wash out from distance. In fact it feel more natural from some point of view. My model still need some refinement but I4m pretty satisfied with it. I've already finish the normal map but I wish to unwrap properly to work the spec and diffuse map on photoshop.

I started working on the story and it's taking shape. I've finish the design of the main enemies and I my post my sketches.

Forgot to say it's the normal map version. check the polycount on the last picture. 6k !!! Pretty nice ?

Simple pillar

2010-02-27T05:17:00.000Z

So still with experiment for UDK. So I may have found the ambiance for my game Skullz. Sort of eastern European punk. It's quite vague but I've gather a lot of reference of Milan ( industrial town but with some history) and Turino ( the same) I visited a while ago.

So this is a simple pillar not really related to that but since unreal level design is about small asset put together I though I might to start with something. I probably won't use this asset but It's a nice pillar. The purpose was more about understanding how to get my asset from zbrush-Maya to UDK. I used photoshop instead of polypainting.

I did pretty well tough I still need to work on the specular map.

Again, it's nothing mind blowing just testing out udk and my pipeline. I might post something using polypaint. I still need to know which one is the fast way to go.

speed painting

2010-02-25T07:45:00.002Z

I've just started to draw with my tablet and this is my first try using photoshop.

Procedural building with custom meshes

2010-02-25T01:09:00.001Z

Just trying out some old building with the new tool in UDK. I still wondering if I should integrate it in my pipeline.

Senate

2010-02-22T18:17:00.000Z

I have a level called the senate where you are in a middle of a huge parc surrounding the senat. I thought it would be a huge opportunity to use terrain and speedtree.

I'm far from being finsih and I won't probably do any major improvement in the next week but it was just a try. I manage to use geocontrol and udk so bear with me. It's by no mean a finish level or anything.

Procbuilding experiment

2010-02-22T06:52:00.001Z

Just showing off some of my experiment using procBuilding volume.

After testing out for a whole day I came with some really good result still messy but I get my way around it. I jst started modeling a sort a city using udk content. It's really difficult to come with womething interesting sing the same pattern again and again. But I start having an idea on what I'm going to need to create my city. A lot of mesh and building part but the Procbuilding will allow me to place it easily.

Level design

2010-02-22T06:46:00.000Z

I've been working on level design or to be more precise unreal level design. Bsp modeling was and still a bit vague for. I want to make the best use of the tool available so I still trying to figure out what the best way to go. So as an exercise and also a good way to gain experiment I started working on tutorial.

So from those experience came this level

I follow 3dBuzz and Eat3d tutorial on level and I got some good insight on how to build a level using the puzzle method of the engine.

I'm still thinking on how to integrate Zbrush onto the process.I which to sculpt and texture everything inside zbrush using poly-paint noise and zapplink. It sound feasible but I still looking for the fastest workflow as I'm planned to do a entire city planning ahead is the best I can do.

Little preview

2010-02-11T16:21:00.001Z

Little preview on my incoming game, Skullz.

I started that project a long ago but I have to ditched it for several reasons.

this is a link of my first environnement.

http://forums.xna.com/forums/p/10562/55425.aspx#55425

The game changed a lot and now it's not a men but a women who's going to hunt down the enemy.

Anyway Krysalid is on hiatus for the moment, UDk is not really good for combat game or at least I4m not experienced enogh with it to come up with a good gameplay. In another hand I've been playing around with the unrealscript and I managed to set up a tps with the original skullz character.

Without any more delay the picture.

more dramatic lighting

that's it for the moment. I may start working on the game play and the enemy design.

update on my latest work

2010-02-11T16:04:00.002Z

long time without updating. christmas, grilfirend, family boredom, so much reason for me to not work during a long period but I here to update my little blog.

Yhe point is I did work a bit on a side rpoject a hurling game demo with a small team and I decided to post what I did for them

Nothing too fancy but I kind progress on texturing (in terms of speed).

Some sketches

2009-12-21T10:43:00.000Z

I've found some sketches I've made when the game only in early stage. I found it interesting to post them.

So much progress but so much to do

2009-12-08T05:28:00.003Z

So I'm going to talk too much about how I did it. It's more some experiment but, I'm using a lambertian toon shader with an outline. Since the game was supposed to be realistic I still have some trouble figuring out what art direction, I4m going to take.

I just finish the shader and I'm pretty happy with it. The material editor is a real force for game development. I know a bit of hlsl but I wouldn't be able to recreate such a complex shader with it. So the cloth are made with a different shader that the skin. I use and environment map to create an highlight and some specular map and ambient occlusion map. It give me a really nice look on the cloth. I try to create a spec map accordingly to the fabric.

So this character so is going to be change. I'm starting a new design for him. Someing less realistic or at least a bit more manga. I'm not looking for something cartoony just someting like realistic manga from mamoru ooshi. or satoshi kon. I still don't know where this game is going so I may do a lot of tryout to found out what art direction is more fitted for my ambition.

Duo Progress

2009-12-01T14:11:00.001Z

So I just finish the Zbrush version and I atempted to do a retopo version of the model but since I didn't change that much the overall shape of the body I use the target moph mesh I stored before I start sculpting. Even if you think of doing retopo having the basemesh store into the morph target is always useful.

so this is the reatime version of my character model.

This is my first attemps for hair but I'm not really happy with so I'm looking for some way to improve my styling

Showing off some art

2009-11-30T06:35:00.000Z

So this is the first character for my incoming game. I've been quite busy lately I didn't get a chance to make more progress with UDK but I did work on my game character Tachibana aka Duo. So there it is. I did a hair test in maya but it didn't work the way that I wanted so I may sculpt the whole hari cut into zbrush.

it's quite far from being finish but I will keep the update. So I need to sculpy his hair, retopo the whole mesh do a new uv map, then bake the hi mesh intothe low rez one and texture it.

I will rig it and animated it but it's another story. So there is a lot of work left but I'm pretty happy with my character right now.

screewing with the parking level

2009-11-18T08:08:00.000Z

While improving with material editor in UDK, I start appllying new effect to my level. I might change the overall look of it in the final version but still a better image than in my previous post.

lightmass with bsp

2009-11-17T09:19:00.000Z

I'm getting really great result with bsp and emissive static meshes. I'm quite happy with the result of my last tryout. The problem is my level won't use bsp or really a small amout so I need to get the same amout of detail with my custom models

lightmass testing and level testing

2009-11-17T05:30:00.001Z

So while I'm not working on my showreel I'm learning how to use UDK and it is not so easy to pick up. When I Say not so easy is that if you want to have a good level is not possible right away. I'm kind a learning the pipeline so your progress a pretty slow. but I'll try to get there and put my models into the game without any loss of quality

this my parking level for my future game. I keep the secret on the gameplay but it's something inspired by dynamite dekka. So this is the lightmap. I still try out stuff since there is no lightmass tutorial so it's a bit vague what you can achieve.

the level with the texture. I still working on the ambiance.

My carbonmade Portfolio

2009-11-17T04:39:00.001Z

Just wanted to add my Carbonmade portfolio. Some character level and Cars I've done. This is not new but I thought that it would be nice to have it on my blog too.

http://ribbons.carbonmade.com/

Playing with hair in Maya

2009-11-17T04:28:00.000Z

Lately I haven't got the time to work n my showreel but I'm still doing some progress on photo realistic human characters. So today this the end on my experiment on hair using maya. I went pretty good though. I won't say it wasn't a hell to get there but I4m pretty happy to move forward.

Outdoor-02

2009-11-11T23:57:00.003Z

Speed tree

So this is the second tutorial on how to create tree within the UDK pipeline.

After we create our tree in Speed tree program, we just need to save it and then launch the Speed Tree Compiler.

Once we launch it you’ll be facing a menu asking you to load some trees. Add the tree you’ve just created and click next.

Then you have to define your export settings. Be sure to pick an appropriate folder.

- Settings

I chose to create only 2 billboards for my tree but the default settings are just fine. For the sake of the tutorial I choose to create a billboard which will be replacing the tree in game when the camera is far away. Pretty neat feature but it doesn’t always work. OR maybe I haven’t found or to properly set it up.

You can create overhead billboard also if your character is flying over the trees. I leave you choose what suit best your needs.

Click next and then your see two your texture in two different folders. The fist one is the leaf and the second one the billboard. Don’t worry if nothing appears in the second one the billboard will be out put properly.

Be sure to change the prefix name if you have more than one tree in the folder.

Set the other parameters to fits your tree like the target size for instance. I did merge my billboard to the texture. But you can keep into another file.

Then my Tree is ready to compile. Press F7 or the start button compilation in your compilation menu.

The open up your UDK editor and import the all components to the appropriate package.

You need to create the material first and then load them into the speed tree editor

So first pick the bark texture right click create a new material

Open it in the material editor in left clicking twice on the newly created material. For those familiar to this process just skip to the tree editor.

Connect the material to the diffuse channel

Select the normal map tree bark and with the texture highlighted right click in the material editor and select new texture sample. I usually set the compression of my normal mapping to uncompressed. Then connect the normal mapping to the normal input. Valid your change and close the material editor.

Repeat the process for the leaf and the leaf mesh, if you have one. The leaf mesh can be the collision model you add into the speed tree and export along. For this example I used a Rock I found in the apricot game made with the blender game engine. The assets are free of use so if you are looking for a grass or a rock you might want to check this game.

www.blender.org

Arrange your shader like in the picture and set your blend mode to BLEND_MASKED.

Valid and close the material editor.

Open the speed tree editor and the right just add the texture corresponding. The branches go with the bark material, the leaf mesh with the corresponding texture. Assign the leaf card, the card and the billboard with the leaf material.

You should end up with something like that.

Then just drag and drop your tree onto the level.

If you want real-time shadows, just press F4 and under the speed tree panel look for lightning and uncheck use precomputed shadows.

Trouble with the pipeline

2009-11-09T06:57:00.000Z

I managed to get my tree in the game engine but it look nothing like in the editor. It's kind of sad for a tool shipped with the editor.

Anyway I was trying to get some experiment to work and so far I try dynamic shadows, which is more fitted for outdoor game and guess what. this was terrible. Setting the shadow with the tree wasn't the hard task I was trying to push the system a little bit more but nothing really evolve. I think the reason most people use baked shadow is because it's was more stable than dynamics's. Unfortunately for me and people willing to do an outdoor based game we need more effort to get a decent result.

this is my first attempt to get something close to the result I got in Cryengine2

So yeah this is quite far from the realistic environment I had in mind but I have a lot to learn.

Crysis scenery

2009-11-09T04:42:00.000Z

I was working on the cryengine for just a week but it was enough to taste the wonder of its tool. I try a lot of tool and I really think it's my favourite. Crtengine2 is one ( maybe the most) powerful engine with dynamic shadows and a lot of eye candy feature like dynamic sky scattering water caustics and such. Most of those shaders don't have to be set by and. Sadly the engine is just available for modders so even if you game is greate only the small crymod community will be able to play with it.

There is also few problem like the like of global illumination or lightbaking. But as far I know I never experimented so much ease in game developpement.

Theses are some experiments. I'll try to create such environment in the UDK.

I did pick the cryEngine at the beginning because I was trying to create a dense forest for my game. The first part is set in a forest and the main character is pursed by enemy and unarmed.

So the cry engine was the best choice for my game. I even gave up the idea of selling this game in order to get a powerful engine ( tired to create my own tool) but the UDK came so I'm trying to port everything I had so far and learn how to do realistic outdoor.

progess and a little bit of deception

2009-11-09T04:24:00.000Z

Even if I knew the UDK wasn't a terrain based editor I'm a bit disappointed with its terrain features. It's really buggy. One advice never edit the order of your texture layer or you'll find the end of it. The engine get really messy and even some layer you've erased while ago are painted instead of the one you just created.

But enough with whining, Epic games did a great job offering us the UDK.

So I'm posting some progress with the terrain editor.
I created my own textures and normal map. I use imagesynth which is a wonderful tool to tile your texture. So for the sake of my pseudo tutorial experiment, on creating realistic outdoor in UDK, I'll try to progress as fast I can.

it's barely nothing but just to show off my texture in game.

there is a lot of tutorial on how create a terrain so I was wondering some people wanted me to do a quick one.

I'm not really versed in tutorial on UDk atm bt I could give few trick to ease the process. I was also think of releasing few material in pack. I think except for cliff an rock ut3 material aren't really realistic, so maybe people need som good and tiled realistic textures.

Outdoor- 01

2009-11-07T21:58:00.001Z

Speedtree quick start.

I don’t know if people are facing problem using Speed tree but I’m going to start a sort of tutorial.

It’s going to be mainly outdoor oriented because there is sufficient indoor references. I’m in learning process so feel free to correct me if I’m doing anything wrong.

So the first part will be focused on creating speed tree and the basics features and functionality. I’m not going through all the details because there is a documentation accessible from the help menu.

So start of by creating a new document. You should see a blank screen appear.

To start building your tree just start by right clicking on the tree icon on your generation window on your bottom right

From there you can select what type of part you want to add to your tree. The menu is self explanatory.

For this mini tutorial we will select the “standard RT” trunk. You’ll see the trunk appears in the main window.

In the top of that window you can toggle the display features.

You can switch on and off the light, the shadow, the tree, leaves and other features. You can also change the display mode in the generation tree.

In the asset/ material menu in the top right click on the [+] sign to add a texture. Select a texture in your browser and load it

Click on the tree then, in your property menu roll down and select material/ branch, open it. Switch none to the texture you just created. (Takes time to load)

You can tweak the tree using the property tool but I’ll let you try those setting. Just want to say you can use a curve and the curve presets to change the setting over the height

So after you tweak a bit just add a branch node on top of the trunk in your generation window. I pick the branch “standard RT “for this tutorial. As you can see the texture material from the branch is derived from the trunk.

I just tweaked few feature like length ( self explanatory) , the frequency ,the start and start angle.

The start is a sort of scale to represent where the branching start. So if 1 is the top of the tree and 0 the bottom so 0.5 is the middle. I set my start to 0.3 for example.

The star angle is self explanatory and you can tweak the curve to create variations in your branching.

I decided to add another set of branch to create a better look. So on top of the first branching I add another one. I had to resize it to ten to make it look realistic.

So that’s my favourite part. We need to add leaves. So on top of last branching just add leaves. I pick wild but you could pick just standard.

Then just add a new material like for the trunk. Asset-> material-> [+] -> select texture . Then in the property panel look for the leaves option, open it and the select the texture in the type setting. Change it just like for the trunk and you should have something like that.

I change the size a bit using the curve to create a sort falloff. You can multiple set of leaves on top of branching.

So that’s it for a pseudo tutorial. I may post some other experiment small tutorial just to starters like terrain and such.

Next is to get those trees on the game engine.

Edit speed tree UDN is accessible now so maybe you can have a look to have a official paper explaining all the feature in it.

**credits for the textures. Dim arboretum on blender game engine forum and apricot game.

in-game model

2009-11-07T16:06:00.000Z

First in-game render with the simple diffuse and normal mapping. I'll try to add the subsurface scattering shader working.

Model showcase

2009-11-07T13:10:00.000Z

Some experiments with Zapplink and polypainting. Came out really nice.

I'll try to do a low rez model render using UDK, to show what my game character will look like.

model showcase

2009-11-07T12:02:00.000Z

This is a character I modeled while back on maya and Zbrush. The hi poly model is 10 milions poly but I bake into a 1500 poly model.

I may use it in my demo game using the UDK

UDK

2009-11-07T11:42:00.000Z

the UDK or unreal developpement kit is like a Xmas gift form epic.

I was struggling between finish a decent framework in xna or use a game engine when this news came out of nowhere.

Just one week after Unity becomming free, UDK is released and it's huge.

I wasn't really interested in the Unity because of its lack of render to texture function( no shadow, env map or post processing effects) and building tool from scratch is really time consuming. As game developper you prefer spending time on gameplay rather than tools.

I try moding with the cryengine ( since I own the game) and I was pretty impress but the fact you couldn't publish the game and share it to rest of world was a bit depressing.

So this middle ground solution was excatlly what I was wishing without believing it may actually come true.

So how much for that ? It's a bit tricky but it's free until you want to ship it.

but rather than explaining in my terms this is the link

http://www.udk.com/

be sure to check the licensing.

first post

2009-11-07T11:36:00.000Z

So this is my first post. This blog is going to be mainly game developpement oriented and I'll try to post my progress in my game my experiments in both Xna and UDK and also my 3d models.