Japanese Zen / Stone Garden
polycounter lvl 9
Hey guys.
So, this will be the piece that demonstrates everything I have, all I've learned. It's going to probably take me 2 months to finish this by myself, with all the hours at work I have and all the crap I want to put into it.
I've been doing really rough top down sketches to help myself understand what kind of composition / layout I should be going for, and trying to obtain a sense of scale of how things should be. I'm not a concept artist by any means ;p I'd like people to point out what could be a more interesting composition / add elements / take away things, whatever. Since this is a top down image, the sense of different elevations of the land is hard to distinguish, but I plan on having that come into this somewhere along the way.
I'm trying to include as many interesting elements as possible, and get the best layout I can before I start anything in 3d.
My plans are to take all the meshes I make into UDK, and create materials, lighting etc all in UDK. I want this to be real-time friendly, however, I'd rather have more poly's than not to get the detail across.
The idea of building modular for any environment seems like the best idea, but I know almost everything will have to be unique, save the trees, stones / rocks, and maybe a few other things.
Here's a gallery of my current online reference. I'm using many books as well. If anyone else knows of any other really good examples of this kind of imagery, let me know please!
http://img51.imageshack.us/gal.php?g=1127m.jpg
Help me out guys. I'm really devoting my life to this now.
Thanks
So, this will be the piece that demonstrates everything I have, all I've learned. It's going to probably take me 2 months to finish this by myself, with all the hours at work I have and all the crap I want to put into it.
I've been doing really rough top down sketches to help myself understand what kind of composition / layout I should be going for, and trying to obtain a sense of scale of how things should be. I'm not a concept artist by any means ;p I'd like people to point out what could be a more interesting composition / add elements / take away things, whatever. Since this is a top down image, the sense of different elevations of the land is hard to distinguish, but I plan on having that come into this somewhere along the way.
I'm trying to include as many interesting elements as possible, and get the best layout I can before I start anything in 3d.
My plans are to take all the meshes I make into UDK, and create materials, lighting etc all in UDK. I want this to be real-time friendly, however, I'd rather have more poly's than not to get the detail across.
The idea of building modular for any environment seems like the best idea, but I know almost everything will have to be unique, save the trees, stones / rocks, and maybe a few other things.
Here's a gallery of my current online reference. I'm using many books as well. If anyone else knows of any other really good examples of this kind of imagery, let me know please!
http://img51.imageshack.us/gal.php?g=1127m.jpg
Help me out guys. I'm really devoting my life to this now.
Thanks
Replies
Some personal suggestions/thoughts:
1) In my oppinion a project the purpose of which is artistic showcase will not depend much on how well the top-down plan is thought (if any at all).
The impression from such enviro is mostly affected by a different kind of composition. This is "view composition", e.g. what and how the shapes and colors are seen from actual
I second Matro, look through your reference which you have collected a great bit of and pic out things in them that can be used over and over again in your scene
I like to build a list and then work through it
keep goin cant wait to see the final piece
The biggest challenge for me is getting to know how things are laid out. I'm dying to just start making the assets, but I didn't want to get ahead of myself.
I was thinking just that, Matro...get my rocks, bushes, etc all going, so I have something tangible to work with. To be honest, the conception part kills me. I need 3D as quick as possible, I just work so much better with it.
So I guess I'll shift my gears into making the rocks first, as they are such prevalent pieces throughout the environment.
Will be posting updates as soon as I get a few meshes going (In between my stupid work hours!).
Lol...I know, no big deal.
I'm just moving along, trying to keep motivation up, and do at least one piece a day. I figured that since rocks, stones etc are so prevalent in this environment, I'll make a modular ground piece that I could maybe place under the flowing river / stream that runs under the bridge. It'll easily be duplicated etc, so, that's 1 down, tons more to go lol. Oh, this piece is about 3k tris.
I plan on producing all the pieces for the environment with more "beauty" in mind than anything else, but ofcourse I'd still like it to run well during play.
Oh, yeah, trust me man, I've been dabbling with mesh painting a lot lately, thanks to 3D motives' adv mesh paint tuts...I can't wait to be at the point where I can use it!
Even though they're going to be under a stream you might want to break up the pattern with some smaller props (grass, lilly pads, on the other rocks etc.)
Tank you.
Nah I totally agree with you Matt.
The way I'm planning to deal with this environment is to just have tons of modular pieces and sets that I can mix together and throw on top of one another. I'm going for max detail, brah.
More rocks...lol...
A rock wall...
Misc...
I have a feeling I'm gonna be pretty decent at making rocks by the time I'm done with this...lol...
My next pieces will be...MORE ROCKS!
Hey there man, thanks for the c&c!
I completely agree with you. The super-sized rock there is just to see how well the texture holds up, and who knows, maybe I'll have it jutting out of the ground in the forest somewhere etc =D
As for what you just said, totally true. All this stuff currently will probably be somewhere along the river / streams' side or actually underneath the water, hence the discolorations. But definitely going for what you said as well! My next pieces, the actual stones for the walking path, will reflect that.
The problem being, it takes years of research and probably a lifetime of study to create your own Zen Garden, being such a highly specialised subject with infinite amounts of nuances.
You get all that knowledge for free modelling something that already exists.
(Ryogen-in)
To help you google, Zen Garden in Japanese is 石庭 (Rock Garden)
龍源院 is Ryogen-in (famous example)
Thank you so much hawken...no, I won't pooh on your idea at all, because in all honesty, coming up with my own zen garden layout is something I knew would be a major hassle! I was gonna just go off of all my reference, but didn't have a solid understanding just yet of what the composition / layout of the garden should be.
So thank you =]
Some very nice rocks derived from photo reference! I could see this scene looks very nice if you keep this quality level up and find some clever ways of repeating these assets, for a very orderly garden like this it shouldn't be too hard. Do your rock assets currently tile? Doesn't look like they do from the outer shape of the mesh and texture. That should definitely be one of your core concerns right now. Once you have just a few very tileable meshes/ materials you can slap this sucker together in no time.
Also, are you planning on using other techniques for creating a tiled walkway or other more orderly rock formations? Going out of your way to hand craft everything for these types of surfaces (tiled walkways, tiled decorative rocks) definitely makes it easier to tile them and you can often get more bang out of your normals/ displacement maps without having to use so many tris.
Anyway, interesting thread and subject manner, curious about the techniques you're using and where this is going!
I definitely intend to get going from this point forward with only tillable, modular pieces. These pieces here were in part just me finding my flow with things. Now that I got it all down, you will start to see much more modularity in my models and textures.
If only I didn't have to work so many hours...could get this done so much quicker
As far as technique and pipeline goes for me, it's quite large as far as pipelines go. I feel sometimes I should try to narrow down to less tools / programs, for the sake of keeping things a little more orderly and error free, but whatever, tools are tools and if you know how to use them right it doesn't matter what you come out with as long as it looks good and plays well!
I'll explain my pipeline a bit later, because I'm on break at work
Maya, make rough base of my model, Export as obj, import to ZBrush, sculpt the shit out of it, export from ZBrush, retopologize in 3D Coat, export new retopologized mesh, UV it in Maya and Headus UV Layout, go to xNormal, load my LP and HP, bake a normal map, occlusion map and convexity map (to further enhance AO details), and then begin my texturing and importing process into UDK.
In between all this, I might use Crazybump to give some more detail to my normal maps, and I might also use some other 3rd party apps for misc things.
To me, it sounds like I'm going overboard in using all these different tools, but maybe that's normal? I don't really see too much talk of what people use to make things nowadays besides Maya, Max, and ZBrush. Like I said, it works for me, so I'm all good. What matters is end product anyways, right?
Stepping stones are done, unless anyone has anything to say about it!
And yes, these are individual, placeable stones, not a set.
I made 2 different ones, and I think that may get across the illusion well enough that there is variety.
And, changed the nasty rock texture I had before. I'm now going for a more bluish and greenish color palette for the garden. I want calm, easy earth tones. And added some wide stepping stones. I'll definitely be adding more to those Standing stones later on....moss, more dirt, maybe, we'll see, I'm not sure yet.
The modularity is very slowly taking shape. Still nothing breath taking yet!!
Ah man I'm not even thinking about lighting just yet, other than the fact it will be day time, or possibly early morning. I've always loved the blue / orange color scheme with mornings.
All this crap I'm posting ATM should not be taken at all as the finished scene, not even in the slightest bit. Right now, it's a "playground", a "placeholder" if ya will, to test how my models and textures can deal with the engine itself
So, the next pieces will probably be something like a wall perimeter, like in the image posted by hawken.
The scene will start to actually take shape once I've got my most essential building blocks in place.
I'll build some walls, a gate and / or fence to mark as the entrance to the garden, and then I will probably model and texture the actual sand pit to house the rocks, stones, etc.
Trust me, I'm gonna make it awesome. Slowly but surely.
I'm pretty speechless with the rocky ground (http://img41.imageshack.us/img41/6900/rocksr.jpg), well done. Does it tiles ? Did you followed a tutorial or something ?
Also, I'm not really digging the redish stone model (http://img30.imageshack.us/img30/4696/damnrocks.jpg). It looks likes the scale of the texture is wrong or something but maybe it's just me..
Keep it up
Cremuss,
Thank you for the kind words
Actually, those pieces don't tile yet. I could make them tile, but those pieces were actually gonna be used for something else I had in mind. However, I could very easily tfix them to be tiled if need be later on!
The way I did them was using PhotoSculpt, which is a brand new tool that I don't think many peeps know about! Google it, it's pretty impressive.
I spent two days trying to figure out how to import into UDK what PhotoSculpt exports out. It was hard to get it to look just like what the program spits out, but I found out how to get it all into UDK and it works amazingly well
The other method, which is exactly the same but 100 times harder, is to just take a image, bring it to your favorite modeling package, and model off from that (As in, taking the cracks in a wall and extruding edges from a poly-plane to give the illusion there is depth.
As for the red stone model, I totally didn't like it much either.
Here's a shot of the current model, reference, and silhouette. (In maya, right before bringing it to ZBrush)...
So, this is the first variation of about 4 or 5. This one was definitely the hardest of the bunch. The top and bottoms were a bit difficult for me to pull off, but I just needed a base for ZBrush anyways. The other pagodas will look like these...
http://img695.imageshack.us/img695/1737/pagoda4.jpg
http://img37.imageshack.us/img37/1709/pagoda5.jpg
http://img706.imageshack.us/img706/2784/pagoda6.jpg
http://img269.imageshack.us/img269/2585/pagoda7.jpg
http://img841.imageshack.us/img841/9717/pagoda8.jpg
So, now that I've got the most difficult one out of the way, I'll probably just re-use some of the elements for the others, because none of the other ones are even close to as complex as the one I just did (Thank god.)
Now, time for ZBrush. ^_^
90% done with the sculpting in ZBrush...
Gotta add a few more details, then on to texturing it. The flat planes near the top are going to be alphas, with a stone grating style decoration. Like in the reference.
Check out this lantern, nice and beefy, with plenty of nice flat areas begging for zculpting.
You lantern ain't bad though, don't get me wrong, just not sure if the time spent in zbrush is all it could have been. Keep it up!
No, I totally hear ya Crazyfingers. I always take critiques well, no worries =D
And as such, I will take every word you just said with me to the next variation of the Pagoda, because I do believe you are 100% correct. I feel the same way about this one! I goofed up a bit, that's for sure!
My excuse, is that this is my first "practice" pagoda, ;D
I will really try and go for the one like in the pic you have shown there. I won't copy it, but I'm going to take it's concept and apply it to the others that I have in mind, in the future.
With all the comments as of late, I suppose I'm going to go back to it, and put some more details in the areas that could use more (Like the one you mentioned crazyfingers). I'll push it further and see what I can do. I think I tried too much to stay exactly with the reference, so I'm gonna stay aware of that from now on!
Like crazyfingers said, details are difficult to judge at times make them too small and nobody can see them from a normal viewing distance and they wont show in the normal maps rez.
Also if you are making a full level rather than just a scene to be viewed from set angles, I would spend some time creating a block out and navigate it.
Japanese guardens have huge amount of foliage in them and that can be hard to create in games.
They also have an intresting design of 'hide and reveal', they try and reveal features in the garden so that you are standing exactly where they want you to view it from(that was the worst explanation ever).
And dont forget the gravel "pool" are suposed to be symbolic water.
Well its about time I shut it now
Good luck!
Haha I was looking forward to your post Kharn!
I actually just started learning all those hidden aspects of Japanese gardens from this book I'm reading for research concerning this project - just learned about the gravel being symbolic of water and it's ripples
As for the long intervals between work updates, that's because I am always working late hours to keep up with my living situation....if I only had more time to do the art I wanted to... *sigh*
Regardless of that fact, it is always in my mind and inevitably this will all get done.
When I was thInking about the project as a whole the other day, I was considering making it just a scene rather than a whole level. I'd like it to be walkable and viewable from all angles, and big enough to incorporate lots of different element, but just a lot more cased in a tight fitted. Kind of like the more box shaped encased ones in my reference pics. We'll see.
And this drives me nuts but, since I started modeling so many damn rocks, I realized how simple and effective a scene I could make with some more specially sculpted ones in a scene depicting Stonehendge!!
Stonehendge would be super easy to make IMO - that'll either be my next project or an intermediary one right now...just seems so fun.
Thanks for your tips Kharn, nice of you to share your knowledge on this subject to someone who doesn't know quite enough about these very very particular gardens!!
My current Pagoda...
To clarify something, if I take my diffuse into zbrush, I can paint the seams away kind-of...right? That's what I have in mind.
Let me know what you guys think of this project so far!
Mind you, I've been making this thing with no reference at all, using only my memories from Japan....Why, you ask?...
It's only the base, but to be honest, I CANNOT find ANY pictures that have a decent shot of a Zen Garden wall in them! Seriously, it seems impossible to find good reference to use for making something like this!
If anyone has any pics or knows where some are, PLEASE help me out! I'm dying to get better reference pics for this!...
Here it is ATM...
Curious why the pagoda is split up the middle, but the polys on the grate aren't. I think the lighting on those could look weird due to them fanning out from one corner.
As for ref for a wall this isn't a true ref but maybe it could help you visualize it, check out this thread
http://www.polycount.com/forum/showthread.php?t=80172
And yeah, I agree the polycount is higher than what seems necessary on my Pagoda, but truly, I wanted all the detail of the curves, so it's actually intended to be that way. The player can get really close to it so I want it as a hero piece of sorts.
@Hayden - That worked, thanks man
@Baddcog - Oops
Yeah, the grating isn't split down the middle because for some reason when I'd light them in UDK, they wouldn't look right....not sure why, but just having it how I have it atm doesn't produce any weird lighting problems so I just left it
And I fixed what you mentioned, Baddcog
I think it's ready to be thrown into zBrush.
I agree, and thanks for helping me out there, I really felt something was off, and I'll definitely be going back later to polish up all the textures
Also, how do you guys think would be the ideal way to go about taking this into zBrush? Should I import each piece as a subtool, then merge down the pieces later or is there a better way?
The pagoda will look great up close. If the lighting looks better that way stick with it. You should make a lod model for that too. You could easily shave it down to 2000 tris I think and at a distance the player wouldn't notice.
Thanks man!
You're definitely right about the LOD - could only help. I'll get to that when I'm in polishing mode too.
Working on the tile wall right now, still WIP, no color yet!...
is there any reason why you have such different variation in your gloss map?
id try to add more varation on the wall part in your gloss map. differentiate the wall gloss between the stone. id make the wall part very dark gloss and make the stone a higher gloss with maybe patches of high specular and high gloss for a granite type look. might look good
The wall stones in the bottom wont tile well because the pattern is creating a blank space at both ends.
Also texel ratio on those stones seem uneven in your close up shots compared to cement (stone are blurry).
Can we see your references for your roof here?
The lantern pedestal looks very thin and weak between the legs.
Great observations - I can easily see what you mean by the pedestal looking weak, as in it could crack any second holding up all that weight.
Here is the reference I used...
and
As far as the Tile Wall goes, I was planning on adding a wooden or metal strip of some sort that runs along the whole edge of the wall, covering most of the part that you see as a blank space as well, mainly because I'm so lazy...but at least it will tile well afterwards! Do you think what I'm thinking will work if I do it right?
Like this...
I was trying my damned hardest to get it to read like metal as well as I could, but I must not have the right idea of how a Gloss map and Specular work together to produce such a material...I need to read up more about all that. If you know of any good resources, toss them my way!
Right now, my gloss looks like that because I wanted to see variation in the hotspots where the light was hitting it, but maybe it just isn't working...should I tone the saturation down on my Diffuse as well?
Updates...
Pagoda - Recreated my diffuse, moss comes from the bottom and gradients upwards instead of being splattered randomly everywhere. Lichen rests on the top portion.
Wall - Now tiling better, and materials read more how they should (Metal, plaster etc). There is still a very very minor seam remaining where the plaster wall ends and begins, but you have to look quite hard to find it.