Lack of Environment Artists
polycounter lvl 20
It came up in Strangefates post http://boards.polycount.net/showflat.php...;gonew=1#UNREAD
in pimp/preview section that there are very few that are interested in environment work in the ganes industry.
Env work has always been regarded more as the shit work or the work that people use as a means to traverse to a position they consider as higher up on the food chain, which is to be a character guy.
Characters are ego stroke work, but they generally occupy 20% of the screen real estate, the other 80 percent is the world in which they run around.
Despite that, artists generally would rather make characters.
It's my opinion that theres a good few reasons for this.
As mentioned, characters are an ego stroke, its more widely accepted that more time should be taken on a character than on a lampost for instance.
We don't see close up's of a lampost on the cover of magazines or gamebox's so theres the ego / pride coming into play there.
It's not something I think artists should be ashamed of, we all want to get acknowledgement of our ability, its a natural part of the drive to succeed.
The problem with this perception though is that its more widespread than the art team and as a result companies often don't have the foresight to proportion the necessary time for environment work that they do with characters which makes it tougher for people to excell in the environment area.
Or, many companies have people that do a bit of everything so artists rush through the environment work ( the shit work ) to get to the character work ( the fun stuff!).
Sadly, environment guys are the unsung heros of most games, in the previous generation of technology it was the case that often the map texture guys were the backbone of the games art direction in that the responsibility for tying so much of the game together rested on their shoulders but few of us knew who the texture map guys were.
Compare the awareness we have of who the skin artists were on the characters and you will see my point.
Another problem has always been , and still is, that usually the characters get a higher portion of the memory resource. If the game is not able to fit on the disc or in memory, the first thing to lose resolution or polycount are the environments, then often the effects, the particles and so on before the characters are sacrificed.
So with this lower spec coming in to play, its simply another reason why many use environment work as a stepping stone to get to a discipline that has a higher spec so that people have a greater chance to excell.
Collectively, these are the reasons many games environments end up being quite weak or have a mixture of quality levels that are not as consistent.
Anyway, the industry is changing, as it always is. Next gen offers( demands perhaps?) the chance to articulate every aspect of the game to a great level of detail.
It's thanks to this that environment work is becoming a lot more demanding and as a result the old attitudes of putting the most skilled artists in the area of character art is starting to create a wider division in the quality level of games environments.
I'd say this is the best possible time to start getting more heavily into environment creation, its actually an area thats more challenging, diverse and rewarding that others now that hipolygon creation is the driving force for its quality level.
Being an environment artists gives you the chance to recreate the world in your own style and allows you to take inspiration from any geographical location in the world and any time period of artistic movement.
It's true that you can argue the point that so does character creation but taking characters to a higher level of detail is more subject to the law of diminishing returns.
I can respect that someone has taken the time to model dimples in the skin of a face with zbrush or mudbox, I'm impressed by the closeup of it in a render or on a magazine cover, but I dont really appreciate that at all in the game because its so tiny.
With environment creation, theres the chance to walk right up to the increased level of detail and appreciate it and that is what is so rewarding.
Next gen likewise demands that far more subject matter is articulated in 3d by the environment team because less faked 3d on flat walls will hold up with modern normal map driven higher polygon worlds.
Environment works is also much more relaxing than discplines like character or vehicle creation now I think because to finish a single character or vehicle the time spent modelling it out can be anywhere between 2weeks to 2months by the time you are finished the low poly, unwrap and the hell of getting it all processed correctly.
The longer a single piece takes the more stressful it becomes, most of us have a barrier for self perceived efficiency that disappears once we hit the week and a half barrier. Anything beyond the length of time ends up feeling like we are wasting time, going to slow, not fast enough or worry sets in that management wants you to finish quicker.
While some environment pieces, like the hero assets can take the long, a lot of time take a week, or less and the balance between piecemeal snack like construction periods and then long marathon endurance test assets helps to promote a more relaxed and varied workflow.
For me personally, I thrive in the research of other cultures, time periods and artstyles that is necessary in order to convincingly articulate a new or wellknown world.
I joke to myself that others create characters, or guns, or blades, or vehicles... but I get to play god and make a world.
Anyway, if you have opinions, it'd be interesting to hear them.
in pimp/preview section that there are very few that are interested in environment work in the ganes industry.
Env work has always been regarded more as the shit work or the work that people use as a means to traverse to a position they consider as higher up on the food chain, which is to be a character guy.
Characters are ego stroke work, but they generally occupy 20% of the screen real estate, the other 80 percent is the world in which they run around.
Despite that, artists generally would rather make characters.
It's my opinion that theres a good few reasons for this.
As mentioned, characters are an ego stroke, its more widely accepted that more time should be taken on a character than on a lampost for instance.
We don't see close up's of a lampost on the cover of magazines or gamebox's so theres the ego / pride coming into play there.
It's not something I think artists should be ashamed of, we all want to get acknowledgement of our ability, its a natural part of the drive to succeed.
The problem with this perception though is that its more widespread than the art team and as a result companies often don't have the foresight to proportion the necessary time for environment work that they do with characters which makes it tougher for people to excell in the environment area.
Or, many companies have people that do a bit of everything so artists rush through the environment work ( the shit work ) to get to the character work ( the fun stuff!).
Sadly, environment guys are the unsung heros of most games, in the previous generation of technology it was the case that often the map texture guys were the backbone of the games art direction in that the responsibility for tying so much of the game together rested on their shoulders but few of us knew who the texture map guys were.
Compare the awareness we have of who the skin artists were on the characters and you will see my point.
Another problem has always been , and still is, that usually the characters get a higher portion of the memory resource. If the game is not able to fit on the disc or in memory, the first thing to lose resolution or polycount are the environments, then often the effects, the particles and so on before the characters are sacrificed.
So with this lower spec coming in to play, its simply another reason why many use environment work as a stepping stone to get to a discipline that has a higher spec so that people have a greater chance to excell.
Collectively, these are the reasons many games environments end up being quite weak or have a mixture of quality levels that are not as consistent.
Anyway, the industry is changing, as it always is. Next gen offers( demands perhaps?) the chance to articulate every aspect of the game to a great level of detail.
It's thanks to this that environment work is becoming a lot more demanding and as a result the old attitudes of putting the most skilled artists in the area of character art is starting to create a wider division in the quality level of games environments.
I'd say this is the best possible time to start getting more heavily into environment creation, its actually an area thats more challenging, diverse and rewarding that others now that hipolygon creation is the driving force for its quality level.
Being an environment artists gives you the chance to recreate the world in your own style and allows you to take inspiration from any geographical location in the world and any time period of artistic movement.
It's true that you can argue the point that so does character creation but taking characters to a higher level of detail is more subject to the law of diminishing returns.
I can respect that someone has taken the time to model dimples in the skin of a face with zbrush or mudbox, I'm impressed by the closeup of it in a render or on a magazine cover, but I dont really appreciate that at all in the game because its so tiny.
With environment creation, theres the chance to walk right up to the increased level of detail and appreciate it and that is what is so rewarding.
Next gen likewise demands that far more subject matter is articulated in 3d by the environment team because less faked 3d on flat walls will hold up with modern normal map driven higher polygon worlds.
Environment works is also much more relaxing than discplines like character or vehicle creation now I think because to finish a single character or vehicle the time spent modelling it out can be anywhere between 2weeks to 2months by the time you are finished the low poly, unwrap and the hell of getting it all processed correctly.
The longer a single piece takes the more stressful it becomes, most of us have a barrier for self perceived efficiency that disappears once we hit the week and a half barrier. Anything beyond the length of time ends up feeling like we are wasting time, going to slow, not fast enough or worry sets in that management wants you to finish quicker.
While some environment pieces, like the hero assets can take the long, a lot of time take a week, or less and the balance between piecemeal snack like construction periods and then long marathon endurance test assets helps to promote a more relaxed and varied workflow.
For me personally, I thrive in the research of other cultures, time periods and artstyles that is necessary in order to convincingly articulate a new or wellknown world.
I joke to myself that others create characters, or guns, or blades, or vehicles... but I get to play god and make a world.
Anyway, if you have opinions, it'd be interesting to hear them.
Replies
A year or so ago I started to make some environment models for a mod . I modeled some tree's and buildings and rocks . I then showed them off on the mod team forum . The same day the character artist showed off his character . I got 2 responses " hey those look pretty good " and of course the character guy got like 20 ! I can see the discouragement someone might feel doing environments .
Also from the perspective of a none-pro who is trying to break into the game industry Environment art is a bit confusing . It's like do I need to make 10 buildings and some rocks and ground textures and some sky domes ..etc etc in order to impress an Art Director and get a job ? Do I need to know level design well ? Another thing is that there are VERY VERY few tutorials and instruction for doing environment art so it's like stumbeling around in the dark , but for characters there is a vast resource and entire forums dedicated to it . Also a lot of resources for level designers just from the sheer amount of people who make mod levels for games like Never Winter Nights ,Unreal..etc but just almost nothing for environmental art.
Maybe you can explain what impresses you and even EPIC in a portfolio of a none-pro try to break in, because as you said there is a shortage ( meaning the industry needs fresh blood ) and obviously nice published game levels and environments from an artist with experience would impress you but noobs would not have that unless they did some stuff for a published mod maybe. How much does seeing good environment models from none experienced mod makers impress you and say EPIC for that matter ?
Just a quick look through a few game job sites and I see that most companies want Environmental artists with at least 3+ years of experience . I know what people say that it's just a precautionary but like you said there is a shortage but it does not seem to be effecting hiring practices ?
Hmm I dunno in my opinion the road just seems clearer when it comes to breaking in as a character artist . I mean it''s easy to say what a really cool character looks like but very hard to say what a really cool building or tree looks like
When developing atmosphere and really communicating a certain style environmental takes on a lot of challenges so i really respect what a lot of studios/artists that come up strong with in terms of strong environmental aesthetic.
Aesthetic, mood, atmosphere is all very difficult to communicate and develop that sensitivity to. And i think a lot of artists think it's about making props and filling a world with as much junk as humanly possible while missing the big picture. You are creating a world and there's so much opportunity to make it impactful.
I think a lot of it stems from unprofessional and immature attitude frankly. Perhaps it's because the industry is still quite young and hasn't developed the respect and image as the film industry has. I think that kind of stuff really degrades the quality of work and give the industry bad rep. Environmental art can do so much more and some artists are really taking it to that special level, while to others it's an afterthought.
Sorry in advance for derailng.
I think i'm rambling on now, i need some sleep ..
Anyways, yea, just wanted to say that it's great to hear this and it really helps to encourage me some more to continue working on my environmental skills.
[ QUOTE ]
'I joke to myself that others create characters, or guns, or blades, or vehicles... but I get to play god and make a world.'
[/ QUOTE ]
You have a lot of power as an environmental artist as far as defining the look and feel of a game goes. Once could say the visual quality, look and feel of the whole game is in your hands ...and in the hands of the other enviro artists of course.
Depending on the team structure, you'll be either doing single props (Buildings, decorations etc) or be able to look at a level and go in and model/texture everything yourself at your hearts content.
It is important to have an eye for the big picture or the stuff wont come together well
Whether you'll be doing crates or awesome hipoly center pieces depends on your skills i guess. Personally i like to go all across the board. If i spend 2 weeks doing something fancy and hipoly, i often want something quick and simple after that.
Also, modeling a tree or some gras sounds boring at first glance, but so does modeling fingernails or nasal hair. It's the big picture, knowing how cool the gras or the tree will make your scenes look that makes even those little things exciting.
There's as many boring and exciting things in environment work as there are in character modeling, it's the big final picture in your head that makes that nasal hair so awesome to model (not that i ever saw anyone modeling nasal hair).
I agree that the importance of environments is sadly underestimated, when 80% of what you see in a game is environment and the other 20% whatever, it should be a no brainer, you'd think anyway.
I hope that games like GoW will help us a bit, as most of the reviews compliment the environments over everything else, or at least, they seem to take up most of the review space
i for my part can say: i love doing character work.. the design the modeling and texturing.. it´s kind of funny but the best thing is to "play" with this new toy once you finished it.. its like creating you owne actionheros
but for my current project i´m doing a lot env-work and i love it too.. but only bc my chars can use it to present them self (they are so narcisstic..) but on the other hand.. for regular industrie work i think i prefere the more relaxed env-work
my tip: make the best char-models in your freetime (if you have some
And this is why i love making weapons, the one thing that is ALWAYS on screen and close up. Very rewarding making something badass that will be seen 100% of the time
[/ QUOTE ]
I've not seen any guns in Tony Hawks. Or Bully. Or Need For Speed. Or Mario. Or Monkey Ball. All of those do have environments though.
Ror, you've made a very important point. We sort of spoke about this a while ago, and some of it certainly is ego based. The animimation people and vfx guys probably fell the same way to an extent. The character artists can point out "I made HIM!", while the others get to say "I made that wall/smoke/crouch animation."
It's worth remembering that when this new generation of consoles arrived on the scene (2 years ago in studios, not in November 2006) companies started hiring. Most of those character art positions have been filled, but the lack of environment artists mean that I can list 8 or 9 companies that I know FIRST HAND who are activley trying to find map artists.
How many games ship with editors? 3dsmax costs £3000 or whatever, and Unreal Tournament costs £30 and you get the editor for free. You can build and texture a level and run around in it with learning who to rig and export.
Yes, most places I know build their levesl n a 3d package, but for the first forays you can do it cheaply, easily, with 100s of tutorials.
Theres no shame in going back to Quake or Quake 2 to earn your initial stripes.
The main reason for this is dead simple,I find enviroments easyier to do.This doesn't mean they are easyier to do but I personally find them easyier and I think there afew reasons for this.
1.Everywhere I go im in a new enviorment,for example I look outt he window of the house,I see the house opposite its garage clean streets etc,whereas I know if I go round the corner id be in a different enviroement,bigger road,perhaps dirtier pathments,then for something totally different I could visit the old city of Oxford and study/take pics of the old achitecture taht surrounds the city.
I find it harder to do this with charcters,despite being surrounded by different people 24/7,but to me most people look the same,for example,in the summer guys will wear Jeans and T-Shirt.....but so does everybody else,different colours logos,messages etc but still tehy all look very similar.
2.Its much more likely as an enviroment artist id be doing different things,I don't think anyone has touched on this point yet,but I mean as a charcter artist say working on a skater game,all your doing is creating "just another skater". we havent seen that before have we?
Whereas for that same game you could be working on assets for a garden,for a buissness park,for a factory a wide range of oppotunities to do different things.
3)And proberly the simpilist reason of all,theres a need for enviorment artists,ive read this countless times on Polycount and from other sources,I figure therefore thats far and away my best chance.
I do wish however as McIlroy has said there was more info about enviormental art work for games avalible.But having had a small brainwave while typing that line,a question has arose.
Do we not see many specific envioroment tutorials online becuase of the broad range of how game engines handle them and therefore it all comes down to the old quote "It depends"?
John
I find myself more and more thinking about environment designs these days than character designs, and when I can look at a whole in-game terrain and think "I made that", it's pretty cool.
for a long time as stated before, the split was mostly some guys doing the textures and another one doing the brushwork, layout... The visually best maps were those were the same person would be a very good texture artist, too or had someone else doing work specifically for him.
With the move away from lots of "brushwork" to the static-mesh stuff, mappers basically need to be modellers. Before it was sufficient to work like lego with boxes, you could do some CSG stuff and so on, which is a bit more "natural" modelling of structures, compared to thinking on triangle/vertex level.
I guess that just too little of the mappers made that transition yet. When you mention unreal-ed, those mappers would mostly work with assets the engine gives them (static meshes..) and "just" do layouting work mostly. Not building their own detail meshes and such.
And then of course is the mentioned issue of the ego, character art getting more coverage... eventhough it's completely stupid as enviroment is what gives major atmosphere to a game imo. Thinking about the lucas arts adventures (indy jones, monkey island..).
Maybe we need easier tools that make modelling of world geometry more natural as the lego-building in quake1 was... (when working with CAD software I really enjoyed the perfect booleans, the adding of features... however the drawback was that the system would need more "planing", ie changing a concept would have been harder, but if you had a plan before it was pretty cool and fast to work with it)
i started with leveldesigning fantasy maps for a hl1 mod and it was really cool to recreate your own vision in 3d. back in those days you could pull off whole maps as one person only, including textures, lighting and models. these days it gets quite complex and making a complete map by yourself takes up a lot more time. nevertheless its still one of the most interesting jobs in gamedev - thats my opinion
heres a list of cool environment artist portfolios. if you got links aswell then let me know:
(copy/pasted from my link list)
www.chrisholden.net
www.polycreator.com
www.stellan.be
www.pernroth.nu
www.jsalvador.net
www.diduct.com
www.warbeast.de
www.thomashess.net
www.mangledpoly.com
www.philipk.net
www.littlelostpoly.co.uk
www.waveswrip.com
www.nathanielbell.net
onlinewarriors.beyondunreal.com
www.tylerwanlass.com
www.mikeyspano.com
www.demondrogers.com
www.generalvivi.lawgiver.us
www.michalmonart.com/wilinski
www.marshmonkey.com/David_Marsh
www.jsmars.com
www.squirrelyjones.com
www.tgz3d.com
hem.bredband.net/b224223/portfolio
Rick: Being an environment artist and being a mapper are two very different things, man. Env artists as I see them discussed here have little to do with making levels, they make the art assets that are placed within levels. And so need max or whatever.
[/ QUOTE ]
I guess that all depends on where you work and how that place works then. In the 3(4?) companies that I have worked at, my description is more valid that yours.
not to forget that on every project i've worked on, the env guys were working longer hours and had to deal a lot more with engine-issues. or even had to use those nasty inhouse-tools. brrrrrrr.
that being said, character work can be annoying as well if you have a chain of managers and producers letting you make endless revisions. i'm working on a lead character these days, i get the full treatment!
I also think that a character artist if needed, can do perfectly the work of an environment artist. An artist must be a 4x4, at least is how this works in spain. The better employers doing characters end doing the main characters, the others with less level or artistic quality do other simplier tasks as props or scenery. A character is the "image" of a game.
Tomb Raider, for example, is very well known due to Lara. Character got all the importance although good environments are also something very very important. Just look for example at prince of persia, the scenery is what calls more attention. I enjoyed a lot the game, and i disliked the character work.
Good points there.
That said, I've always had an interest in environment work, and while it may not be the dream job character art would be, for me, I'd be more than happy to get hired doing environment work instead.
I guess I`m skipping through the `I wanna be a game-artist`, coming from `I wanna be an artist` perspective. There`s an axe for grinding, what what.
Apologies to any `emo crates` that may be reading.
and hopefully we will see more of those places... imo you can put that sort of life into enviro art, too. Just compare sterile modern enviro (glassy, steel, cleaned up) with rural old/whatever places, where every building is "off-shape" and tells a story too. Likely it is the time constraint that simply says, we need xyz buildings, trash cans... but wouldnt that also apply to the non-lead characters ?
Big up to ror for that long ass foreword very useful
I'll repost my old speech from StrangeFate thread just to get a starting point:
[ QUOTE ]
I'm relly interested about this discussion, I mean do you guys think there is an oversaturation of character artists out there?
I thought it was animation since all the assmonkeys I know at school are wanting to become animators.
I think the lack of environnement artists comes from the fact that there aren't a lot of available ressources. I mean the only tutorials I can think of at the mo' are cholden's ones (non Unreal related, saw there are a lot of books and dvds on the subject already).
It'd be cool if someone makes some sort of introduction to environnement making for character modelers especially by mentioning all the usual stuff beginners aren't aware of (little problems like scaling, uv mapping large surfaces, tiling... I think there is way more but these are the only things I can think of at the moment). And it'be cooler if it was generic for major softwares (max maya modo...).
Sorry if this post looks stupid but you env guys got me interested
Ciao!
[/ QUOTE ]
So as a character artist wannabe (and very interested in env art) I found it very difficult to find infos on the matter. I'd like to learn to make a full working environnement in my chosen 3d app without having to deal with any engine issue as a starting point, just to take more care of the aesthetical aspect of my environnement but still keeping it under normal games limitations. Then later on, learn the UT2k4 or whatever popular built in editor.
So here are my problem that are keeping me from trying env art;
Let's say I want to make a little rts map in max:
_The first issue that comes into my mind is the terrain
In conclusion all the aspects of Mapping without using an editor may be problematic for someone whose been using on working on single meshes at a time.
_Then comes the actual env art assets;
Modeling is still not a problem and since the buildings are supposed to be small they'll be uv mapped into a couple of texture sheets, but what if those building were supposed to be for an FPS? Should I tile my walls or whatever texture so I don't have to use a big ass texture again (since the Pov is pretty low)?
_And at the end how should I go with placing my assets? are there rules (especially since I've most engines calculate polygon interpenetration as if they were booleans)?
As an end note I must say that I personaly got introduced to characters making thanks to poop's vid tuts (It's his fault I SWEAR!!) then I saw that there were so many ressources for character artist that I kept this direction.
It'd be really cool if any of the confirmed pros here talks a little bit about the technical process involved in making a full environnement and why not a full walkthrough of different typologies of envs
Sorry for the big rant and for my lousy english
Cheers! and keep discussing
I just looked. I just did some searching, and the dearth is quite staggering. Most of the useful information is actaully posted in the P&P forum of this messageboard.
It's not something I have the time to do, especially at this time of year, or for the next 8 months (I'll be hitting deadlines), but I've help out if others wanted to work on it.
I think I like it more because of the scale of the objects I'd usually make, and all the extra layers/effects you have at your disposal to make something look really good.
Now that I'm coming off the character team, they've had me do all kinds of statues and things with skulls. I guess most env artists aren't prepared for that kind of thing, usually.
Modeling a character is easy, modeling an awesome cool looking and believable character isn't, same with archietcture, vehicles, whatever.
Being able to model toons and cars after a modeling sheet with nice front and side poses is a rather isn't gonna help you anywhere else. There's no modeling sheets for enviros.
Then again, a good artist is a good artist, not matter what he's working on... to some extent.
I think one of the reasons there's less material for enviro art is that there isn't so much to know. You don't have to worry about joint deformations etc. Also, enviros don't try to kill you, you can walk around or fly around with cheats in any game and look at how they did things.
Some ideas and tricks come with experience, but that's another story.
[ QUOTE ]
_The first issue that comes into my mind is the terrain, I think I can deal with the modeling aspect of it, but how should I uv map it? how should I texture it? I don't think the best way is to put a big ass 10k*10k pixel map on it, And if I had tiling, how should I deal with the different surface blending (you know grass-->dirt)?
In conclusion all the aspects of Mapping without using an editor may be problematic for someone whose been using on working on single meshes at a time.
[/ QUOTE ]
First, it doesn't matter at all how you do thing, it doesn't matter MUCH how the polygon count is. Like some like to say 'it depends' which means that it doesn't really matter.
If we're talking portfolio pieces, it just has to look good and ingamish. We just want to see that you can model nice enviros and have enough imagination and skills to do a fishing market or plaza all yourself.
Whether it runs fine in any engine, i don't care,if we like what we see we can talk about polygons later. It IS that hard to find good enviro artists. If someone's good, there's always enough time to learn specifics in house.
Terrain, it can work like any other surface using tiling textures. You can either use an alpha channel or vertex blending (which we use) to blend several materials together.
The only rule there is i think, is to work clean. Don't use more polygons and texture space than you really need to convey the shapes you want, and if an expensive shape can be redesigned into something else that looks as good but is cheaper, it's always worth a thought.
As in, if you have photos of architecture with tons of spheres everywhere... try to come up with something else for those spheres
[ QUOTE ]
_Then comes the actual env art assets;
Modeling is still not a problem and since the buildings are supposed to be small they'll be uv mapped into a couple of texture sheets, but what if those building were supposed to be for an FPS? Should I tile my walls or whatever texture so I don't have to use a big ass texture again (since the Pov is pretty low)?
[/ QUOTE ]
Tiling is your friend. The buildings in my post use only tiling textures (not for the window glass or electric boxes).
I use vertex blending to break up the tiling and add the usual dirt/weatheration, blending 2 textures/normalmaps together.
The normalmapped architectural trim details on them use a second UV channel, and the normalmap and occlusion maps there get overlayed on top of the tiling materials.
Still, you could skin a building i guess, if the skin work is awesome, i'd asume that you can make tiling textures in the same quality.
[ QUOTE ]
_And at the end how should I go with placing my assets? are there rules (especially since I've most engines calculate polygon interpenetration as if they were booleans)?
[/ QUOTE ]
hmm i think that's only when you work with BSP or CSG ?
THere's no rules, just bring them in and have fun.
Each engine works different tho, some will let you use several UV channels, some wont, some will expect this or that.
But again, for portfolio purposses you can model everything in your favorite 3d app. I think it's wrong nowadays to expect from anyone to have their work in a game engine.
It's not worth the time learning all the tips and tricks for a complex engine when only a small percentage of companies use it.
You'll learn the engine and it's features, limits and resource restrictions when they hire you.
Now that I'm coming off the character team, they've had me do all kinds of statues and things with skulls. I guess most env artists aren't prepared for that kind of thing, usually.
[/ QUOTE ]
True, there's a lot of cool stuff to do in environments, it's your skills alone that will decide whether you'll be doing the cool stuff or the crates.
As an enviro artist, you can decide what cool things you want to model. You're doing a park ? well, it would look a lot better with an elaborate old statue with beautiful cloth or perhaps, it's a statue with a horse, or general, angel or it's a fountain with mermaids, poseidon or... you get the point.
You may not be doing the ingame characters as an enviro artist, but you can often do anything you ever want otherwise. Only limit is your skills, and the time available.
These are common problems with the games industry and I can clearly recognise the need for an increased recognition of quality environment art, to help promote the role and avoid the need to hire half a team of newbs to fill up the environment positions. There really shouldnt be this perception that somehow its less to work on environments, I have allot of respect for my friends on the environment team. It amazes me at some of the work they are doing but what is scary is that I wouldnt have a clue how to acheive half of what they are getting in game.
To contrast Fate's approach to buildings and further the point that method of construction is just the road that leads to the destination, the destination being a cool environment, I'll explain how I made buildings in Gears.
A building is just a cube or a rectangle, which means you only need to model onc side of it, much like you only need to model or paint half a face on a character. Much like a face though, it helps to unwrap a little more than half so as to avoid the middle mirror seam being so obvious.
With so much of the environment being 3d now, I try to extend the purpose of tiling 2d textures into 3d and make the 3d forms modular. I split a building up into its modular bits, building tower top, clock face / windows, rows of windows, side pillars, sectional trims, floor level windows and door
Once the modular chunks are decided upon I model everything up in low polygon form, but I do it in a lazy fashion in that I leave lots of unneeded edge loops because I'm sketching out my plan for making sections modular and laying down my best guess at the main forms. I know that I will use what Im modeling out right now to generate the hi poly and the low poly later on.
I call this type of modeling medium poly because it is essentially low poly modeling without the tidy up and it can go either way, I suppose you can call it lazy modeling too.
At this stage there are 3 key things for me.
One is the identification of which main forms can be broken into modular chunks, like the pillars that separate the windows horizontally from each other, they are all the same so I only need to model them once and instance them later at the hi poly stage and process them all down to one small area on the texture later at the lo poly stage.
Two is the restraint of my layout to the grid. The grid is god for an environment artist if you want to be a good one and if you want the level designers to be able to use your work and of course like you enough to take suggestions and trust you enough to think yor will understand further requests they might make of you later. I work on a 16 grid, I try not to go below 8 though I dont leave the grid on for more than the planning / modular stage of creation because then it interferes with the fine details. To be clear, I ensure the modular blocks can be manipulated and combined with others when conformed the a grid based pivot point because thats what the LDs will use in the editor where they cannot changed the pivot point like we can in max.
Three is simply thinking ahead enough to understand how the asset parts will be used, reused and where I can put the seams in such a way as to avoid them being obvious. I also take the time to forward think which elements will need a pass in zbrush so that I construct the part in a way that support equal subdivision of the mesh.
Once all the ground work has been planned out I begin splitting the main building into as many different separate objects as I can for more manageable polygon sizes per object.
Splitting up parts is a major part of working in hi poly and a considerable time saver when working on environment parts.
So once I have everything split out in medium poly, conformed to a grid, in preplanned modular pieces that I know can be instanced I generally split off a duplicate of the parts and finish out my low poly so I know what kind of polycount it will be and so I have guide, or a framework to create my hi poly within. This is crucial, often it is easy to forget that the hi poly is just another way of drawing out the design on to the low polygon like painting used to be.
So once this is all done, I know it all works, it all fits together, the polycount is defined and I am confident and assured enough to begin the fun stuff, the design 
Usually if working on a large building, I will then split it up into separate files defined by what pieces will go on which textures, typically a building will eat up half a dozen textures, but because Ive planned it out to be modular, those textures and model chunks can be reused and recombined in dozens of ways which also ensures continuity of style and art direction which leads to a cohesive and convincing world set.
At this point, to make a whole building of about 4 floors, plan it out, low poly, medium poly, unwrap etc, Ive probably eaten up a full working week but its all solid, approved by my leads and I know how much texture res will go to each area so I have a yardstick for how much detail is worth putting where.
Creating the hi-poly for all of this will take a further 3weeks or more.
The hi poly itself is made up of multiple chunks so because of this I tend to flesh out all the chunks individually at the medium detail level, getting all the curves and chamfers set up so they flow and connect to each other. Once those are done I take on the design elements such as bass relief, patterns, engravings, bolts. The best way I have found to do this is to simply made a spline shape of one leaf, model it out with the shell modifier, then chamfer the edges, put turbosmooth and then deform multiple duplicates of this along a spline path that mirrors the main forms using the path deform modifier so that I can inline a leaf deco base relief design around the window frame trims for instance.
I then continue over everything like this, find a main form, model out a repeating pattern, deform it to the path, repeat for next form.
Once this is done I take all the concrete elements I need to into zbrush and use a 24bit alpha texture ( grayscale PSD) as a stencil for my concrete detail as an overlay in zbrush and then go back over it all by hand choosing where to add dents, scraps, cracks and so on. Once I have done this to all the concrete elements I export the objs from zbrush into a separate folder and run a bath process through polycruncher plugin in max, I set it to about 60 percent.
Once this has been completed I bring those chunks individually back into the scene in max, apply materials to everything, optimize more with poylcruncher the pieces that can handle it and once it all looks good with materials assigned to it I begin the often nightmarish task of getting it processed.
The advantage of applying materials that are representative of the final design in max is that then you can set up a lighting arrangement or 2 and process a diffuse map, specular map, lighting map, height map and normal map all at once and it gives the texture artists a great headstart.
I tend to rely on a hand made lighting scene for generating as lightmap that indicates where the shadows might fall based on top down lights. I then do one using a skylight only which indicates the depth of each area relative to the normals of the low poly which is very helpful also.
Once this is done, I make sure all the UVs are correct, then unwrap a 2nd set of UVWs on channel 2 where each piece has a unique area so that the engine has a lightmap UVW channel for generating shadows and lightmaps.
Its a long tiring process, so making it modular ensures that the assets can be used many times. If any of you saw the UT E3 demo of OnxyCoast this year, it might interest you to know that every single building or piece of concrete in that level was on a mere 2 textures. Making it modular works, and it helps justify the extra time taken to articulate the environment assets to the same level of detail as characters.
Environment work is fucking awesome fun though. I've had the chance to research cultures , artstyles, geographical locations and periods in history that I've learned a lot about that I otherwise would not have.
I've remade orwellian britain, played with steampunk, did the ww2 renaisance style, egyptian, scifi, fused Islam with Giger, looked toward russian / siberia bunker compounds, dug up info on rome / greece before all the paint came off, got into all kinds of weird art styles I never would have otherwise.
It really is quite diverse and challenging and rewarding work, the difficulty simply is in how much more of it is down to you and your willingness to read a book and research how things work. Most difficult is the ability to see the overal, you dont aim to hit a home run every time you come to bat, often you need to punt it, hold back so that the whole team comes together at the right time.
Individually, some assets do not impress, some of them impress more than characters, but collectively, its an incredible thing to see it all coming together in a level and then be able to run around in it.
r.
[ QUOTE ]
And this is why i love making weapons, the one thing that is ALWAYS on screen and close up. Very rewarding making something badass that will be seen 100% of the time
[/ QUOTE ]
I've not seen any guns in Tony Hawks. Or Bully. Or Need For Speed. Or Mario. Or Monkey Ball. All of those do have environments though.
[/ QUOTE ]
Obviously talking about shooters, but thanks for the well thought out response anyway. I do lots of environment work on my project, but what i was saying is weapons are the most fun to me becuase they are always going to be onscreen. I figured it was on topic because we were discussing the joys of the rest of the developement besides characters, unless you want to split the discussion up to levelwork vrs prop work as well.
I think that whether a person prefers doing character art or environments depends a lot on the person. If you like doing social things in real life and find people interesting, you are probably going to like doing character art a lot more. For me, in real life I find other people boring, but I have run top speed clambering over Etruscan tombs and Mayan temples, squeezed through natural caves and storm drain systems with nothing but a flash light because I love exploring real life environments.
One other aspect I like about working on environments is that it helps you see what is actually there in real life. Since I have started doing natural environments I pay way more attention to the plants and animals I see around me.
Obviously doing character assets and doing environments are very different. The technical needs of the environment you are working with change a lot with the engine, where you are usually doing the same basic stuff with character modeling. If you like the technical aspect of game creation then environments are great. In the project I am working on now I am able to make sugestions about how the engine works, so that I can get the technical trade-offs that work best artistically-- of course, it helps if you are on good terms with the programmers.
Since I have been working with smaller companies I have been able to do concept art, characters and environments. I would be lying if I said I loved tweaking every little plant texture for dozens of plants, or that modeling a trashcan made me bust a nut; it wouldn't be true, but tweaking character assets can be a chore also. That is why I say it comes down mostly to who you are.
thank you thank you thank you thank you thank you thank you thank you thank you thank you thank you thank you thank you thank you...
more env workflows from other people would be cool too!!
thx again!
To prove it, I can show you the horrible pink environment I did for the excellent MoP's Xemakos exmaple...
I've been struggling with this destroyed building for a week now, trying to reach that GoW quality. I ended up breaking into pieces and working on them individually because I have never done anything as complex as this before.
Though Ror's walkthrough of his process was very useful, a proper tutorial would be great (for myself and others).
I know how to make high poly and low poly characters, uv map, mirroring, approriapte polycount, joint deformation and all that, but don't know much about environments. Can you build the entire level in something like Max and then import that into the game (or level editor)?
Many more questions, but I think it would really just be a case of seeing someone do an environemnt that would give almost all the answers. I understand everyone is busy, I'm not demanding a tutorial, just saying it would be great if there were some up to date ones about
Simply understanding that things should be split up into smaller chunks and the overall workflow is the thing that really is illuminating when people are new to hi poly.
The rest, like everything else is just about putting in the hours.
r.
The rest, like everything else is just about putting in the hours.
r.
[/ QUOTE ]
Thanks Ror for the info good stuff !
Even as that may be, there's nothing else I'd rather do. I love environment art. As Ror stated - it's 90% of the screen - you see it at all time. If it's not nice looking/convincing, the whole visual falls to pieces.
If you are building a large streaming world, you'll have 2 types of textures on a building - unique textures and generic textures. Generic textures can be kept in memory for much longer and shared across many buildings, then each building will have it's own unique textures.
Avoid perfectly straight lines. If you are building a row of buildings, set some further back, and some further forward. Use extended extruded trims to break up that line, put low wall or orante railings outside them.
You want to think of a second level of 3d. Buildings don't just have height, width and depth, they have extruded windowsills, intruded windows, mantles, buttressing, drainpipes, overhangs, balconys, air conditioning. I can see every one of these features right now looking out of my window.
Avoid hard edges between materials - grass to pavement, ground to buildings. Always blend these - either with alpha'd oveylays or with vertex alpha. GENERALLY your camera will be at this level, so don't worry too much about this the higher you go.
CHEAT! Cheat as much as you can. Reuse work and textures as much as possible - its better to spend a week on something you plan to reuse than half that tme on two things. With reuse, even a simple texture flip or multipass overlay will hide the fact thats it's a reuse ost of the time. If you place two identical items next to each other, people will notice, but if you copy an entire building and put it on the other side of the street 4 or 5 buildings further down the street, no-one will notice.
I have a couple questions though, first off, Ror, where is that first tutorial you said you made last year?
Secondly, I really only have experience doing simple characters, and because of that, I've never had to get into doing anything like vertex alphas for blending various tiled materials together. Anyone feel like giving me a short "how to" on blending tiled materials together properly? Or at least point me in the right direction?
you guys have raised the bar with that title.
my initial goal was to work as an environment artist in the film industry..but after seeing what's possible with next gen games that goal has changed somewhat
I've also noticed a distinct lack of environment artists in the game forums..everyone wants to be zbrushing the next great character..I prefer zbrushing an old stone column.
Cheers
Stefan
When comparing making an extremely creative environment or character, yup, both can be very rewarding in terms of creativity. However when making a game that's set in a modern day, or similar setting, I'll take making characters to making another dumpster or brick wall, anyday.
The environments in the Final Fantasy games, or in Psychonauts, I wouldn't have minded switching to environments at all. I think it would have been really rewarding, stretched my abilities and taught me new stuff. But if I was on the team for Def Jam, or Madden, I'd rather practice my anatomy and drapery on the characters, than make a gym floor, or a modern day city street. Too often the environments have the potential to be really awesome, but because they are viewed as the "lesser area" of the art by everyone up to the producer often times, you would never feel fulfilled as an environment artist in my opinion.
If we are bored by a subject, we are bored by a subject regardless of what part of the boredom we get to pick at.
r.
The rest is a matetr of whether the engine supports it or not.
We paint vertices in blue, and the engine then knows that it has to use the second texture specified in the material there, but again, i asked for that feature so it's there, i don't know how common it is in other engines.
The alpha would need like a second UV channel, the first for the tiling textures, the second within the UV space nicely unwrapped (like for a skin) where you paint in grayscales the amount of blending. Again, engines do it differently.
And that's why, it doesn't really matter how you do things for your portfolio as long as they look good and right. Techniques can be discussed in an interview or be learned after being hired.
Just use whatever techniques your engine of choice gives you, or you can stay within Max or any 3d app and blend things in the way they work there.
[ QUOTE ]
When comparing making an extremely creative environment or character
[/ QUOTE ]
You're the one who has to be creative, you can create any environment in a incredibly boring way, or make it very fun to work on. I think part of the problem we have with enviros is that some of the artists doing it see it as a boring thing in the first place and don't try to make it fun.
Take GoW, it could have been the most boring setting to work on since gaming history, the 1000th game to use some retro scifi look but... they went and made sure it's great to work on, rather than just sigh and get it done, they made something awesome of it. The same can be done with pretty much any art set. As an enviro artist, you do have usually enough influence, and nobody will try to stop you from making things look better and more interesting ...they usually wont push you to do so either, which kinda sucks and ends up with uninspired 'meh' enviros..
Granted, something like modeling a ring for Fight night 25 isn't very exciting, but modeling the characters for that game doesn't seem any more creative and exciting either.
It'd take 10-15 minutes to make a list of 5 Environment art Tips/Techniques/Goddamn rules. We all do this then link to each other. Job done.
I also can't think of anything more challenging than making realistic people/face textures. Its a lifelong challenge and keeps me motivated.
Fundamentally I believe the two disciplines are very different and although they have things very much in common, you need a different mindset for each of them.