Creating metal surfaces

 wrote this short text some time ago to help people when they were creating metal surfaces and thought it could do with being reposted here. I've also added a section on reflections at the end so without further ado...

Creating metal surfaces

This is just a couple of things I keep in mind when creating metal surfaces. They are by no means set in stone, it's just a few things I've noticed might help when creating metal surfaces for next (well, current really) gen use.

The most common problem with normalmaps for metal is that it's easy to overdo it. Depending on the engine you are using, you might not be able to get sharp, crisp normalmaps unless you use fairly large texture sheets so for the most part, the fine grainy detail will have to be left to the specular. Scratches and such do work however and usually need to be in the normalmap to make it look convincing (just keep in mind that texture compression might ruin the really fine scratches, those that are just one or two pixels wide).

Below is a compilation of different types of painted metal surfaces:

As you can see, painted metal can look very different depending on what type of paint it is and what it has been subjected to. It's important to decide what type of paint your surface has been painted with since it will affect both the normal and the specular maps. Having a history in mind when you create the textures is useful since it will help you decide how to treat it. Is it an old surface with lots of scratches, new with a few dents, so old the paint has started to flake, was it painted without being cleaned up first so you have really lumpy paint, has it been polished often? Having a history will also make the object look a bit more interesting rather than "generic metal box #6". This is of course not mandatory, it's just a mental exercise that might help when creating the textures. I should add that having a history in mind works for any kind of texturing, not just metal surfaces.

A common mistake is to create a normalmap that has a lot of noise in it but this makes the surface look more like leather/elephant hide than metal. If the texture is a 2048 instead of say, a 1024, the noise might have looked more like the grainy paint above but in most cases, it won't look like convincing metal. What I would do is to tone the noise down considerably and just keep the scratches where the paint has been scratched away to reveal the metal. What we need to keep in mind however, is that scratches that are just a few pixels wide will most likely look lumpy and not very pretty once the texure has been compressed. So keep larger scrathes, dents etc in the normalmap but be careful with the finer detail.

In addition to that, a specular with fairly high contrast can work well since paint is usually fairly matte and non-reflective (there are of course exceptions, it is as always down to what type of surface you want to mimic) whereas scratches can be very shiny since the metal surface beneath the paint shows through. Another little tweak can be to make the diffuse colder. Sometimes a texture can have a fairly warm feel, with brown and almost light orange tones as a base which can make it feel slightly wooden. By simply changing the hue on the base colour layers so they are more blue-tinted it can feel more metal-y. It's a cheap trick but it works. It is also important to tweak the specular curve of your material since the way a game engine renders the light shining on your material can have a large impact in addition to your textures.

When it comes to the specular, it's almost better to put a lot of the fine detail there rather than in the normalmap for the reasons outlined above. Some detail needs to be in the normalmap and mimicked in the specularmap, such as scratches while otheres shouldn't be in the specular at all. One such example is if you have a plain metal surface that has been dented. Again, this depends on what kind of effect you want. If it's a recent dent, there won't have been time for dirt and dust to collect so it hasn't become dull yet. If it's an old dent, it will be less shiny than the surrounding areas since it won't have been cleaned as well as the rest of the surface. And just to add to the confusion, some things should be in the specular but not the normalmap such as oily surfaces, grease marks, handprints etc. This is where the idea of having a history of the objects comes in handy again. Has the object been manhandled a lot, dropped, left to stand in an oilslick etc. All these things combine to make the surface look interesting and if you put the right detail in the right map, it will look fantastic.

Of course, metals need to reflect something so having a good reflection map is the last piece of the puzzle. Here is where it can come down more to what looks good than what it would realistically reflect. Unless your surface is very clean you can get away with using a reflection that doesn't have anything to do with the setting. As long as it looks good, that's all that matters. It's also down to memory usage. If you take a cubemap for example, it's made from six separate images so the memory usage is fairly large for just a single cubemap. If you can reuse one cubemap on many surfaces you'll save memory which can then be spent on other things, such as higher resolution textures on selected objects. As always it's a trade off between what looks good and how much memory you have left for the shiny stuff. That's also why slightly dirty, painted metals is a bit easier to cheat with, you won't get a clear image of what the reflection looks like. Clean environments are always harder to get to look good and interesting in my opinion. By the way, a very good example of succesful clean environments is Mirrors Edge, it has great usage of lighting and a very clean style. Well worth a look for the visuals alone (apart from some of the fighting sequences, it's a good game in my opinion).

Since this whole text is about metal surfaces, I have to mention Dead Space. It has a very well designed look and the way they use specular and normalmaps in that game is almost exemplary of how to do metal surfaces right. The normalmaps are clean, crisp and there is very rarely any noise in them. It is a very well made game, excellent design on pretty much all levels. Great atmosphere, lighting, texturing and a good story with good voice acting. Speical mention goes to the integrated HUD and the sound design as it is, in one word; amazing. Playing it with headphones late at night was an expereince I won't forget any time soon. As an example of Dead Space texturing (and a cheap way to get more images into this post), here are two screenshots that show their restrained normalmaps and nice use of specular. It looks a lot better in motion, and the lighting is pretty much all dynamic so you'll see some really nice effects all through the game

Until next time...

"The Unspoken Truth About Managing Geeks" - Excellent article

A poster over at The ChaosEngine (a forum for people working in the gamesindustry, more about that below) posted a link to this article about managing geekss in IT companies and it is a very good read. It definitely applies to the gamesindustry as well and I've seen this happen in companies were I've worked. Makes me wish managment teams all over would read this and take at least some of it, if not all, to heart. Pretty much everyone I've talked to that have read it agrees that it's pretty much spot on! So, without further ado, here's the article:

The Unspoken Trust About Managing Geeks

The ChaosEngine

The ChaosEngine is a very good forum for people who work in the games industry. It has a private section where you have to prove that you are a developer before you can join it. The way to do this is to send a mail using your work email and if possible, to get another developer already registered on the forum to vouch for you. If you work in the games industry, you could do a lot worse than to join, it's an excellent place to talk about the industry without worrying about rabid fanboys butting in and just have a good time with fellow developers. A lot of information is avaliable there, from art to programming to managment and business advice as well as information about various companies. It also works as a watercooler where you post stupid/fun/cool/useful links and articles, such as the one above!

As an example of what this community has accomplished, the Tim Langdell and Mobigames dispute over the name "Edge" (and if you haven't heard of this, where have you been?). The blog ChaosEdge will hopefully be familiar to you since they've been posting a lot of interseting information regarding the case. It is maintained by a few people from The ChaosEngine and a lot of the stuff that has been dug up on Tim Langdell has been done by the excellent internet sleuths over at the forum. In short, it's a pretty awesome place.

Attitude and Criticism

One thing you will face as an artist in the games industry is criticism. It's unavoidable. Your co-workers will tear your work apart, your leads will tear your work apart and your art director will tear your work apart. Maybe not devastatingly so, but they will still find things you can improve and undoubtedly you will be told to change things. So the right attitude is pretty important, on both sides if you want to keep the artists at least civil with eachother. Giving and being able to take criticism is probably one of the most important skills to have in order to stay moderately sane when working in this industry since you are working with a group of creative people who (hopefully) want to create the best looking game that they can.

In order to do so we try to help and push eachother to new heights and there will inevitably be occasions when someone will look at your screen and, depending on their social skills, say anything from "That's shit" to "Looks good, but you could chage that bit to this and this bit here to that, could be cool" to "The publisher has changed the date, we need to cut this down". That last one is especially fun. What I'm getting at is that no piece of work survives contact with your co-workers so being able to distance yourself from your work is very important. I'm not saying don't be passionate about what you are doing, you have to be if you want to be good but don't let your work equal your self esteem. I've seen a lot of young (and not so young) artists starting out and as soon as someone points out a flaw get very defensive and try to fob it off with "It's supposed to look like that, his nose was grafted on after a dog bit it off and the doctor was a cross-eyed drunk with parkinsons so he couldn't do it very well so that's why the nose looks like that, be quiet". Which isn't really helping anyone. Of course, you don't have to rely on your co-workers since there are a few simple ways to check your own work. One of them is to just mirror whatever you're working on, either in your tools or by looking at it in a mirror. After a while your brain gets used to the way things are placed in relation to eachother so flipping it will force your brain to re-evaluate the image, helping you spot flaws or maybe if you nudge that bit there a bit to the left it will look even better. Or you can flip it upside down. Another way is to play with colourschemes, again forcing your brain to refresh.

Being able to look at your work with critical eyes and a bit of distance is going to help you in the long run and make you a better artist. Being able to look at other peoples work and point out things that could be better in a nice way is also something that will help you in the long run, especially on the social side. As an example, CGTalk has battled with this for some time, on how to get people to give constructive criticism instead of the usual "AWESOME!!!11" and "That's shit!" and it's just as important, if not more so in a studio. You don't want your co-workers seeing you as an obnoxious idiot so when commenting on peoples work it's a good idea to articulate why something is great/shit and what could change. Preferably at the same time. Criticism is usually absorbed a bit easier if you point out something that is good at the same time as you point out something that is bad. Of course, once you get to know the poeple you work with you will know who you can walk up to and say "Jesus, what the hell did you do to that?!" and have a laugh and who is a bit more sensitive about their work where a more diplomatic approach is needed. I'm not saying you have to be overly nice about it but you do need to be honest. If you give honest criticism and explain why and perhaps think of a few ways it can be improved, people will (hopefully) appreciate it more. When you ask for advice they will hopefully treat you the same way since that is the best way to improve. Being nice to someone can actually do more damage than good in some cases. Tell the truth, but do it in a polite way and you get the best of both worlds. There's no need to put someone down just because they made a mistake, that is not constructive and will only generate ill will which might come back and bite you later.

Speaking of attitudes, having an open mind towards new techniques, tools and programs is really important in my opinion. In this industry the technology is always evolving and there's almost always something new popping up that looks cool and interesting so to dismiss it wihtout giving it a look seems counterproductive to me. On one hand it's easy to get stuck in your ways and not really try to evolve ("it works for me and I'm happy with it") which isn't necessarily a bad thing in itself but it does mean you can get left behind in a way. Take the not so recent transition where we started to use normalmaps. It's been used in games for at least 4-5 years by now but there are still games where the normalmaps are a mess. Wether this is due to lack of knowledge or lack of passion, the result is the same, it looks like crap. Having an interest in learning and sharing the knowledge is as much as anyone can do since you can't actually force anyone to learn new stuff unless they want to (well if you're the lead/AD you can force them...) and it will show that you're trying to improve both yourself and the general quality of work which you will hopefully be encouraged to. There are of course places where this isn't appreciated or might even be actively discouraged but you can't win all the time.

In conclusion, being able to give constructive criticism is very important as is being able to take it on board. Having an attitude of "I'm the greatest thing since sliced bread!" and ignoring what your co-workers are saying isn't going to do you any favours. Being open to knowledge no matter the source and being willing to share is the way to go. This might be a bit obvious to you but from what I've seen and heard, it really isn't that apparent to people that we are evolving as an industry every day, every week. New tools, new ideas and new processes will always come along and complement the old ways. Being able to seize on these and bend them to your will is not a bad thing and it saddens me when people treat them with suspicion. Don't fear the change!

Last of all, I want to apologize for the lack of images. This is supposed to be an art related blog and no images? Shame on me.

-Day after posting edit-

I remembered a few other ways to refresh your brain in regards to your own work. Taking a break if you can is a good way (look at this addendum as an example). A coffee break or ever better, a day or two without looking at it will make you come back to it and see it in a new light. Of course, in the games industry you are usually working to deadlines so being able to leave something alone for a day or two isn't always possible. If you're working on a texture, taking a step back and unfocusing your eyes will draw your attention to areas that stand out. If you're doing a 3d model, assigning it a completely black material will allow you to focus on the silhouette, making it easier to see how your model reads from a distance without any detail to distract the eyes. Team Fortress 2 is worth noting in this since each of the nine classes have distinctly different silhouettes. You can spot which class you're up against from a mile off thanks to the excellent character design. The Australian painter Bob Abrahams has a long list of things you can do to re-evaluate your work on his blog, worth a look if you want to read more about tricking your brain.

Worth noting is that this being the games industry, odds are you'll be revisiting things you've created in the past to give them a bit of extra polish and to optimize the assets. For example, those 16 sided cylinders that seemed a good idea at the time has turned into a problem and needs to be reduced to 6-8 sides. This is again something you'll get used to, things inevitably change. The only thing we as artists can do is to try and do the best job we can with the information we have, accept the changes imposed on us with dignity and then have a rant at the pub after work. In an ideal world we wouldn't have to change our work but then again, changes where you can add more detail and polish isn't always bad. The kind where you have to chop a couple of levels up and mash the pieces into a new one is less fun however but sadly it happens. In the end you'll just have to roll with it (but no one says you have to like it).

Consistency and Efficiency

One of the problems when creating environments (and any kind of assets really) is to keep things consistent, especially if you have more than one person working on a project. When I say consistency I mean the general feel and design of a scene, level or asset. Texturing, lighting, modeling, shading and all the other little bits that go in to creating these little works. The reason for this post is a question I got asked on CgTalk and it's an intersting one since in theory it should be quite simple, especially if it's just one person doing the creating. However, in reality it isn't always so easy, especially as your team grows.

For the single artist working on a scene it is usually fairly straight forward to keep things consistent since it's just you and you have your own workflow. But even then you tend to develop your techniques as you go along which can introduce slight inconsistencies in your work. For example, take a project I've been working on for some time, a 2d shooter. I started out with one style in mind and created a few ships along the guidelines I had set myself. However, after a couple of ships, the style started changing as I tried new ideas and approaches. After a month or so, the new ships looked nothing like the old ones which led to more work since I had to redo the old assets. This is of course my own fault since I could've stuck with the original style but truth be told, it didn't really work so it was a good thing the style changed.

So where am I going with this? The importance of concepts and style studies/guides can never be overestimated. Had I done proper concepting and tried out various styles in the beginning I might not have had to redo those ships. The fact that it is 2d doesn't really matter since the same principle applies to 3d as well. It's easy to get excited when you have an idea and just get stuck in with the modeling but what might happen is that you start to flounder a bit later on when the intial idea has been brewing in your head for a while. It starts to change and you might think of new things to add. Slowly your intital idea will evolve into something else and that's can of course be a good thing. But it's usually a bit more efficient to do this evolving on the concept stage since it's a lot faster to do simple drawings instead of modeling an environment. And realising halfway through that you're not sure where you are going or that the first part you made doesn't actually fit with where you're going now can be really frustrating and in the worst case mean that you just drop it.

The same idea is even more important when you have a group of people working together on environments or a bunch of assets. Having concepts that you can refer to, quick mood paintings with colour schemes, detail studies to show the more intricate parts and surface types is worth its weight in gold. And to bring it all together, a Lead Artist or for larger teams an Art Director that can guide the rest of the team towards a common style. The Lead or AD needs to have a vision of how the different parts will come together and be able to guide the rest of the team towards that vision because if they don't, our personal styles will undoubtedly start to make their mark on the stuff we produce and you'll get a disjointed feeling to the visuals. Some parts might look a bit like, say, Call of Duty whereas another part might look more like Killzone or Half Life 2. I'm not advocating that a Lead should walk around with a whip, forcing each artist to obey his every whim but rather that the Lead/AD will keep an eye on what is produced so they can step in and nudge people in the right direction before it would have to be completely redone. Especially before things are commited as a final version in your source control there needs to be some sort of sign off.

To go back to the single artists working on a scene, this is one of the things I'm trying to do with my island project. To go through each step along the way and hopefully end up with a nice little scene in the end and a decent example of a thorough workflow. It does mean it takes a bit longer to have something to put in an engine but I think that the time I save later on when I would usually be at a loss as to what to do is worth the work in the beginning. A good example of the ideas above is Bram Eulaers Unearhly Challenge winning piece: Abandoned. Below you can see the concepts and paintovers as well as the final result.

He has some very nice examples of environments where he shows the concept that comes first and then the finished 3d environment, it's definitely worth having a look at his site if you haven't seen it before.