Guide to keeping 3D artists happy?
polycounter lvl 9
G'day poly count. I've lurked here for a little while and it's time to post a thread to get the ball rolling.
I'm a 2D artist, illustrator border-lining a designer roll. I will be working in a project group heavily over the next few months and I haven't had much of the pipeline experience that I'd like, so I'd love to know a little bit more from you 3D people on how to do my Job properly
So as a 2D artist there are certain things that I've noticed are easy ways to make a 3D artist hate you. For example, Painting indicative details rather than clearly illustrated line work makes lives difficult because the purpose of orthographic art/concept work is clarity for the modelers. Because clarity is key and ambiguity doesn't help anyone.
I know the best way is to practice the work of people next in the pipeline to you to know what kind of information makes your job easy/difficult, and I've done a little bit of that, again not as much as I'd like. Nonetheless, are there people here who've had any experience with illustrators/designers who perhaps don't understand the needs of the modelers/technical artists, and what are some common pitfalls that you've run into on our end that might make you want to strangle us? How do we avoid them? All advice welcome
I'm a 2D artist, illustrator border-lining a designer roll. I will be working in a project group heavily over the next few months and I haven't had much of the pipeline experience that I'd like, so I'd love to know a little bit more from you 3D people on how to do my Job properly
So as a 2D artist there are certain things that I've noticed are easy ways to make a 3D artist hate you. For example, Painting indicative details rather than clearly illustrated line work makes lives difficult because the purpose of orthographic art/concept work is clarity for the modelers. Because clarity is key and ambiguity doesn't help anyone.
I know the best way is to practice the work of people next in the pipeline to you to know what kind of information makes your job easy/difficult, and I've done a little bit of that, again not as much as I'd like. Nonetheless, are there people here who've had any experience with illustrators/designers who perhaps don't understand the needs of the modelers/technical artists, and what are some common pitfalls that you've run into on our end that might make you want to strangle us? How do we avoid them? All advice welcome
Replies
1. Make your orthos line up!!! If you make an ortho of a concept and the details don't line up it actually harms the 3d artist's process more than it helps. It depends on how closely they are expected to match the concept, but if it's precise, and things don't line up, he's better off trashing the orthos you spent all that time on. Sad for you and for them.
On a serious note, pretty much your standard stuff. Avoid abstract concept art, define the forms, line, etc.
Depends on the artist, but many peeps know their technical stuff, so working WITH them to make ends meet would be ideal.
For example, I read this somewhere, so memory is hazy, but basically, the concept lead made a monster which had alot of tentacles coming out of it (mouth, back, hand, etc) and a couple of artists who had some robust knowledge in what the budget was and the technical aspects of the project, told him that the monster would seriously need to be scaled down, UNLESS it was going to be a SINGLE BOSS.
Long story shorts, everyone hated the guy and had to get some other leads involved to make him understand that this monster wasn't going through, and the monster ended up having only a couple of tentacles on the mouth and being boss, with some radical changes from what it originally was.
Exaggerated foreshortening from obscure angles.
Not every 3D artist likes working like that though. I generally won't model over an ortho so the proportions don't need to be exact and I prefer concepts with suggestive detail because it gives me something to figure out and make my own instead of modeling out ideas exactly as someone else had envisioned it.
1. Limit flowing cloth to the main character in the game.
2. Tentacles suck to animate, it's typically a perf restriction as well.
3. No long dangling thing between the legs, like this: Link
4. Show the animator the final concept so he can do a sanity/technical check on it before it is passed to the modeler.
5. Characters with long fingers look funny when they try to lift something.
6. Lot's of straps on a character is very tough to skin/rig.
+1 So true.. hehe
omg yes... gimme
one thing i would say though, is to actually have a crack at modeling and even animating one of your concepts. you'll quickly come to realise the things that do and don't work in final production, and that will make you a better concept artist.
If a concept is requested to solve an art or design issue, talk to everyone who is requesting the concept to make sure you know what needs to be solved.
I think the biggest theme is that concept can't work in a vacuum because otherwise the concepts will be useless when it actually comes time to apply them to the game.
1/ Don't make impossible shapes. Things don't have to line up perfectly in an ortho, but as long as they roughly work, then it's fine. Leave the non-euclidean stuff to the designers. Generally if you can sketch out a few views of different components or sub-sections it'll weed out the impossible forms.
2/ Remember, unless you're drawing something small or a background, your place in the pipeline comes after the designer, not before. Grayboxes are great because they're generally indicators of a) what goes where and b) what is likely to STAY there. This also applies to characters and enemies - your designers and art directors should be talking about shapes and forms that signify things (a weak enemy will be, a strong enemy will have this look, etc)
3/ For most things you make, props and what not, WILL be abused. They should work at half-size and double-size.
4/ Nail down style early; this saves your modeling and texturing team a lot of work.
5/ Technical limitations aren't just poly count and texture resolution. Post processing, draw distance and overdraw can also conspire to ruin concepts.
Apart from that, have fun and try your best.
So things to remember for me from this is to keep your ideas practical within the scope of the game, understanding the limitations of the engine or mechanics and balancing the distribution of geometry/bones/skinning data where you need it most. So when Hazardous says, "Start crazy as it will be watered down later" I guess you kinda have to trust your team to keep a conservative scope, because sometimes I'm working with our designer and he's always layering on new mechanics and systems and I interject with 'whoooaaa slow down, we only have a few months to make this project" :P
So I guess the process Hazardous describes is a bit like this
I do talk to our modeler as well and he seems to be pretty happy with what I produce for him at this stage, but to a certain extent I feel like there's not -enough- feedback - my brief is 'this is our game, we're doing penguins, go nuts.' So I suppose I'm viewed as the game aesthetics expert, but it feels like without proper feedback I'm flying blind :P
So I also guess it's handy to design to people's strengths - if someone has a lot of trouble skinning/rigging a particular type of 'thing', then limit your use of it until they have more experience with that 'thing'? Or if some animators are better at animating exaggerated expressions and gestures, then similarly scope the designs of characters to be likewise exaggerated.
And yes, I have been doing my own characters and environments as part of my course (I go to Qantm College, Brisbane.) I have my own limitations and weaknesses that some other modelers and animators don't have, so I guess my pool of experience isn't that great of a yardstick for figuring out what's practical and what's not, given I'm still a student.
So short of it is, make things scale-able, practical, play on your teammate's strengths, work within the context of the project, and let as little as possible rely on high-power systems like cloth physics and hundreds of belt buckles, etc.
also hearing that some concept artist's refuse to do line drawings because they think its not and epic mood painting piece ugh
1. Learn basic 3d modeling
2. Block out shapes in 3d
3. Paint over 3d shapes to create concept, no "cheating", stick to the actual shapes in 3d.
4. Present modeler with 3d blockout.
You can also use the blockout to paintover accurate orthos too, weee! This is mostly for hard surface stuff, but might help for organics too? Really productive for anything mechanical.
Bonus points if the blockout mesh is checked ingame for proportions before any of the heavy-lifting type modeling work is done, this is more of a project thing than a concept artist thing though.
"Translating" a concept with half-baked perspective and lines drawn to "look cool" regardless of 3d space can be a real pain.
If you're not able/do not have time to learn basic 3d modeling skills, work with the 3d artist while doing sketches, have him whip up quick blockouts to match your sketches, then do final renderings over said blockouts.
+1 :thumbup:
Like for example, if I ask the concept artist to draw up a window and a room that the player will walk past and be able to look in, and they send me this:
Then they're going down the well.
Also Brendan, I don't know if it's just my permissions on this forum or something but I'm getting a 403 error on your image, so it's not displaying. I'm getting a lot of those on this forum so it could be an issue on my end. Maybe upload to imgur or something?