Modeling for Games vs Other Mediums

Zarathustra

I lack experience modeling game assets and I've heard lots of contradictory info so I'd like to hear from those who know, those who do it, especially if you've worked in games and other mediums (ie - film, broadcast, etc). Are there differences to how you model? Are you generally working with straight poly models, or do you build a "proper" high res, subd model and then res down? Are the answers to these questions dependent upon other factors (ie - props vs "hero" objects/characters, scope, time, etc)?

I'd appreciate any insights those of you could provide me.

Thanks a bunch!

Computron

what is your game? a ds game? a low budget 360 game? or a AAA title? in the coming years most AAA title's Props and modular pieces will most likely be done from a high poly just like in film, with some procedurals for things like organic environments.

you will model just the same, except you will need to bake down to low poly models.

Zarathustra

I assumed hero characters would be the same, but I wasn't sure about hard surface objects.

Thanks.

maze

doesnt matter what your game, film, documentary is. Learn to build things as they are done in real life, a window might be seem stupidly simple to model, yet a an inset box doesnt make a window (..or a really bad one) there is more than that,.
Specially when building real life stuff. I work in film and yet I do a lot of low poly modeling as well. (I have the impression that many people think that film is just about high poly, when its really not the case always.)
I remember couple months ago when working on red tails show. Seeing the way planes where constructed helped a lot in modeling, something really important is to always have a good sense of scale. If you are to start something set yourself a "master" scale reference and stick to it., it ll help you a lot. Also well proper topology, looping.

Do not jump on any sculpt application if your 3d modeling skills are not good. It ll just make it worst imo. And about hero assets and non hero assets, the same rules still apply (good topology, form, volume, logical construction)

A well done low poly model can always derive into a good base for a high poly model, and in a production pipeline, that is very important.

Lets say you model a low poly window (to take the same ex..) and your lead likes it and tells you this shot will need that window "highresed" then if your low poly is well done you can reutilise that mesh and keep adding detail.

Hope this helps man!

Zarathustra

Good stuff. Thanks. I totally agree with you. Furthermore, I always stress building to real world scale. It makes your life easier building and can make your life easier in the future if you have to quickly comp or bang out a project and can just load past objects in without playing the eyeball-guessing game of relative scales. I also agree with the need for understanding topology. It might be great that you can bypass all that by working entirely in ZBRush, but eventually you have to get it out of there, which means retopologizing. The best tools for that are useless if you don't understand how the polyflows should be.

To give a bit more insight into why I'm asking all this, I'm tasked with hiring people for whom modeling isn't a primary requirement, it is nonetheless a requirement (and as a modeler myself, I don't feel comfortable having someone who isn't a solid modeler). Anyway, others around me are dismissing my concerns and saying that if someone can model for games, they can model for film and I don't believe that's a universal truth. Could you? Sure. Can every game artist model for film, or conversely, can every film modeler model for games? No, at least not at first. Each medium has their specific needs that you have to be aware of.

Now character modeling should go without saying. Topology is critical for proper deformations. Hard surface modeling, however, could be different. There, topological issues relate to your edge proximities for defining the sharpness of an edge for subd modeling but that's not a concern for straight poly modeling so if all you know is hard surface poly modeling for gaming, you may be out of sorts having to do subd hard surface (unless you're employing some kind of edge weighting). Likewise, if your mind works in subd hard surface, your models are needlessly heavy as a game asset.

maze

sure each industry have its own rules, technical limitations. That I dont argue. But a good modeler is a good modeler. If you are a good jazz musician you can probably learn to play blues gracefully as well.

I dont think a modeler of one industry can model "right away" for another (film, games) because of technical constraints, workflow. But imo those things are "easier" to overcome than having to teach someone what I explained in my previous post. It depends on your pipeline. Sure if you can get someone good at what he does and already with knowledge of the game industry technical constraints in your case thats a plus timewise. Good luck with that!