Good Poly Flow

Hellspawn

Hey there! I'm new to the site and this is my first post. I'm currently enrolled in a Game Art and Design major at the Art Institute of Pittsburgh. This quarter I got to take my intro to 3D modeling class which I love! But I seem to struggle a little with good poly flow. I can get the shape of the objects but when i go to unwrap them there's so many faces it becomes annoying to try and get them all textured.

I was wondering, could you guys give me some pointers on how to maintain a good poly flow? I would really appreciate it =]

Ott

Some more info on what you are trying to model, what software package you are using, and perhaps even a screenshot of a certain object you are referring to would be a good start. Smart polyflow depends immensely on what exactly you are modeling and what the purpose of its polyflow needs to serve.

Hellspawn

I don't have any screen shots of what I've been working on recently. I can tell you some of the things though. The object i had the most trouble with was our last project which was a bottle of windex. Before that it was the usual beginner stuff, Chess set, hammer, dinnerware. Things of that nature. I just have a really difficult time telling how much I do and don't need when it comes to the actual modeling, if that makes any sense.

Ott

Well, there really is no set in stone rule for this sort of stuff....like I said it really is dependent on the type of mesh you are doing. For simple environment stuff you really need to focus on the best edges that are identifying your main shape and silhouette.

If you have a lot of excess edges that aren't really defining the shape, get rid of it. For small game props, you have to be even more careful about going overboard. If I put a bottle of Windex on the counter and I am playing a 3rd person game, do I really need to chamfer every single tiny bevel? Will it even be seen onscreen and ingame?

Try to get a visual idea of how large your object would realistically be seen ingame and ask yourself if you can cut a lot of geometry out of it and still keep the main shape. One of the biggest mistakes a lot of young Game Artists make is thinking that we can just brute-force throw a half million tris onscreen because it's "Next-Gen" and ignore the fact that we still face budgets.

If you want to start with a basic understanding of actual onscreen game art, start looking at games and get an idea of how many polygons they might have used for the objects, or get a copy of Gears or UT for PC and look at some of the assets they create and get a feel for actual triangle counts. (Keep in mind that for every console and every asset and platform you are going to have different limitations and budgets)

I would also suggest you frequent some of the "What are you working on" threads here, as well as this thread http://boards.polycount.net/showthread.php?t=50588 to start getting an idea of how triangle optimization physically works from a development standpoint.

Of course, your best and fastest stream of feedback would be to simply post a screenshot of your asset and let the Polycounters rip it apart

Hellspawn

Haha! Thanks for the great advice Ott. Yeah, I think i just need to practice more. I've already learned alot from the boards here and continue to learn more every day. I'll keep practicing and checking forums to learn abit more aout how it all works.

Once i start getting into more complex shapes and things, I'll be sure to upload them on here for feedback and crit's.