Home Technical Talk

Is topology still important?

Ever since I picked up 3d topology was my favorite thing but judging by how every thread and resource I find is ancient it feels like I missed the moment when good topology became automated or irrelevant... Or something. Can anyone tell what's going on?

Replies

  • sacboi
    Offline / Send Message
    sacboi high dynamic range
  • PolyHertz
    Online / Send Message
    PolyHertz polycount lvl 666
    Good topology is still important for characters and anything with deformation, or if you're going for a custom normals based workflow (like Star Citizen or Alien Isolation). Outsourced work with messy wires is generally frowned upon as well as it makes cleanup a real pita for whoever the asset gets handed off to. But yea a lot of artists who have complete control of an asset from start to finish (especially if using a boolean based workflow) choose to ignore topology these days so long as they feel it wont impede their ability to get the results they're after.
  • FrankPolygon
    Offline / Send Message
    FrankPolygon grand marshal polycounter
    Topology is still important and best practices should be used when appropriate but it's important to remember that it's still part of a larger process. Simplifying topology layouts into good Vs bad, based on arbitrary technical elements like quads and triangles, can be useful for learning the basic concepts of modeling but it's less helpful when it comes to evaluating the tradeoffs of individual topology layout strategies.

    There's a popular academic approach to topology layout that tends to focus on manual control and technical perfection while ignoring certain practical considerations. That's not to say that this approach is invalid. It's perfectly fine for learning and exploring how the underlying geometry effects smoothing behavior, animated deformations, etc. The real problem with this approach is it tends to over emphasize the importance of relatively minor technical elements. Often at the expense of certain art fundamentals and process efficiency.

    A purely academic approach to topology can be appealing because it provides a clear, logical structure that has a number of technical elements to fall back on when scoring model quality and individual learning progress. Where this can really fall apart is if an artist focuses more on perfecting minor technical aspects, like having all quad grid topology with evenly spaced segments, while ignoring the important fundamentals, like having accurate shapes or working within resource constraints.

    Starting with manual topology layout and focusing on the basics of all quad geometry makes it easier to teach the fundamentals because it tends to eliminate some of the variables but there's a fairly wide range of modeling workflows and topology strategies that produce usable results. What's acceptable often depends on on the specific goals for the project. Without the appropriate context it becomes necessary to operate on assumptions that may not provide a fair or true picture of which approach best serves the need of a particular project.

    When defining good or bad topology it's important to look at:

    • What will the model be used for?
    • What are the technical constraints?
    • What are the time constraints?
    • What does the end result look like?

    It can be fun to manually create clean, even topology that's nice to look at but all of that work is functionally useless if it takes too long to create or if the extraneous geometry compromises the underling shapes. Another important thing to consider is both tools and processes continue to evolve so there's significant competitive pressure from alternate workflows that produce comparable results in less time.

    Generalist workflows still exist and are still relevant in certain situations but a lot of modelers also choose to specialize since this allows them to optimize their process and create competitive advantages within their field of expertise. What's optimal for characters is different from what's optimal for hard surface props and so on.

    Institutional inertia and hysteresis tend to perpetuate negative emotions towards newer tools and workflows. This tends to be especially true when new tools or workflows provide a significant time advantage over existing tools and workflows that required a significant time investment to master. Whenever that's the case any gate keeping or trash talking is more of a reflection on the individual making the comments than anything else.

    That's not to say that every new workflow idea is a winner or that a conservative outlook on upgrades is bad. There's specific reasons why some artists and companies continue to use older tools and workflows. It can be difficult or expensive to change tech and workflow process in the middle of a large project and there may be specific competitive advantages to using established processes in an existing market.

    Arguing over potential theoretical issues can be a real waste of time. Just because a popular artist or a teacher at a certain institution does things one way doesn't mean that's the only way to do things. Results that impact the player's experience and make the art creation process more efficient tend to be more important than untested personal opinion and academic ideology.

    Making samples can be a really powerful tool for evaluating something before deciding whether or not to use it on a project. Objectively compare the results to the amount of resources used then choose the tools and processes that best fit the goals of the project.

    When it comes to building your own art process the important thing is to research, test and evaluate.
  • poopipe
    Offline / Send Message
    poopipe grand marshal polycounter
    Untidy geometry ruins the life of everyone who encounters it.
    Including the gpu
Sign In or Register to comment.