Confused about high poly asset creation

SmilingMountain

Hello Polycount,

I'm trying to improve my prop creation skills with a view to eventually creating some half decent portfolio pieces. I've got to say I'm a little confused as to what the current workflow is for creating high poly assets (hard surface). I originally learned the SubD workflow (manually positioning edge loops), and I still prefer it to any other method for hard surface I have researched so far. I see a bunch of artists using Zbrush for hard surface stuff, which is intriguing, especially the boolean/ dynamesh workflow on the wiki. Personally I find that for creating an accurate High Poly, for most objects, creating and placing correctly shaped boolean operands can be just as time consuming as SubD. I'd be interested to know how other people feel about this.

I've also tried out the double smooth/ open subdiv method which I find gives a good level of control and allows non-destructive editing of the underlying mesh without worrying about edge loops. However you still dont get the level of manual control over edges that you do with subD. I haven't tried using quad chamfer yet but that seems like an interesting technique too. There also seems to be some boolean witchcraft you can do in modo to create high poly meshes but I haven't looked into that yet.

So does it actually matter which of these methods you use as long as the HP bakes down well? Or would it be beneficial to proritise learning, for example, the ZBrush or Modo method, if that method is going to be more widely adopted in years to come?

Do artists working in the same studio all tend to use the same technique for creating High Poly? It could be my imagination but it seems to me that the different HP modelling methods all give subtly different edge definintion. Therefore surely maintaining uniformity across a project would require all the artists to use the same modelling technique? I'd really appreciate it if someone could make this a little clearer for me, at the moment I'm really not sure what I need to learn to keep up current methods.

Thanks for taking the time to read this.

Eric Chadwick

Which method you use depends on what you're comfortable with, and whether you get good & fast results with it. 

People within the same studio often use different methods, in my experience. Depends on the management, whether they are open to different workflows.

Depends on whether there is a lot of back-and-forth between different team members, usually not though. If there's a problem in a highpoly, usually teams will tell the original artist to fix it.

Depends on how important the asset is, vs. how fast it needs to get done. The boolean/double-smooth methods are great for making fast assets that don't need a lot of finesse.

Some people are really fast in Zbrush, and prefer the organic workflow. Others are faster in subd, and prefer the precision.

My advice is to get comfortable with multiple methods, see which you prefer.

SmilingMountain

Thanks for your relply, that makes things a lot clearer. I'm currently getting to grips with the boolean/ zbrush method and like you say for some things it is crazy fast. At the moment I think I probably get faster results using SubD modelling so I'll stick with that method for the time being. Thanks again.

Dan Powell

As Eric has said it's down to preference and the studio's pipeline, but here's the pros and cons to each I've found personally. 

Traditional SubD:

• Very fine control over topology; it will do exactly what you tell it to
• Non-destructive and flexible, you can make adjustments easily 
• Can be slower and require a lot more knowledge to create certain tricky shapes

Dynamesh:

• Less fine control; you rely a lot more on the automatic processes in ZBrush such as Dynamesh and Polish.
• Less convenient to make adjustments (because you'll probably be back and fourth between ZBrush and 3DS Max for example, to make a small change, then have to do the whole Dynamesh-Polish process again.)
• Fast. Really damn fast.
• Difficult shapes can be easier to make with Dynamesh than SubD - because ZBrush handles the smoothing and geometry for you. 

I prefer to use SubD as much as possible - just because it's the process I'm most comfortable with. I'd be tempted to use the Dynamesh method when I just wanted to get a shape done and know it won't require much back n' forth between 3DS Max and ZBrush.

Plus, sometimes I like having clean, Isoline High Poly wireframe to show off - the wireframes Dynamesh produces  can be filthy (but they still work fine in practice!) 

NoRank

I usually use some of subD and some of Booleans workflow, it kinda varies depending on what model I'm doing. 

You will get different different edge definition anyways, even if you only use one kind of technique it may happen.