Best way to retopo a tree-like mesh?
In ZBrush, I sculpted the base mesh for an alien "tree" using Dynamesh. I want to retopo before I sculpt the details, but ZRemesher gives me spirals everywhere, even though I used a ton of guide curves (made with the crease curve brush and the frame mesh function). I tried the adaptive size and curve strength at either extreme and the adapt option turned on and off, but I still get spirals, which I can't see any way of easily fixing. On the other hand, manual retopo in 3DS Max would probably be quite difficult (as I understand it), due to the numerous intersecting "branches" and Max's inability to hide part of a mesh. But I have an idea!... maybe.
A sphere with good topology is essentially a subdivided cube. When you need to unwrap it, the UV edges that get broken are the ones that were originally part of the cube. I call them non-looping edges, but I'm sure they have another name. On any mesh, these non-looping edges form the "skeleton" of the topology. So, if you needed to retopo a sphere, the ideal solution would be to designate the non-looping edges, hit a button, and watch the program automagically fill in the rest. Is there any tool that works like that?
In the case of my tree, a good topology would consist of separate "tubes" for the trunk, each root, and each branch, plus some additional polys to connect them all. That way, I can break them all into separate UV islands. If the aforementioned tool doesn't exist, what would be the best way to retopo my tree?
A sphere with good topology is essentially a subdivided cube. When you need to unwrap it, the UV edges that get broken are the ones that were originally part of the cube. I call them non-looping edges, but I'm sure they have another name. On any mesh, these non-looping edges form the "skeleton" of the topology. So, if you needed to retopo a sphere, the ideal solution would be to designate the non-looping edges, hit a button, and watch the program automagically fill in the rest. Is there any tool that works like that?
In the case of my tree, a good topology would consist of separate "tubes" for the trunk, each root, and each branch, plus some additional polys to connect them all. That way, I can break them all into separate UV islands. If the aforementioned tool doesn't exist, what would be the best way to retopo my tree?
Replies
There are hide options inside of max, a lot of them actually, all over the place...
Hide/Unhide objects: Its in the quad menu, right click the viewport.
Isolate selected: Again, in the quad menu. This will temporarily hide every other object that isn't selected in the scene. You can then "end isolation" and everything comes back. It is very handy, especially if you already have objects hidden and don't want to do an "unhide all". Doing an end isolation restores the scene to the hide/unhidden state it was at previously.
Hide/Unhide sub-object pieces:In Edit Poly there are buttons to hide whatever is selected in the current mode you are in, be it verts, faces, edges or elements. The buttons are labeled "Hide Selected, Unhide All, Hide Unselected".
Customize your interface: You can add these "sub-object" features to the quad menu, so you can hide/unhide objects and sub-objects all in the same place. I used to do this until I started setting hot-keys.
Main Menu > Customize > Customize UI.
Hide/Unhide by Layers: The layer manager and the new scene explore have the ability to hide/unhide objects based on what layer they are on, you can also freeze layers and organize them by nesting folders. So for example you put the branches on one layer, the roots on another and the trunk on its own layer and then nest them in a layer called "Tree". Click the tree layer and they all select, hide the tree layer and they all hide.
Seriously this isn't a problem in max, just your understanding of how to use it.
Some retop options in max for this object:
1) Crunch it.
If it's a static object and doesn't deform its important to know that all objects are broken down to triangles so you don't necessarily need quads. So export your lowest sub-d out of Zbrush and use the pro-optimize modifier and crunch it down to a manageable level. Unwrap using point to point seams, relax it, arrange it and bake.
2) Model it and use the conform brush.
You can rough in the shape using any technique you like, box modeling, edge extrusion and classic modeling technqiues. Maybe use wrap-it to snap your low poly to the surface of the higher while you model. Maybe use it's automatic quad based mesh generator, or if you don't like paying for scripts use Max-retopo. If you need more suggestions on how to model up a quick mesh I can vomit out a few more. It doesn't need to be prefect, just roughly hug the right shape then you can use the conform brush to get it to hug the surface.
The conform brush is in the graphite modeling tools, freeform tab. You select your low poly object, change it from grid "Draw on Surface", pick the target (the higher poly) and paint over your mesh to get it to conform to the surface. You can adjust the offset to push it below the surface or above it. I use this tool to create collision meshes also, a higher offset value will help create a collusion mesh above the surface while a value of 0 will hug the surface and a negative value will to below.
Unwrap, pack it and bake. If you're slacking it might take an hour, if you actually know what you're doing and have done it before 10-20min.
I knew about this option, but like several of the tools in Max, it does absolutely nothing when I click it. The problem is that, since I don't want to buy Max, somebody else (who has school connections) is letting me use their computer with Max already installed (presumably legally). However, I can't go messing about with their settings or installing plug-ins, so I have to work within certain constraints. I never know when I'm misusing a function or experiencing a bug, so I naively assumed that it was impossible to hide sub-objects on some types of meshes (ie. editable poly).
Thank you for the quick (and lengthy) response!
There isn't one way to do everything, be very skeptical of the people that say there is. There are ways that might work better but it all depends on a lot of factors . You can get there by many roads, some might take longer than others but they will get you to the same place, eventually.
I'm guessing you started off with Zbrush and are learning the rest of the process? "pixologic" is different than conventional 3D logic that rules all of the other major pieces of software. Many people try to apply their zbrush learning to a standard 3D application and fail miserably. You need standard 3D logic and "pixologic" to get it all working, all of one and none of the other won't get the job done.
About easily available tutorials...
The bulk of easily available tutorials are often made by people who are just starting out and learn something cool and want to share it. That doesn't mean it's the best for everything, even if it is their hot-new-toy. After they come down from their high, they'll use it as intended and probably a lot less often. Most people who do it for a living, don't get excited about what they see as mundane modeling practices and they don't bother to record tutorials because the majority of their peers work that way and it's not new or exciting. So what you get from "independent users" is often pretty janky, overly hyped and not exactly going to cure cancer even if they say it will, but sometimes there is value in it so they are worth watching, just not worth basing a religion around.
Watching other people work (while incredibly boring and pretty creepy) is one of the best ways to refine your methods and pick up a toolbox full of tips and tricks. It also helps to record yourself, but don't publish it, review it. Find the flaws, the road blocks and what wastes time and could be improved? If your mouse travels across the entire screen every other second just to hit a button, maybe it should be in the quad menu, or a hotkey. If you have multiple hotkeys that do the same thing maybe you should combine them into one function. That isn't sexy or as exciting as concept sculpting but you figure out how to get the job done incredibly fast.
All skinned cats lead to Rome?
The methods I outlined are what I think are the easiest for any user, based on the info I have in front of me. But there is a lot I don't know about the object and it's end goal so that opinion could change based on new info. Either way, knowing many methods and picking the right one for the job at hand is a valuable skill to have. They are pretty standard and easy to do, so is strip modeling, so use whatever works best for you and the project you're on. You might be more familiar with a slower workflow, but it might take you longer to initially learn a faster way to work, so in that particular case it might make sense to stick to the slower method. If you wanted to make a career out of it then it would be important to learn the different ways to work because they are invaluable later on down the line.
If I had to guess, you aren't creating the right conditions for the tools to do what they are intended to do. It's probably a misunderstanding of what the tools do, maybe you don't have anything selected, maybe you don't understand the difference between sub-object mode and object mode? It could be a lot of things but it is probably not a broken feature of max. I and many other people who use max everyday aren't repoting massive problems with core tools that have been in max for over a decade and are used every day (by a lot of people). The chances the problem is PEBKAC is pretty high.
F1 = Chucky (your friend till the end, hiddey-ho) but without the murder and mayhem.
The 3dsmax help files (F1) do an incredible job of explaining each button. It even does a decent job of detecting what tool or menu is activbe and taking you to that when you pop open then help file. The help doc often gives you clear examples of how the tools work and most of the time a quick tutorials on how to use them. Seriously F1 is your friend, many pros still punch F1 every once in a while, don't feel stupid hitting it and don't assume its all junk and that you know more than the guy who wrote the help file. Be careful not to be the "I know more than the creator, the creator is stoopidz" guy. That is just the The Dunning Kruger Effect on full display for the world to see. Don't be that guy. Be humble, realize that you might not know everything and be willing to learn new things, even from people you think are beneath you. There is a reason the help files exist, they are... helpful =P
"But, but... I wear many hats so I can't focus..."
This is poisoned soup that is about to be served, in the dinning car of train, that is about to leave the rails. In other words a recipe for disaster and a train wreck all rolled into one. It's entirely possible you are spreading yourself way too thin and need some help. The types of games you can make in isolation are very limited (I'm thinking of ET for the Atari 2600). That pool gets even smaller if you aren't a true master of all disciplines.
The learning bar for each discipline is coming down and games are getting easier to make with smaller teams but its still a lot of specialized work done by people that are dedicated to their particular craft. When one person spreads themselves across it all, with superficial knowledge of each thing, it hurts the end product.
It also hurts their chances of landing a job because they are terrible at everything. I'm not saying everyone has to specialize but you do have to be good at something in order to be hired.
Often the one-man-show people, are doing that not because they are awesome at everything and can pull it all off, but because they lack a support network that can help them deliver those things (see Dunning Kruger Effect and or,the anti-social loner who thinks trusting people is dumb). You might want to spend some time networking and find some people that are interested in the over all project but are fairly competent at one or two particular things. Those connections and that team experience are very important. It's the next set of questions that gets asked after "can this person do this particular job".
So be careful and don't spread yourself too thin, it can really sink your project and hurt your chances of landing a job.
They are all interconnected. The more you know about the not so fun stuff, the better you become at the fun stuff. For example if you know how something will deform it will help influence your design. If you don't, you'll probably inadvertently make a nightmare for someone else (or yourself).
Knowing proper topology helps you create models that will rig, skin, deform and unwrap well, knowing how that plays out while concepting really helps you create fast and efficient models that do their job well. It also helps you keep your design entact when and if someone down the pipeline tries to steamroll you with changes.
Them: "You have to redesign the main component that gives the character its characters it's messing up all of this technical stuff!"
You: "We talked about that because I had the same concern when I was drawing up the concept. You said we would handle it by ..."
Them: "Damn it, I was hoping you forgot about that."
If you want to texture something the best way possible, you need to know how to unwrap, to unwrap its critical to know how to model and its really helpful if you have a rough plan of how you'll unwrap the object before you even start sculpting. It helps you plan your UV seams and your material breaks, what details go on what sheets, how much flesh do you have, opacity maps, leather, metal, what gets packed where with how much padding, what kind of shaders you'll be using and how to optimize it all so it runs with the smallest foot print possible.
"I only like doing the creative bits so I'm not going to worry about that other, junk."
Very few places will let you dump a sculpt in another persons lap and call it good. You have to be an amazing-fantastic-super-awesome-god-like sculptor for that to happen. They basically have to hire two people to do the same job that it normally takes one person and there are lot of people that can fill that role all by themselves. Knowing that other stuff really well, will influence your design choices. Also creativity isn't relegated to only drawing and painting. Some of the most creative people I've worked with couldn't draw a straight line to save their lives.
The best Programmers, managers, animators, creative directors, art directors, designers, script writers, technical artists, even HR people and CFO's will need to be creative and leverage that creativity to do their jobs to stunning effect. It's something that permeates everything someone does and it helps make the not so glamorous jobs, fun. If you hate unwrapping you might not realize it takes creative skill to do it well. Most creative people are going to love it because it is another outlet to express their creativity. It doesn't matter that no one will appreciate their masterfully inventive way that they unwrapped something, they got to use the one thing that wakes them up every morning. If that isn't you, then I suspect you're attracted to those glamorized bits of the industry for the wrong reasons.
Strip modeling? Bridge, is probably your friend. Personally I would probably start out with cylinders, hack off the ends, wrangle them into the rough shape, extrude some edges, then weld/bridge the pieces together and figure out the topology adding edges, creating faces, cutting things and making it workable at the transitions. If you want to use strips, knock yourself out, it might work out great.
Or are you looking for a detailed tutorial that does it exactly how you think it should be done and can't find one? I think you might be waiting a long time for that... Not too many mind readers around here that can also reconfigure max to work the way you want it to.
BUT you might be able to get a few different takes on the same task if you post the lowest level from zbrush in a common format and then see how people tackle it. It's not that complex, most people could do it on their lunch break.
It informs design choices because you know what will work well and what won't in the next stage of production. It makes your co-workers not want to swear about your work behind your back.
Whoa! I stand corrected. That was lot of helpful advice. Thank you very much!