Reusing your own base mesh
polycounter lvl 11
So basically I have a head and a body that I worked a few weeks on until they were good enough to use in my characters. Now im wondering, would it be more wise to re-sculpt the entire anatomy on the base mesh again per character, or just use what I already have but modify it depending on what I want? I mean if I was to make a spartan and a buff guy in a T-shirt, id take the same basemesh and firm/thin it out a bit for the spartan or buff it up a little more for the t-shirt guy.
Same goes for the head, should I be remodeling this stuff? The head was problem a 1-2 project at the most because I wanted to have my own base mesh I could use all the time.
Im just curious, how does the industry deal with this?
Same goes for the head, should I be remodeling this stuff? The head was problem a 1-2 project at the most because I wanted to have my own base mesh I could use all the time.
Im just curious, how does the industry deal with this?
Replies
1. In industry, you often do whatever is the most cost/time effective method so reusing things is a given.
2. In personal development, every time you redo something from scratch you will improve exponentially as you don't skip any learning exercises that happen when taking shortcuts.
Take that information and do whatever you want.
mask everything but the lower jaw, rotate.
Then under the geometry tab click the Edge Loop button a couple times to add loops around the border of the hidden edges, this should add the geo you need to the selected area of the mesh and then you can tweak the mouth shape more effectively at different levels.,
-also do what Zac says to actually open the mouth.
@Dreamer, I try not to take shortcuts because ill feel guilty, but luckily ZBrush is easy enough to jump back into and fix up proportions or anatomy problems for the next project!
can you wip up a half way decent head in polys in an hour or so? if you can then using the same base mesh with some minor edits is a bit quicker then starting from scratch every time.
if you can't I would say model your characters out in polys from scratch untill you can get what you need in a few hours and you know the tools and workflow well. fast efficient poly modeling is still extremely important for any modeler. its still the skill that determines how fast you are at getting work done.
well said
The base mesh's for both were created in max through poly modeling, I just used zbrush for easier shaping.
If this is your first basemesh that you're finally happy with the proportions of, model it again each time and you'll find soon you're able to get better proportions and faster.
I don't think you need to poly model at the level people were at before sculpting was available. back then if you wanted a wrinkle you had to model in that wrinkle with loops etc. it was painful. I think what people are getting at is that at the end of the day you hand off a low res poly mesh that needs to work for animation. so you still need good poly skills. any chance you get to develop those skills you should exploit to give you an edge. imo guys who can sculpt and have good poly skills have an edge.
another thing is that the more you experiment with you base sculpting mesh you see what work and what does not when you subdiv and sculpt. so even if your happy with your base mesh there might be another layout that sculpts better at a lower resolution then one your useing now.
I have about 3 base meshes that I use for most stuff. one biped that is very low res and generic that I can edit into any creature, one that is higher res and more detailed specifically designed for sculpting humans. and a few others that I would use for specific models.
You CAN do one yourself, and if you feel like it go right ahead. Nothing wrong with it. But there's also nothing wrong with using base meshes. Maybe cause I get topology, and low poly work for animation, and all that, but the only times I worry about making a base from scratch is when it's anatomy is not humanoid. Then, zsphere's kick ass to start it, and finish it in max. But you can still use base mesh hands, and heads to kit bash on that mesh if you want too.
Point is, don't limit yourself with too many "rules". Find the way that works for you on your personal work, and use that. I break "rules" everytime I start a project, but that's only because I found a way that works for me. Specifically, I jump to about 3-4 subdivs right off the bat. Most people say that's wrong, but Rick Baker does it, and several other high profile artists do too, so who am I to say they're wrong?
- In a dev situation, you are almost always going to be working with game meshes that have to be fairly decided upon before you finish all sculpting detail.
What I mean by that is, you usually can't just sculpt like crazy and retopo after because otherwise you won't have animation/rigging taking place while your working on something. They would simply just have to 'wait' for you to finish which costs time and money. It also loses you the small window of 'contingency time' where a problem occurs and you realise its gonna need a fuckload of fixing that you didn't have previously scheduled.
SO in that respect. I strongly recommend learning how to make a correct mesh properly (by doing it over and over) and before sculpting, instead of depending on retopo tools.
~ Another point is that there are few people who may or may not be in the industry saying "It doesn't matter if you use some other mesh just get to sculpting".
This is bad practice for so many reasons. The studio I work for has an inhouse art test that requires all character artists to model & texture a full character in 2 work days. Photoshop + Max/Maya only. This is because there are so many artists out there that have portfolios that hide the fact that they don't understand fundamentals. DO NOT BECOME ONE OF THOSE ARTISTS.
It will just lead to you being someone who possibly never gets a job without ever knowing why, or being the person on the dev team that slows things down because you're so used to taking a fuckton of shortcuts that simply could never happen during production.
If there's a situation where your asked to make 200tri head, or a 800tri head, or 4000tris, or all three you need to know how to do that shit. Be a professional. Imagine your idea of a professional character artist and do fill those tickboxes. There's so many incredible zbrush artists out there, you can find a million of those we see them all the time. Only a small handful are desirable on an actual dev team; because only a small amount are champion modelers/texturers. Problem solving and artistic decision making will be the bulk of your work as a professional artist.
Train that shit up while you have the chance. Be a champion in all aspects. Because believe me when I say if your just 'zbrush awesome' then all companies already have truckload of freelancers the US to China that could do what you do at a faster pace, and for a cheaper price. Which makes hiring you pointless.
I am glad to have read this thread.
I kept trying to scan your QR with no avail and then realised it was space invaders. lol.
On topic, reusing base mesh in my opinion is definitely not frowned on, for efficiency's sake. Having said that, I think it's still very important to practice your fundemental modelling skills because not everything requires a base mesh that needs to be sculpted and everything.
You've made a good point, but I've never run into any situation remotely close to that, maybe I'm lucky? But it's never been a problem for me. Like I said, though, I understand the basics, how to model for animation, good topology, all that kinda thing. I spent several years learning basic poly modeling in max, so I don't have problems with understanding how things work.
For the purposes of having a base mesh to sculpt from, it's not a problem to reuse base meshes. Ive never had a single employer who's had a problem with me using a base mesh, especially when it helps get the job done faster. You need to work the shit out of it, and make sure you're becoming better, and not just making the same mistakes over and over again, though. It's just a starting point, that's it.
This varies from person to person, but for me, I usually end up with a ready to work model in under 8 hours with above method, since it allows to me tackle each part of the workflow systematically and allows me to take breaks between parts as opposed to taking breaks during a certain phase. Tackling the mesh flow and proportions, as well what is a hard surfrace, which is soft and which parts of my model will deform and not really takes a toll on me when I'm tackling them all at the same time, especially for non-human creatures.
w/ Zbrush its WAY to easy to go off concept and out of proportion.
I am curious to see exactly what quality is expected of artists with just 2 days to model and texture a character model in this test. I cant imagine a piece of art worth mentioning coming from such a rush job. Is this using base meshes or prefab bits and they are asked to ghoul things together?
As for character art pipelines we can have entire thread discussions about it.
I've seen and worked on dozens of variations. Most of the time its model whatever the heck you want as long as it fits on the skeleton.
Sure you build efficiencies like reusing base models or premade body-types for the project.
If its using an already made base model that has a specific mesh topology or modular system then its a little more tricky but even those allow the artist to take the pre-made mesh and conform it to their new creations and project the details.
My main beef with artists these days is that most cant model a clean optimized model to save their life. So the 200 800 4000 example is very true. Have fun creating but good lord know how to make a in-game friendly version of your model.
But yea very curious to see what would be expected as a final model for that 2 day test. I cant envision anything pretty...
And for the record; I never said I was against basmeshes entirely. I use them too. I just said, you shouldn't be using them if your in school trying to learn.
Can we not all agree that you shouldn't be using basemeshes until you can make one upon request?
Skiffy is confused now...
But yes build your own base meshes! heck build up a library of starter bits if it floats your boat. But if you just download a basemesh from online and using it while still studying in art school then your just shooting yourself in the foot. Foundation foundation!
As for 2 days I don't know... The second I heard photo slapping and PS2 crowd characters "pretty" left the discussion. As my one artist friend would say "It works as long as you don't look at it"
~ I honestly wish I could show you the awesome shit some of our artists have made in that time. Game ready and looking pretty.
Personally, I find it hard to work form a 'blank' basemesh (kind of a like a blank white canvas for me..)
- I normally make a customized basemesh for each individual i do to get the juices flowing
I guess it wouldnt much of a test if you saw an example of other people's.
This whole talk of basemeshes kinda makes me realise why so much top industry guys say "don't open zbrush until you understand modeling". It's so tempting to just skip all the boring stuff and get to the fun part
This^
Then you've never seen quality production work.
Painting characters here at KingsIsle, I take about 2 days per character. Low poly, toon style, hand painted characters.
More time does not equal a good character.
I REALLY hate that current generation of artists consider work to only be good if it's hyper detailed with zbrush, 16,000 tris, 2 2048 textures, etc.
I agree that zbrush artists are a dime a dozen and the proof is in the final low poly output, but I don't agree that you should shuffle incomplete pieces to the next guy just to keep him busy. With proper planning and structuring of the pipeline and schedule it won't be an issue for the animators to wait until the model is done.
We wait until the models are fully completed before we start rigging or animating. Our process is well refined and well scheduled to the point that it all fires off without a hitch. On rare occasion there might be a material refresh.
Our modeler also has a crystal clear idea of what the base mesh is going to look like and how the topology is going to flow as he's sculpting, which will changed based on different things that could pop up but for the most part he has it all planned out.
Doing it the other way, where you tie the modelers hands and don't allow changes while they sculpt is very restricting and often leads to those tense moments where a change has to go through, then the animators are stuck re-doing a bunch of work. The time they saved was just wasted and it caused all kinds of frustration, animosity and headaches that didn't need to be there.
Rigging and animation is so much more complex than modeling and sclupting, there is so much more that can go wrong and most animators will agree that they probably don't want to mess with your rough sketchy models and would rather work with something that isn't going to change. All of the animators I've worked with and talked to are like that.
I structure our schedule in such a way that the modeler can do their job unhindered and un-rushed by the animators schedule so they can concentrate on their job instead of running around like a crazy people worried about other people and deadlines. That is one thing our industry continuously gets wrong and I refuse to work like that if I can avoid it. If at all possible management should structure the studio in a way that artists can concentrate on art.
So long post short,
- A modeler needs a clear picture of what the low poly is going to look like, more for him/her so they don't waste their own time doing a bunch of things that don't matter. That comes with practice and experience.
- Passing on unfinished products for anything more than testing is kind of a crazy way to work and often leads to more wasted time than whatever it saved.
~ Agreed; although I never said it was sensible to pass on unfinished things. I just didn't feel it necessary to break down the advantage of having a fairly decided on mesh, that changes incremently so the rigging/animation can have skin weights loaded (by world space if necessary) without having to destroy too much previous work.
So YES. Changing the lowpoly as you go should happen; but an entirely new mesh at the end? That is just nonsense.
To clarify; Basemeshes are not evil. They should just be used by someone who can make one upon request. That's when I feel people are most helpful in dev situations and look hirable. When they clearly know their shit inside and out.
.