[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Settings Home
/3/ - 3DCG

Thread archived.
You cannot reply anymore.

File: A01_pillar01_stage4.jpg (915 KB, 3000x1158)
915 KB
915 KB JPG
What's a good level for dynamesh so that I can do some good detail to bake onto a low poly mesh?

I know it depends but let's say I want to have a really good clean model to sclupt some great detail down onto my bakes for a low poly pillar. Would you guys go above 800 dynamesh? Currently I've been Dynameshing at 800 then I CTRL+D to add one extra subdivision level. Is this overkill?
You set your resolution to whatever you need bro. And your high poly can be as high as you want or as high as your PC can handle. I try to keep my high poly below 1m on export from zbrush, but I often go 2m++ when sculpting
Actually Dynamesh's dencity depends how on how big your model is.
For example, If you enlarge your mesh 5times that 800 will be very high resolution.

Btw, when I work with dynamesh and have very good form, which I know that I won't redynamesh, i'm dividing and deleting lower subdiv, to be able to use IMM and curve brushes, but if you dynamesh again, it'll go to resolution you have dialed
While there's nothing wrong with your approach - you certainly can get perfectly good results with an all-dynamesh workflow - I find a hybrid workflow is best:

I use dynamesh for the initial form-finding and perhaps the beginnings of detailing, but it ends there. DM is amazing for iterating and allowing you to make huge, massive changes to your model but when it comes to super-high resolution detailing its not quite flexible enough.
So, after finishing your base and making totally sure you're happy with the overall shape and form;
1. Turn off DM - you're not gonna use from here on out.
2. Duplicate the relevant subtool.
3. Zremesh this down to a low-poly level.
4. Add a subdivision and in the subtool palette hit 'project all'. Do this several times till you're at the kind of super high poly count that will allow fine detailing.
5. Delete original dynamesh subtool
6. Proceed to manually apply fine sculpting detail in top level subdivision. Export this as your bake source.
7. Switch down to lowest subdiv level, export this to be your bake target.
You shouldn't sculpt a pillar to begin with, putting so much work, effort and detail to a thing that's presumably only a small part of the scene is a waste of time.

Just make a box, stretch it to be vertical, maybe add some bevel and draw a depth map on it with photoshop, you'll get normals from the depth map.
This is the kind of shitty busywork/slave labor I want to see 100% automated by AI routines. No one should waste their precious lifetime with this.
Can you really get the result like in OP's pic by using that method?? I never heard of it, please post some source if you know of any. I want to try it out. I was pretty sure people sculpt it, no matter if it is an hero asset or some shit mostly hidden from the plain sight. I mean, look high detailed it is, look at all those cracks and metal frames inside of it, no way can you just paint that shit to some map?
You mean the pillar? I don't understand why I'm seeing comments like these for any organic sculpt posted. Why should it be automated? I see more art in this than in extruding some hard-surface sci-fi armor. Don't get it why are so many people here looking down upon organic modeling/sculpting ("reee just use photogrammetry/automatize it completely")...
>Don't get it why are so many people here looking down upon organic modeling/sculpting

Because most people here are talentless hacks who are too lazy to learn how to use zbrush/sculpt. They shit on organic modeling to feel better about how bad artists they are.
>I'm an autistic fuck that spends 100 hours on a single background prop and I constantly whine why nobody appreciates my work

Nobody cares, anon. We all get our paychecks, one way or another. You are not special or irreplaceable. Nobody is in this industry.
What are you talking about? The opinions on this board are a complete opposite of what you can read and hear from professionals on the internet, I wonder who even ARE you guys. How exactly would you create a ruined pillar like that? How would you create some ancient ruins or damage to objects without sculpting? Everywhere you look, every pro tutorial, people will sculpt something like this, there is no other way around it if you want to get a GOOD result. No, photogrammetry is not always the answer. It is an option, sure, but not for everything.

This is getting really ridiculous, I don't even know what are we arguing about anymore. I've seen people post rocks, trees, any other kind of sculpted object, (even characters!), and somebody would always respond something like "just scan it dude", "just use box modeling, sculpting all those details is a waste of time, just use textures lmao". Unbelievable, seriously unbelievable.
I challenge you guys to recreate this pillar without any sculpting, at least a part of it. If you have some time and care enough, of course. Maybe you'll change some opinions here. Hell, I'd sure appreciate learning about some faster method of achieving this if possible.

Use your depth maps and what not and get the result of the same quality as in OP image. It should be fast as you said so, and I'd really like to see the result.
The entire point was that you're wasting your time for spending a week just to make a pillar, you really don't need that amount of attention to detail for a pillar.

If the "people in the industry" were this compulsive about detail, nothing would ever get finished, they would maybe put 20 minutes into modelling that and be done with it, sculpting or no sculpting. The rest of the detail would be done with clever use of textures, it wouldn't obviously look as good as if you put weeks of work into it, but in the end it doesn't matter because no one will care or probably even notice.

Making assets is all about efficient workflow, you have to work fast if you want to get something done or work "in the industry".
How come have I seen pillars of this quality in games, though? Why did I see art tests which ask you to create things like this? And "a week" is an overestimate, it won't take that long, come on. There's a shitton of different brushed which make this work a lot faster and easier. I bet you haven't even tried making something like this.

>you really don't need that amount of attention to detail for a pillar.
Well, you can't know what significance will that asset have in the scene, so it really depends.

But I repeat, recreate this with box modeling and treat it like it will be some big piece of a ruin, right beside the main character in a cinematic, so in focus and fully visible. Show what you can do in 20 minutes.
File: proceduralpiller.png (209 KB, 1043x695)
209 KB
209 KB PNG
I did do a test years ago trying to use procedural processes to make a simple concrete pillar funnily enough for a background prop but never needed it in the end. Obviously no where near the level off the hand sculpted one but then it's only made from a box and Voronoi maps. Looking back I think I should have separated the plaster and concrete elements so they don't blend together in some blobby mess. Procedural maps are fast at adding lots of detail but a nightmare to control and get the results you want.
Oh yeah right, I also played with procedural maps in the past. Yeah, you can get some okayish results with them for simple things, but it's pretty painful to control them and ultimately the results were almost never good enough for me. Well, honestly, with damage brushes I'm pretty sure you could get this result even faster than by setting up those Voronoi maps.

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.