[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] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / 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: rough.png (662 KB, 2775x2158)
662 KB
662 KB PNG
Hey /3/, any smoothing-group experts here?

I've started experimenting with them to see if I can smooth out from as low-poly an origin as possible, but I'm not sure what my workflow should be going towards the normal bake...

I've not finished the base mesh yet, but anyway... here's the lo-poly....
>smoothing-group experts
nigga, wat?
anyway, just add a chamfer modifier set to "unsmoothed edges", then a turbosmooth on top and you'll have your high-poly
File: smooth.png (822 KB, 2843x2159)
822 KB
822 KB PNG

And here it is smoothed. My question is - can I use this same base-mesh as the LP for the bake or will it need a fresh retop?

Can a base mesh with a fuckload of complex smoothing groups be used in the lo-poly bake?

I'm not planning on doing a shitload to it in Zbrush...

Yeah I know what I'm doing to get the smoothed model out. And the whole point of this exercise was to not use chamfers at all in order to keep the poly count down.
oh fuck, didn't know there would be a second part lol.
Well, I don't see why you wouldn't be able to bake. The way I learned it was to worry about getting a working high-poly, then create smoothing groups on the low poly based on the UV islands you have. Then the normal bake should override your low-poly smoothing groups anyway.
File: rough_slide.png (271 KB, 3714x1748)
271 KB
271 KB PNG

Here's a closer view of the slide without / with the NURMS to give a better idea..
File: smooth_slide.png (356 KB, 3769x1958)
356 KB
356 KB PNG

...With NURMS.

>>>The way I learned it was to worry about getting a working high-poly

Yeah, what my aim was here was to derive as close-to-finished mesh with the low-poly as possible, to avoid need for retop.

>>>then create smoothing groups on the low poly based on the UV islands you have

Ok... why? Not being a dick there in asking that... as I've always, up to now, done absolutely everything in one smoothing group.

>>>Then the normal bake should override your low-poly smoothing groups anyway.

My understanding was, (and this is really why I created the thread), that you can get away with maybe 1 or 2 different smoothing groups in the lo-poly, if they're in inconspicuous areas - but having loads and loads of different groups meeting each other (which this has) will definitely cause problems.
fuck if i know, that's just what i was told to do, and when I plug in the normal smoothing groups just disappear, so I don't even know if they'd matter in the end
If you create smoothing groups based on UV islands, though, you'll only get 4-5 smoothing groups across the whole mesh, and that shouldn't be a problem.
File: bad_barrel.png (502 KB, 2854x2328)
502 KB
502 KB PNG
Is there any way at all this barrel (green object) is going to be any good or is it just too rough?
File: bad_barrel2.png (450 KB, 2515x2269)
450 KB
450 KB PNG

(Incidentally I was always told to use a minimum of 8 sides in a cylinder to get a good smooth outcome... why? 6 has worked ok here)
I mean, that's fine for a third person model, but if you're making a first person model then a lot of areas need extra geometry. The slide, trigger guard, and grip part of the frame (not the side grips, those are fine). Also, if first person, don't go cheap on polys in the areas that will be the most visible (you care about the back of the gun more than the front). For weapons modeling in current-gen games, I've seen cylinders go anywhere from 12 to 24 faces (and sometimes 32) depending on how prominent/close to the camera they are.

You can use 6, or even 4 (might be better for your topology than 8 in some cases).
Just FYI, OP.
One unsmoothed vertex counts as two or more vertices depending how many hard edges meet on that spot, performance wise, since one vertex van only store one normal direction. Same with the "same" vertex on two UV islands. They are two vertices for all intents and purposes even though they don't get displayed as such in your content creation program

So if you have one cube with hard edges it will have 24 vertices despite showing up as 8 in your Max or Maya, while a cube with chamfered edges will also have 24 vertices. The downside to it will be slightly more complicated UVs but the upside is a much nicer silhouette. And shading if you use face weighted normals.

Regarding your question about the cylinders. The issue is the less sides you have the more baking artifacts will appear since the bake will try to wrap the curve around your LP cylinder which is actually just a prism and the distance between the two models will be too large and look bad.
File: YgC0V.png (117 KB, 1239x635)
117 KB
117 KB PNG
Alright OP, I'm gonna answer some of your questions.

>can I use this same base-mesh as the LP for the bake or will it need a fresh retop?

Yes, you could use this mesh to bake your normals on. However, I seriously advise you to add some extra geometry to your low poly. Poly count is absolutely not as important in video games as it used to be, and your end result will end up looking a lot nicer.
Regarding the smoothing groups; the amount of smoothing groups really doesn't matter. I don't know where you got this idea, but the only thing you need to worry about is that everything is smoothed nicely where it would be in real life.

>why create smoothing groups on the low poly based on the UV islands you have?

Do this the opposite way.
Set up your smoothing groups on your low poly so that it looks nice, then unwrap your mesh and keep in mind that everywhere there is a seam in your smoothing groups, there should be a seam in your unwrap. In other words, every smoothing group has to have it's own UV island.
The reason for this is that otherwise your normal map will have horrible artifacts on all your hard edges, pic related explains this further.

>Is there any way at all this barrel (green object) is going to be any good or is it just too rough?

Again, I strongly advise you to add more geometry to that low poly. It'll make the barrel look much more round, and the normal map is gonna work a lot better in general.
You actually solved a problem I have, thank you.
No probs man, the normal baking workflow is not easy to learn on your own.
Goodness knows I've had my fair share of struggling with them.
Is there a recommended tutorial or way of learning how to make normal maps, how to bake them and so on? This part of 3D always eluded me.
Racer445's tutorials are always a pretty good start.

Another channel which I personally really like is Arrimus 3d, he has a lot of great tutorials and still posts videos almost weekly.
He's got a few vids which address some of the most common problems people have with baking, like this:
alright I appreciate the help, I need to check these out ASAP cause I have no clue where to even start on this stuff.
Ah well if you want a full view of the entire process of making a game asset from start to finish I think this would be better:
Ok Ill check this out too, it's all good info anyway

Not that other guy but yeah I can agree that it can be a pain especially with so many conflicting methods that everyone argues.

>smoothing group has to have it's own UV island.
I wish I knew this sooner since I was always told to have EVERYTHING on the lowpoly as 1 smoothing group so of course many of my props looked like ass around the edges or just didnt shade right.

Also i think TEXTOOLS 3dsmax plugin does an unwrap by smoothing groups quick button.

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.