[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] [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]
Board
Settings Home
/3/


File: CJdSyUV.jpg (64.98 KB, 530x600)
64.98 KB
64.98 KB JPG
>There are people on this board, RIGHT NOW, still using ancient modeling techniques such as box modeling instead of just sculpting.

What's your excuse?
>>
> sculpting mechanic things
>>
>>498051
This isn't 2008 anymore. Zbrush has plenty of good tools to sculpt hard surface. See >>497890
>>
>>498049
>game spec art
>quads for deformation/animating
> try baking ngons
>try importing ngons into game engine

>>498052
Yes, now zbrush has modeling TOOLs that are NOT sculpting.
OP didn't say using zbrush as the only tool. He specifically mentions sculpting.
I can sculpt in Maya, 3D coat, modo etc. but they can't do what I mentioned above.
>>
File: KILL ME.jpg (570.22 KB, 2104x1292)
570.22 KB
570.22 KB JPG
Please OP, explain how my decision to sculpt this hardsurface object in ZBrush was not a terrible idea. I can only imagine how much easier it would've been if I started in Max.
>>
>>498052
> sculpt it like shit
> box model all over it again with drawing every quad
Thanks for proving my point.
>>
>>498054
>I have never touched Zbrush: the post

Go look up on zremesher, the best retopo tool in the industry.

And no, you can sculpt hard surface, I'm not talking about zmodeler
>>
It's actually very easy to sculpt hard surface, then polish using tools that pretty much retopo the surace automatically

http://zbrushtuts.com/2015/08/18/marco-plouffe-shared-w-a-s-p-tutorial-zbrush-keyshot-photoshop/
>>
>>498058
While there is some very nice work in that thread, it all has that tell-tale blobby/low-res look. That's not clean hard-surface modeling, no matter how much detail you try to hide it under.
>>
File: wasp_keysculpt_01.jpg (406.58 KB, 1440x1800)
406.58 KB
406.58 KB JPG
>>498059
Are we looking at the same model?
>>
>>498060
>that CA

why

it was so close to perfect
>>
>>498057
Does it retopo hard surface?
>>
>>498049
>There are people on this board, RIGHT NOW, still using ancient modeling techniques such as sculpting instead of just using 3d scanners
>>
>>498060
>ca
>dof
>>
File: 00001.webm (2.94 MB, 1920x1080)
2.94 MB
2.94 MB WEBM
>>498073
I'm new to ZBrush but It can retopo your average object pretty nicely, don't know about super complex stuff.
>>
>>498049

>There are people on this board, RIGHT NOW, still making troll threads about whatever technique is the latest thing instead of actually creating something
>>
enjoy your transpose tool
>>
>>498073
It's unreliable unless the retopo is controlled with the help of new features in 4R7 like the masking.
>>
File: 1420427640947.jpg (13.09 KB, 196x178)
13.09 KB
13.09 KB JPG
>>498084
fuck the transpose tool!
>>
>>498079
fucking this, even tho keeping up with tool is kinda important
>>
Isn't it always the preferred style to box model first even for the rough shape?
I can't imagine sculpting an entire person from the ground up.
>>
>>498108
>I can't imagine sculpting an entire person from the ground up.
Yet you can imagine poly-modeling an entire person from the ground up? How can you think that's quicker than just sculpting it?

If you already have a decent base-mesh and are making a purely human character, then yeah, most people are just going to sculpt on a base-mesh. But for everything else that requires some actual creativity, you want the freedom to explore ideas and play with shapes, which sculpting first allows you to do at a much faster rate.
>>
>>498052
you can box model in maya way faster.
when you box model you dont have to retopo shit.
>dont fucking mention zremesher
its fucking shit
>>
>>498112
>explore ideas

Does nobody concept in 2d anymore? So much time wasted on moving polygons.
>>
>>498119
>Does nobody concept in 2d anymore? So much time wasted on moving polygons.
Concepting in 2D only gets you the main ideas. Concepting in 3D is more about playing with the forms that the 2D concepts are vaguely representing. But also dynamesh does allow you to concept certain things much faster, especially considering you can rotate all around the model to see how these ideas work in a real setting.

Also, time is not wasted. I can throw down a bunch of forms within seconds that would take you several minutes to draw. And as I said, I'll be able to look at those forms from any angle, not needing to redraw it. But 2D concepting is still the best way to get the main idea and feel of the character down.

>>498115
>when you box model you dont have to retopo shit.
Because box modeling is the "retopo" phase. You're talking about the same thing, the only difference is you're taking the long route. Before it used to be: Box Model>Sculpt>Bake.
But now it's: Sculpt>Retopo>Bake.

This newer workflow is faster, because doing your box-modeling on top of an already done model takes mere minutes. And you get the added benefit of having complete freedom during the sculpting/detailing phase to alter the design at any point. Get with the times.
>>
>>498124
>>498115

Also, lets not forget that usually you'd have to make two base meshes if you're going the basemesh first route. One that was game-ready, and another that would subdivide well and support sculpting nicely, as the game-ready mesh rarely has topology good enough for that, unless it's for a AAA game like Uncharted.

But we no longer need to do that thanks to dynamesh and sculpting first.
>>
File: greyscale_concepts_final.jpg (1.73 MB, 1308x1440)
1.73 MB
1.73 MB JPG
>>498119
>So much time wasted on moving polygons.

Concepts just like this can be completed in under an hour. Yes, this particular sculpt might look like shit, but that's besides the point.
>>
>>498124
only non retard in this thread

except sometimes you're really good at box modeling since you've been doing to for years and you can skip the sculpting bit entirely or you're making shit that sculpting doesn't lend itself well to like a keyboard, a pair of pliers, a detailed view of some electrical contraption, a motorcycle, a house, a nuclear reactor core... etc. etc. etc.

not everyone makes characters
>>
blockout with simple primitaves in maya, simple shape and form sculpt in zbrush, retopo in maya, detail in zbrush
>>
>>498128
That is retarded, sorry. Why would you do the blockout with polymodeling? Have you never touched zSpheres or Dynamesh?

>>498127
Well yeah, that should go without saying, if you're doing some purely hard-surface stuff, then sculpting is unlikely to be the answer, unless you're trying to concept some interesting hard-surface forms.
>>
>>498049
>still using ancient modeling techniques such as box modeling instead of just sculpting.
You still use box modeling, Hard Surface modeling for example in Maya.

Boxing out your model in Maya or Max. I.E getting dem shapes then using GoZ to export it to ZBrush to begin sculpting.

Creating ultra low poly models.
>>
>>498126
And you can shit out basic concepts like that in 2d in 10 minutes. Then you make 5 more and actually develop the idea while trying new approaches and go with the final one.
>>
So... reading this whole thread got me lot of confusion.
Is box modeling still used? Or should I say fuck everything and praise zBrush like next gen kiddos?
>>
>>498127
>>498127
Zbrush is used for hard surface too. It's become easier and faster than traditional modeling
>>
>>498143
Box modelling is still a good thing to understand.
>>
>>498143
Learn both. Moving veritces around and knowing how edges flow is still important. Sculpting is faster, more fun and lets you forget about technical stuff to focus only in your art. Zbrush is only a sculpting program so you still need other packages
>>
>>498077
At that density, the entire character can only be used for rendering.
If you want something game spec, guess what? You have to retopo again with much less density, aka manually.

>>498108
If you can, it's faster to sculpt the entire person, then retopo. Things like fingers and toes are a bitch though.
The emphasis is that the final sculpt will be retopologized if it's being used in a game or movie, unless it's a "concept sculpt" like >>498126


>>498119
3D artists are kinda know as not being good 2d artists. pushing polies in 3D is much faster, and gives them variations much faster than 2d. They can then choose which one they want to spend time one and finalize with a much better idea of how the final version will look.
-Then it's easier/faster to just keep tweaking the rough start, retopologizing where needed etc.

>>498126
It's unimpressive. Prolly no more than 20mins was spent actually sculpting"inflate brush. The painting is actually more impressive. The last 20mins were prolly spent in PS assembling the presentation.

>>498130
>Why would you do the blockout with polymodeling?
Was zbrush 3 your first zb or something?
Before zhpheres and dnamesh you start with a base mesh to ensure polies were even for best subdivision.

>>498143
Pure ZB will get you so far.
From a technical point of view, box modeling and moving verts, manually removing ngons, figuring out what's wrong will help you be more knowledgeable about what's going on under Zbrush's hood, and any sculpting tool I guess.

>>498136
Doing the "basic concepts" in 3D is MUUUUCH faster. That pic is not a "basic concept" though. That's something more time was spent on to be showed to a studio, supervisor, client, or portfolio piece etc.
>>
>>498149
>If you want something game spec, guess what? You have to retopo again with much less density, aka manually.
Guess what? It still takes less time this way. The artist was able to concept out some of their design in 3D space, and get all their fine details down too. Afterwards, a quick retopo, a bake and you're done. It's a lot easier to figure out how to poly-model those shapes once you already have them sculpted out. And tools like Quad draw in Maya or Topogun make it take no time at all to retopo.

>Things like fingers and toes are a bitch though.
Sculpt one separate, duplicate a bunch, minor adjustments, dynamesh. They're really not difficult once you've done a few! Though tbh, it's useless to keep resculpting the major forms, one should have some pre-made fingers to basically "kitbash" with their organic models and then adjust the shape from there.

>Was zbrush 3 your first zb or something?
I think you misread my comment... I was advocating against using polymodeling to do the initial blockout, because we have dynamesh and zSphere. I literally even mentioned those things.

>>498143
Box modeling, or more specifically "poly modeling", is still important, as you will use the knowledge when doing retopo, or when doing some purely hard-surface work that doesn't need any concepting done on it, like a mug, a real life car and things like that.
>>
>>498052
>make a bait thread
>get defensive after one post.

Who taught you to troll?
>>
>>498144
Are you seriously suggesting that you would model a keyboard in ZBrush rather than doing it in Modo or something?
>>
>>498149
>At that density, the entire character can only be used for rendering.
>If you want something game spec, guess what? You have to retopo again with much less density, aka manually.

There's an option to zRemesh with less polygons. And who here was talking about game-ready meshes?

Anyway, at some point retopo tools will become so smart you won't even need to look at the wireframe anymore.
>>
>>498165
Obviously squarish things like keyboards, doors, etc are still faster to make using traditional poly modeling. But once you get into sci-fi or mechanical looking hard surface things are much easier with ZBrush.
>>
>>498169
>implying computers will kill human eye and experience
>>
>>498172
They've been killing human skills for decades. What makes you think an algorithm can't do a better job at retopoing than a person? In fact, most retopology plug-ins are already somewhat automated, and Zremesher is pretty much a one click solution.
>>
>>498049
Just to clarify some shit. Box modeling, poly extrude techniques are used for precise reproduction from a blueprint. Sculpting is when u don't have that requirement. E.g. you don't sculpt a real tank - u model it using traditional methods cause it's efficient and precise.
>>
>>498049

I'm a lazy fug who doesn't bother digging through the toolbar to find the sculpting even if in the long run it saves time.

Besides, it was good enough for my uncle, so it's good enough for you.
>>
This thread was interesting. Yes I think Zbrush might be the new 2.0 shit.
>>
File: 1443737795566.gif (1.58 MB, 400x300)
1.58 MB
1.58 MB GIF
>be a rigger/animator
>mfw the retarded modeler hands me a model he retopo'd in ten minutes after fucking around in zbrush for twenty hours
>mfw there's pinching and deformation fucking everywhere
>>
>>499365
which is why i cringe when people on here say that zremesher is good enough.

>im a modeler
but i retopo everything from scratch, because you know... im not lazy as fuck and i want it done right
>>
>>499366

Wait, someone genuinely just zremeshes and calls it a day? That's fucking retarded.
>>
>>499390
how is it retarded?
>>
>>499395

I'll let you figure that one out on your own when something goes wrong a bit further down the pipeline.
>>
>>499395
the artist doing rigging/skinning will murder you
>>
>>499395
this is why when you send your reel to places, if they are smart, they will not hire you. because people can spot zremesher geometry. what are you going to do when you arnt just making a statue that dosnt ever move
>>
>>499410
not really, just don't be lazy about it
>>
>>499413
using zremesher IS being lazy
>>
>>499415
not if you know what ur doing
>>
>>499416
when you know what you are doing you dont use zremesher, thats what we are trying to say.. jesus dude.

people get fired at studios for using zremesher. if you ever want to actually get serious. do the topology yourself. its like if a rigger used an auto rig feature in maya. they do it from the ground up when they want it done right.
>>
>>499421
whats the problem with it?
the only problem i see with using it is straight up pressing the button without adjusting anything
>>
>>499422
by the time you go an fix all of the issues that zremesher has. you could have just made it from scratch, and more often that not it will be faster.

every Once in a While you will get a piece out of zremesher that is ok
>but why not have more controll over where you want more detail/less detail. where you want the 5 and 3 points to be.

these things are important and change from project to project. Just listen to those of us on /3/ that acutally know what we are talking about when it comes to important shit like this.

we arnt trying to be assholes, but we are just letting you know that if you need to rig/skin/animate anything

>DONT FUCKING USE IT

nobody cares what you do for your personal shit or even if you leave it in dynamesh for your own shit. but dont ever give somebody that knows what they are doing a zremeshed object. they will never hire you to do anything again.

cant people just give up their predetermined notion that zbrush is the best program of all time and it can do no wrong
>>
>>499423
m8 you just wrote an entire paragraph without stating the problem

are you literally autistic
>>
>>499421
>people get fired for using zremesher
Any citations ? I'm not even a zbrush user but what you're saying sounds like a load of bullshit
>>
>>499423
>Just listen to those of us on /3/ that acutally know what we are talking about when it comes to important shit like this.

Oh. You're one of those. Show us your credentials and portfolio before talking like you know anything.
>>
>>499425
>spirals,Spirals, SPIRALS
5 points in the wrong spot
3 points in the wrong spot
to much geo in areas
not enough geo in areas
issues in flow when holding areas like the mouth, eyes, nose, ears, knee, elbow, finger joints, pecks, ect. sometimes it just dosnt know they are there and makes a grid.
overall issues with the flow of the topology.
every once in a while when you have symmetry it will make different topology from side to side.

good enough m8080?

when you know how to do topology correctly you will cringe at zremeshed topology
>>
>>499429
>every once in a while when you have symmetry it will make different topology from side to side.
i countered this problem,its not big

ok i agree it might create problem in facerigs but you are exaggerating

also you forgot when zrememsher sometimes leaves fucking holes when its too low poly
>>
File: 4g69mg.jpg (42.99 KB, 640x480)
42.99 KB
42.99 KB JPG
non zremshed topology reference for the zbrush kitties
>>
>>499431
i can achieve this with zremesh, it might take a little longer than id like
>>
File: im so dissapointed.jpg (163.30 KB, 1062x594)
163.30 KB
163.30 KB JPG
>>499432
you still dont have full control over what happens with the topology unless you spend more time configuring it before you hit the button than just doing it yourself.

every 1/1000 times i see decent topoly out of zbrush, but never great.
>>
>>499433
and even if you spend a ridiculous amount of time configuring and drawing out splines for zremesher, it will still turn out not great.
>>
File: 1439456686078.jpg (27.63 KB, 464x464)
27.63 KB
27.63 KB JPG
>>499410
> because people can spot zremesher geometry.
>>
File: simple example.png (627.41 KB, 1130x982)
627.41 KB
627.41 KB PNG
>>499437
heres an example of the type of n work ive seen people give if they are doing the topology themselves or using zremesher.
>>
So far nobody has posted any good Zremesher results, that's the only thing that would end this discussion.
>>
zremesher is good, and produces usable topology since 4r7p3, but usable is still not great. You won't see it used for retopologizing AAA movie models unless it gets made much better, for instance. It's fine for indie games and stuff, even their cinematics, and is one of the reasons the topology on Genesis 1 (haven't seen 2 and 3) looks like it does.

At least you can avoid spirals with it now.
>>
File: Mio_yeah_ok2_ep3.png (261.35 KB, 758x669)
261.35 KB
261.35 KB PNG
>>499438
>>499435
>>499433
>>499432
>>499431
>>499430
>>499429
>>499428
>>499427
>>499425
>>499423

Both sides of this argument are fucking imbeciles. ZRemesher was NEVER meant to be used for final topologies. ZRemesher is a just a smarter Remesher, that is all, you use it to lower your polycount while still mantaining a good level of detail in the model, and at the same time to make your topology flow decently, but you still need to do retopology using conventional methods once you're finished.

The ones shitting on ZRemesher here should've use this argument first. This thread only shows how little you guys know of Zbrush workflows in general
>>
>>499442
or ive been told by my boss, my co workers and everybody in my old school to not use it and to do final topology from scratch.
>>
>>499442
weve been talking about final topology this whole time.
as said before, nobody cares what you do for your personal shit.
>>
>>499442
this, read the press release, read the documentation, this is exactly what zremesher was designed for, some people can push it pretty far and get some pretty decent topology, but it was designed to be a step in a process, not the end result.
>>
>>499443
That's what I said you idiot.

>>499444
No, you've been talking about using ZRemesher for final topology and why it's not very good, well duh if you had any idea of how it's used you wouldn't even be arguing about this shit.
>>
>>499446
I was making that point origionating from the rigger saying that people were handing him basicly just a zremeshed object.

Too often people arnt creating good topo off of the zremesh because they think its close enough. Which is why i was going a little further and saying people need to learn how to do it from scratch before they even try to fix zremeshed objects. And the fact that it often takes longer to try to use the zremesh with your fixed topology and make it one cohesive thing. It seems like a waste of time to use zremesher as a topology starting point that you fix. I personally start from sratch because im still learning and would probably fall into the same hole other do of just leaving some topology that isnt perfect.
>>
>>499438
lol wut?

zremesher can achieve way better than this if ur not lazy
>>
>>499453
see
>>499415
>>
>>499454
what if i told you i can achieve something very close to the manual one lmao
>>
>>499456
is this bait
>>
File: 4r6-020.jpg (133.45 KB, 630x520)
133.45 KB
133.45 KB JPG
>>499459
>>
>>499461
your just proving my point. thanks
>>
>>499462
Proving what point? You don't think that topology is good enough for animation?
Please, point out the problem areas, since you clearly seem to think you're an expert.
>>
>>499465
>good enough
proving my point
>>
File: tumblr bumblr garbage.jpg (222.43 KB, 1280x777)
222.43 KB
222.43 KB JPG
Since there is at least two people arguing in here, I want to break free from box modeling and try to learn how to use some 3D sculpting tools. I want to start easy so my first project would be something akin to the included image. Would it make sense to use sculpting to make a cartoony character with little detail, or should I just stick to box modeling for that sort of stuff?
>>
>>499461

It's alright. Not ideal, but alright. Might be alright for low end games where you have limited rigs/facial animation.
>>
>>499467

That sounds like a terrible place to start, honestly. Good luck finding a good 3d shape for her head that works from every angle without custom shaders.
>>
>>499469

Well it wouldn't be exactly that model, but just something with low amounts of detail while still have a human shape and face.
>>
>>499467
i would do the basic shape you want in cubes, then bring into zbrush. use dynamesh until you get what you want
then retopo the whole thing.
>>
>>499472
but i agree with
>>499469
cartoonie things are harder to do when you are starting off.( not completely sure if you were trying to say that )
>>
>>499474

I've done a lot of cartoon characters with box modeling, but this would be my first attempt at sculpting. I'm constantly being told I should be using something like zbrush when making characters as in the end it would ultimately be easier, but I've never made the jump.
>>
>>499476
try it out every once in a while, see how you like it
>>
>>499461
yeah, it's nice to knock the resolution down.. but that topology is 'wrong' , the nose, brow, side of the eye, if you look closely at the corners of the mouth, and the cheeks, ears are good tho lol. faces are pretty easy to retop , hands are a pain in the ass, but after awhile it becomes second nature, just like uvs. i think zremesher is pretty nice for a lot of hard surfaces, especially if you have a base mesh with a lot of control loops,clone it , zremesh it, crank the res up do your high frequency shit and bake, can be annoying as fuck making a sub-d mesh sculptable
>>
>>499465
i missed your post i'm not the guy you arer bitching at , but you can see my post below, the topology is not good enough, it looks good if you don't know what you are looking for, but it is not where it should be for rigging and weighting.
>>
File: derp.jpg (249.01 KB, 630x520)
249.01 KB
249.01 KB JPG
>>499465
>>
>>499492
i added some to it
>>
File: 1444850618584.jpg (331.37 KB, 1264x1556)
331.37 KB
331.37 KB JPG
>>499503
see this guy gets it. the brows really bothered me, that straight row across, it deforms so poorly.. and the single quad of the inner eye, i missed it until i took it in photoshop and was like what the fuck.. i really don't know how anyone could think that topology is even 'good enough' , because it's pretty shit. i almost never use zremesher, i started using zbrush long before even dynamesh, so learning topology was pretty important , and depending on the application, mine being topogun(maya live surfaces are getting pretty good but i'm a topogun fanboi at heart) topology doesn't take very long, a few hours at most if you are super anal about it, which i am so i like the extra control, my biggest gripe with zremesher is the weird poles it creates, point in case from the thread that started all of this, it doesn't matter, but everytime i see poles like that in my mesh idk it makes my autism flare up
>>
>>499492
>>499503
>>499510
>>499461
i added some more
>>
>>499461
Looks good enough, still some tweaks here and there needed for 'optimal' topology but this could be used for testing purposes or as a starting point.
>>
>>499519
>HERE AND THERE
see
>>499512

it all has to be changed. everything.
>>
File: area.autodesk.com.jpg (114.09 KB, 720x527)
114.09 KB
114.09 KB JPG
>>499520
Nope, this is the main character model for a AAA PS3 game, it has the same "errors" than that screenshot. You're just being nitpicky to prove a point. Sure if you want to be anal about it you could improve the topology, but that topology the guy posted won't get him fired like someone said earlier in the thread. You're making it seem like it's the end of the world.
>>
>>499521
What errors are there ? its fuckin perfect
>>
>>499522
>>499521
>>499461
>>499519

samefriend
>>
File: asdsadsae.png (4.79 KB, 282x191)
4.79 KB
4.79 KB PNG
>>499522
Weird, It has the same topological errors you are bitching about in the nose and eyebrows.
>>
>>499524
>>499523
I'm another guy. There are moer than 2 people in /3/.
But tell me what errors does that model have ?
>>
>>499521
it doesn't have the same topology, drake has really nice topology.. point in case the mouth , the loops that go around the nose and mouth, the loops that go around the eyes properly , there are 3 visible poles that are placed in the best spots. like anon said, it's perfect. http://www.amazon.ca/Stop-Staring-Facial-Modeling-Animation/dp/0470609907 good book to help clear up your confusion and set you on the right path. this art isn't subjective it is commercial, there are practical standards that people adhere to for a reason.
>>
File: area.autodesk.com.jpg (48.83 KB, 720x527)
48.83 KB
48.83 KB JPG
>>499525
>>499526
Just pointing out the same "errors" from the other model
>>
>>499524
it doesn't have errors in the nose or the eye brows.. the noses are completely different
>>
>>499527
i fear you are missing the sublty of what is good/bad topology
>>
File: topology_breakdown.jpg (152.03 KB, 800x548)
152.03 KB
152.03 KB JPG
>>499527
one thing to consider with drake is that he is a pretty high poly asset , but all the important loops are there, point in case. i know they look similar but they are different, and they deform differently.
>>
File: asdasdsade.png (823.55 KB, 649x722)
823.55 KB
823.55 KB PNG
>>499529
Don't tell me tell the expert who drew over the first image, I'm just following the same standards.

>>499522
>>499526
Yeah, very nice and perfect topology, so much that they improved it in Uncharted 4, actually addressing those "errors" you pointed out, which could only do because they worked with a higher polycount.

Come on now stop contradicting yourself.
>>
>>499524
comparing good topology to a game model where they need to save as much detail as they can, because the game wont look realistic enough anyway, and do as much in texture. which dosnt look as good as when something is modeled in. if given the option, they would have modeled the features they are talking about.

i guess we should all do hair in cards and not with fur

I guess we should all not model in ear features

https://vimeo.com/31781249
>>
>>499532
Moving the goalposts already I see.
>>
>>499533
no, im saying that in a video game they cant model things in that they want to. they know its not good, but they cant make it better becuase they have a poly budget set by their boss. they want to use detail maybe on the clothes, or other character assets. in no way am i moving the goalpost. the standard is still there, older games were just unable to reach it

>if given the option, they would have modeled the features they are talking about.
>>
>>499533
i think the goalpost was already moved going from uncharted 3 to 4.. there is a drastic increase in polycount, that being said the basic topology remains, even if it is kind of hard to see in that density
>>
File: goalposts.jpg (26.10 KB, 300x240)
26.10 KB
26.10 KB JPG
>>499534
>That topology is shit ohmygod so full of errors this is unacceptable in professional environments!
>Here have a picture of a AAA game model with some of those errors
>Yeah they had to do that because they have a polycount budget

>>499535
The flow in the eyebrows, nostrils and cheekbones is completely different
>>
>>499536
they are not the same errors tho are they? the errors you are saying the same are poles. lets get old school http://forums.cgsociety.org/archive/index.php?t-347843.html

A pole is when 5 or more edges converge at a single vertex.(which is impossible in a 100% quad mesh)Yes they are bad when simulated by a smoothing algorithm as they create pinching and other nasty artifacts.They can be hidden or placed somewhere where the topology can accomadate them without noticable artifacts.If this is the case then they are deemed acceptable.

the topology is different, i understand that it looks similar , but it is in fact different. the only similarities that the meshes share is that they are both male heads, poles are inevitable and they just need to be placed in places that are least troublesome, which they have been in drakes case. the nostrils are not comparable, drakes nose has the topology to support the shape, the zremeshed head doesn't, at all.
>>
File: lara_tr1_wireframe.jpg (47.81 KB, 423x479)
47.81 KB
47.81 KB JPG
>>499533
>>499535
99% of the issues i had with the zremeshed model and the uncharted 3 model have been addressed in uncharted 4.


>>499536
>yes, the polygon budget. do you know how games work.
i mean if your boss dosnt care if you zremesh and your rigger dosnt care. fuck it, do whatever. but i know people like me and that have a greater understanding will look down upon people that make those mistakes when they have the option not to.

if they were able to make this model better, they would have done so
>>
>>499538
heres the pic of the things i see that got fixed
>>
>>499538
i find the whole thing kind of a moot point anyway, considering naughty dog hand paints all their textures, seems like they probably wouldn't automatically generate their topology.
>>
>>499540
we are trying to get the people to understand that you do topology from scratch if you are in a professional work environment and not just

>hurdur i zremeshed and its close enough for animation so fuck it, ill fix the spirals and im done.

also
>MOOT
COME BACK TO US!!!
>>
>>499541
lol, i agree completely, to by honest i think that fact that a thread has lasted this long without 'BLENDER IS BEST BECAUSE BLARRRRRGGG' nonsense is a massive step in the right direction. I learned modeling , and then I learned animation and rigging, and then i went back and learned modeling.. because fuck me topology makes such an incredible difference. most people just want to make epic models, they don't understand what is going on under the hood, they only see the end result and hate that they can never achieve it.
>>
>>499492
>>499503
>>499512
sorry to tell you but you don't know anything about zremesher,good job sperging tho
>>
>>499548
>sperging
got us

we are pointing things out wrong it

see
>>499539
>>499467
>>499490
>>499447
>>499442
>>499433
>>499431
>>499429
>>499423
before you go around trying to act lik you know anything at all
>>
>>499550
its not me
>>
>>499538
Still moving goalposts. Polycount budget was not even an argument of yours before the picture of uncharted model was shown committing some of the same "mistakes" as the zremesher model.
>>
>>499553
they said its perfect tho
>>
>>499553

let me break this down for you with an analogy

when you ask for soup at a restaurant, you expect a certain standard. set by other restaurants of the same quality, and price ( the level of the restaurant ). now let say they bring you soup from a can, sure they can add spice, salt, and other products to make it 'good enough'
>BUT
would you not want a soup made from scratch that will, in the end, taste better. would you, if you were a restaurant owner, not want the chef that knows how to make the better product.

now when you are ordering the soup. do you ask what type of kitchen equipment they are using and how old their technique is? no, because you expect a bowl of soup made by a professional chef, at a respectable restaurant today to have standards set by today.
>not yesterday
>not last year
>not in the last decade.

does this preschool analogy help you understand.


the reason it wasnt an issue that was brought up until that photo was posted, is because nobody fucking does that shit anymore at a high level. why am i holding people on /3/ to high standards? because fuck, maybe i want to have people that know what they are doing.
>>
>>499558
are we debating if something has to be perfect or good enough?
its the same thing with food
>>
>>499553
the polycount became an issue with two different models with different polycounts came in to the mix, drake from uncharted has proper topology, if you can't see it you can't see it, but others than do see it see it, there are books and tutorials out there than can help you see the difference. the only issue drake has is poles, verts where 5 edges converge , most models have them, and they are not a super huge deal but they cause problems. all of that aside , the zremesher example had bad edge flow, which is the real issue, but again it you can't see the difference thats fine, but there is a difference.
>>
>>499560
let me try to understand

the spherical loops around the mouth and the eyes assist in better mesh deformation around the cheeks and the eyes
is that right
if everything is quaded equally you cannot get a proper deformation
>>
>>499562
oh you definitely can get fine deformation if the polycount is high enough, but for example when it comes to rigging and weighting, most rigging is done on a down-res'ed mesh, and edgeflow can help in making weighting a lot easier. generally speaking art imitates life, and character art is no different, proper edgeflow and loops mimic the muscles of the face
>>
>>499562
correct, its like crumple zones on a car.
basicly you are making weakpoints while keeping some paths ( referd to as flow ) tight, which keeps detail while the object is deforming
>>
>>498049
I'm new as shit and everything is fucking confusing
>>
My computer can't even load a material on the screen for zbrush. Box modeling FTW!!
>>
>>498049
>What's your excuse?

I'm really bad at sculpting

I'm sorry 3:
>>
>>498064
newb here, what's CA?
>>
>>500683
Chromatic Aberration.
>>
I must have access to individual vertices.
>>
>>501361
wut
>>
>>501402
you can't grab a single vert and translate it numerically it in a sculptor software
>>
You can get faster results for certain objects, like hard surface stuff, using standard techniques. Why "sculpt" a chair, then have to retopo it, when I could fuckin model one faster from scratch?
>>
>>498049

Most shit dont need that lod and make box modelling is faster
>>
>>499421
>its like if a rigger used an auto rig feature in maya. they do it from the ground up when they want it done right.

Actually Pixar starts with a number of base meshes that conform to various rigs to speed up both modeling and animating.

I mean of course some individual tweaking goes on, but they make use of every shortcut they can
>>
>>501524
You shouldn't need to adjust it numerically unless you're engineering an actual product.
>>
>>501556
i get where you come from but "You shouldn't need to x" is not a valid argument
>>
>>498049
You can't sculpt UI meshes
>>
jewelry design

rhino masterrace fight me irl
>>
>>501952
Jet fuel doesn't melt steel beams
>>
File: ....webm (597.09 KB, 1280x720)
597.09 KB
597.09 KB WEBM
because i'm shit at art but good at technical
>>
>>502339
>basic softbody
>good at technical
>>
>>502387
>every pic is related to post
>>
>>499395
hope this is bait
>>
File: 1446382677966.jpg (54.26 KB, 1159x653)
54.26 KB
54.26 KB JPG
can't you just make an automated retopo tool like zremesher do biased decimation along the guidelines as well? that alongside adding edgeflow arrows to the guidelines would improve it greatly.
>>
>>502744
It already does biased reduction based on red/white/blue polypaint zones. And guides already dictate edgeflow...
>>
>>502745
i meant relative to each other so that poles aren't formed, and if it can't it just gives you an error or warning saying you need to adjust.
>>
>>502744
that program actually just was announced

https://www.youtube.com/watch?v=U6wtw6W4x3I&ab_channel=OlgaSorkine-Hornung



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.