/3/ - 3DCG

isnt there some sort of preserve edges button

>>386054
yes, the result isn't much different tho. it works but it doesn't give you a mesh that you could import without errors.

non-zbrush user question-

see how the retopo'd quads have 2 diagonals on them in places? does this mean the tool resolution before retopo matters? just seems weird to me. wanted to ask if it isn't ok after export, but apparently it isnt. care to show a screen of how it looks imported into any non-sculptor software?

>>386060
its a cad obj export. so I can do it a number of ways. firstly ngons. pic related.

I know this isn't the same as starting with correct topology and subdividing it - THEN using remesh to unsubdivide it.

but then isn't the point of this exercise to retopo a poor mesh to something usable.

>>386064
heres an alternate export of tris n quads.
the ngon model actually gets converted to tris n quads anyway when zbrush imports it.

>>386066
heres what that looks like inside zbrush

>>386070
if I remesh it to be the same I get this.

there was also the tri to quad command I ran that before, doesn't make a difference.

>>386072
if I remesh to the lowest setting I get this.

which is interesting that it managed to fix the top. this is without edge preserve

>>386073
and if I remesh with edge preserve, I get this which is the lowest polycount it will allow.

>>386072
this one looks good to me. needs a bit cleaning up manually, and polycount is far too high ofcourse, but other than that not too bad, topologywise.

>>386073
oh god this is just shit. guessing too low polycount?

I cant even find edge preserve off the bat, but isn't situations like this where you would use zremeshers adaptive size? From the tutorials ive seen it's what allows more detail in places.

if not, sorry mate, just learning myself.

>>386080
>72 is not bad but I cannot get the polycount down any lower as the result is >74

edge preserve is actually freeze border.

I don't use this program often enough to know the best way of using its features,
so maybe there is a better way to get it to work, I just don't know atm.
which is why I posted here.
many experts here.

I cannot find good tutorial for zb 4R6 remesh that explains everything.

>>386083
this is the one i watched on zremesh. first hit on youtube, did a good job or remeshing several times with different settings to show the difference. was enough for me so i didnt even click on the lower ones.

http://www.youtube.com/watch?v=CVoFi2mnyhA

>>386086
ok will watch this, back later.

watched vid, used curves, re meshed multiple times with different settings.

worked best when it was split up.
cant do it in one go, you have to go down in steps with different settings each time.
still aint found the perfect way of doing it.

im done messing for tday. out of patience.

>>386136
Have you tried using polypainting around the edges to define density there? It could work out nicer that way, along with curves that have very high curve strength set.

>>386137
I didn't try the polypainting method yet. I don't know it that would make a difference because if you look at the first image I posted, its not that it cant do the topology when it wants to its that it cant position the verts correctly afterwards.

im pretty tired so feel free to experiment on your own and post results. may have more of a mess tmorra or something.

Is all this really worth it to work with hard surfaces in zBrush? Wouldn't it just be better to retopo the shape manually in another application?

>>386149
That's the question. z remesher was meant to make retopo easy, doesn't seem that's the case for hard surface objects.
manual retopo is essentially modeling something twice.
if there is an easier option im certainly willing to give it a go.
atleast now I can well and truly tell someone to stfu when they suggest loljustremeshinzbrush

>>386152
>atleast now I can well and truly tell someone to stfu

i wouldn't bet on it. many people will likely say you're just doing it wrong or call you a noob before suggesting that the application isn't specialized with hard surface models as well.

going from this

best i can do is 1700 tris. just remeshing produces the same shit you had, dynameshing it first seemed to help. target polycount all the way down, adaptive size all the way up, curve strength all the way down and then slowly move polycount and adapt. size towards the middle. still requires a bit of cleanup but doesnt seem too horrible, compared to making it from scratch i think.

>>386168
>best i can do is 1700 tris

....for such a simple prop, this is truly pathetic. You need to optimize for a game, because thats where you'll make your money

>>386052
>automatic retopo for hard surfaces
do manual retopo

>>386169
>theres only money in games

as for the rest the rest, uhh yes cpt obv sir, that's why i mention the polycount.

>>386173
there is only money in games. Thats a fact. Nobody will buy your 3 minute short film you made. No one

>>386174
>implying i can make even a 3 minute short film.

theres still people that make money in AAA movies, ads, product presentations, realtime architecture,... whether or not these people use zbrush, is a different question.

but srsly, as for games, if zbrush cant get the polycount down on a simple cylinder, why should it do it better for complex objects? what i'm saying is, the anon who said manual retopo might be right.

>>386168
I did one more test b4 I go to bed as I was curious about something.

if u start with good topology, then subdivide then decimate. it works. BUT that requires you do that workflow. have a good basemesh first.

for me as a cad user, doesn't really solve anything. most people who polymodel dont do boolians to end up with tris. so most may never run into this problem.

so it may be a viable workflow for some.

>>386175
Why are you responding seriously to a troll?

>>386175
>theres still people that make money in AAA movies, ads, product presentations, realtime architecture,... whether or not these people use zbrush, is a different question.

you will never make it in aaa anything, and if you do, you'll be making close to minimum wage for a few months, then the job will end, and you will be on the street.

realtime architecture? prepare to give all your money away to renderfarms.

You fucked up.

>>386178
didn't consider that. assumed he's genuinely retarded.

the zremesher isn't for lowering your polycount, its for creating a better topology with more evenly sized polys, for better sculpting/deformation etc. if all you care about is lowering your polys use decimation master

>>386188
Amazing Post.

>>386177
well, so what, i can only have a good topology in the end if i already have good topology in the first place? isn't the point of retopo to get you know.. good topology?

i'm telling you, we have like three diferent outcomes now based on approach, we need someone who actually understands the software to tell us where we fuck up. it's likely we're doing that.

>>386189
and end up with ugly triangles? that's no fun.

>>386189
albeit free, but still addon. are you saying there's no way in vanilla zb?

>>386194
All official ZBrush "plugins" are part of the installation

>>386196
fuck me. thanks will try.

>>386052
are you doing the projection to maintain the shape?
also learn to use the curve function for the retop or even the retopology brush

When you use Zremesher, You can use the curves to create a sharp edge around whatever you need.
Click on the edge of whatever it may be, hold shift, drag to another side of the shape, when you see the ring around it form let go of shift, ta-da.
Might help, if you haven't already done it.

>>386201
Here's an example.
Use the Zremesher Guides for this.
Also mind the line going down the center.
Shouldn't be like that, just had to stretch my pinky far to hit the printscreen.

>>386189
Wtf are you talking about. You don't seem to know much about zRemesher. What you're talking about is Dynamesh, completely different. zRemesher analyzes your surface and tries to distribute polys more where it is needed, while also computing what it thinks are optimal edgeflows. You can lay down guides to further dictate edgeflow, and you can put on polypaint to further dictate density where you want it.

you could do this cleanly in a couple of minutes if you were using a real 3d app

>>386240
nigger millenia, a piston with a bottom? guess weapons have no internal combustion engines.

besides, this is but one example, there are complex things as well where zbrush will be faster.

>>386241

Was looking at >>386066

and yeah it's a case by case basis which is more efficient but usually for clean hard surface you want to do poly modeling

Also, you get very clean subdivision if you merge two relatively high density cylinders so that the segments match, clean it up a little and add a few supporting loops:

>>386243
i think most everybody here knows how to do it in a "real 3d app", question is can we save time by adding zbrush to the workflow.

so far as you can see, nothing leads to a result as clean as yours, but i still maintain we havent found the right switches yet.

i mean, assuming it would produce the same quality topology and be as clean, but always take you pretty much two minutes instead of increasing linearly with the complexity of the object, wouldn't you use it? because if not i would get the feeling that you're an anti zbrush elitist.

>>386242
>and yeah it's a case by case basis which is more efficient but usually for clean hard surface you want to do poly modeling
Please don't come on here pretending to be an expert on this when you've really only done texturing work and haven't worked in a real studio environment as a modeler.

>>386243
You're teaching babby level stuff that most people know. What's happening in this thread is trying to see how capable zRemesher can be. Trying to discover new workflows that could possibly be faster. That is what a proper 3D artist does. They don't just stick with the tried and true methods, they try new things.

>>386245
> That is what a proper 3D artist does. They don't just stick with the tried and true methods, they try new things.
Proper artists use the most efficient method for their goals you retarded faggot. Pretentious failures who try to hide their lack of modelling knowledge do autistic stuff like that under the excuse of 'experimentation'.

That is the cleanest mesh you will get, and his method is the correct one to get it. Anything else is inefficient, wrong or simply stupid.

>>386246
not that i mind you shitting on the guy who uses the expression babby level, but if youre saying you can decide what the most efficient workflow is without trying for yourself, youre dependant on other people's opinions and will never surpass the average youtube you get your information from.

>>386248
Meanwhile you waste hours on your 'experimentation' and end up with a spastic result, while I already finish my mesh with perfect topology.

>>386249

after which, i can reasonably explain why zbrush might not be the best tool for this task, while you stay dumb and parrot back what other people told you. whatever rubs your buddha

>>386249
>implying a person works 24/7
>implying this is for a job

This is called practicing on your own time, furthering your skills, something you don't have a chance to do so much of while working. Stop being so retarded, please.

>>386250
>>386251
sad...

>>386252
is this the damage control ive been hearing about?

>>386253
> Using proper topology and efficiency in models
> Damage control
Kids these days...using words without knowing their meanings...

>>386253
don't reply to trolls. holy shit how hard is this?

>>386254
Don't reply to kids. Especially such illiterate ones

>>386243
>>386246
Those look like the buttmad jellyboys who got way too jealous and made the shitposting thread about your spaceships Milly. It never fails to entertain me to see people cry over your great work.

>>386240
YOU FUCKING IDIOT the original model was a CAD export. If you were modeling from scratch you wouldn't need to retopo.
The shape is just a random example.
I don't have a problem with you but alteast read the fucking thread before you reply.

O lawdy, I see this descended into arguing as usual..
>>386201
>>386202
Oh okay.. when I get some time I will add this to things to try.

Even if this doesn't achieve what I wanted originally, its certainly useful info that will help me model better.

>>386271

i fucking know right?
i tried decimation master like >>386189 said. it does give you control over polycount, but ugly tris like the other anon said. you can zremesh afterwards though, if you put adaptive size to 0 it'll actually keep the same polycount. of course theres wiggling room, but definitely something to play with.

okay I was able to add some loops using shift.

some just didn't want to go where I wanted them like the cylinder edge so I had to draw those by hand, which is why they wonkey. I think this affected the end result.

after one remesh same, then second remesh down to 0.1
I got this. im wondering if its possible to use some other zbrush tool after step 2 to clean it up a bit.
ie redistribute the polygons in a more equal way. so the loops aren't so even.

last image is imported into blender to get rid of those triangles that zbrush seems to display.

>>386052
For such a simple object, I don't see why you're trying to remesh this in zbrush...this would literally take 5 minutes to make in any 3d package, even blender

>>386442
Kids really are too impatient to read shit these days.

>>386442
You are on /3/, everything here has to be made in zbrush or else you wasted 00.4 secs and because of that you will never find a job in "the industry"

>>386457
>typical auto-retopologize algorithm hate bandwagon.

Maybe OP just wants to understand how it works so they can use it on a more complicated object?

>>386460
im not hating on zbrush, I like zbrush.
but some faggots here seem to think that you have to use it for literally everything

even stuff like sketching

>>386460
somebody who isn't a tard. its for high poly cad models.
tho im thinking now maybe theres a max workflow that's better. or can be used in conjunction with. pic relation
I don't use this app much either so gotta do some reading first.

>>386516
>this image
Is this some kind of bad shoop?

>>386591
in what way? I don't use max or the nurbs in it to have a clue what its representing doing.

>>386601
seems like labels have been switched

>>386606
lol. yea I got round to watch the vid its not exactly a re-quadder, it skins curves and its showing it does it better than max.

last vid.
http://www.scriptspot.com/3ds-max/scripts/polynurbs-1-0-released

doesn't work in 2014
theres also this but its pay.
http://www.scriptspot.com/3ds-max/scripts/polygon-creator

and de triangulate that I couldn't get to work.
someone says quadder and optimize do the same thing but I have not been able to get them to produce a nice quad mesh either.
http://www.scriptspot.com/3ds-max/scripts/detriangulator

>>386611
Lmao, Max needs scripts to do that. Check out Maya's Nurbs to Poly feature, it will retain quads and edgeflow. It can also do Nurbs to Subdiv.

>>386060
the "2 diagonals" mean the faces are not planar