[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] [Search] [Home]
Board
Settings Home
/3/ - 3DCG


Thread archived.
You cannot reply anymore.



File: hand subdv3.png (217 KB, 960x540)
217 KB
217 KB PNG
is it acceptable for sculpting and animation?
>>
File: hand subdv4.png (170 KB, 960x540)
170 KB
170 KB PNG
>>
File: hand subdv1.png (226 KB, 960x540)
226 KB
226 KB PNG
>>
>>678260
the only thing that you need to worry about when it gets thrown into animation is having the proper definition in areas, otherwise it doesnt entirely matter how the mesh looks, and you dont throw a subdiv model into sculpting thats usually the final model. you would have to take the low poly version of that subdiv mesh and subdivide it even further to sculpt the details at a higher poly and then convert the details into a baked normal.
>>
anyways this doesnt look like subdiv work, it doesnt look smooth and the wireframe isnt bending smoothly, it looks like your low poly version not subdivided. that and also the wireframe is atrocious simply because the mesh is heavily uneven and non uniform, its chaotic as fuck and just look dumb.
>>
File: hand base1.png (238 KB, 1080x1080)
238 KB
238 KB PNG
its a subd mesh. here is the base
>>
File: hand base.png (240 KB, 1080x1080)
240 KB
240 KB PNG
>>
>>678266
ok that is even worse, its just plain ugly and uneven. relax that shit and or use the sculpting grab tool to put that shit in place.
>>
>>678268
:C
>>
>>678260
nope, and thats even ignoring the too small fingers
i mean you can animate that, has enough tris to skin it, but i hate the topo flow, those ngons piss me the fuck off
>>
>>678266
yikes
dat is nut gud...
>>
File: hand subdv 6.png (255 KB, 1080x1080)
255 KB
255 KB PNG
somewhat cleaner?
>>
>>678272
you haven't animated before right?

that's overkill for animation, even tho the people here will tell you it suck.
we use topology to make it easier for the animator, the animator can work around anything but its just easier when the weights deform properly with enough loops in deformation areas. there is no "un-animatable" mesh.
>>
>>678272
Is the character going to be fat?
>>
>>678272
>amerimutt hands
>>
File: feet.png (121 KB, 963x822)
121 KB
121 KB PNG
The only way I do human digits these days
>>
File: hands+feet.png (1.32 MB, 1183x1765)
1.32 MB
1.32 MB PNG
>>678298
nice

>>678272
i like this approach of modeling too it is pretty fast and i like to deal with polygons instead of brushes.
the tools that we have are triangles, quads and 5-gons more than that will translate to ugly poles that can mess the shading, result in bad deformation and bad topology
>>
>>678304
learn about aesthetics because those just look ugly.
>>
>>678304
>the tools that we have are triangles, quads and 5-gons
firstly those are not tools
secondly your pic is a prime example of bad topology
and thirdly learn how to model because that wire frame is just as atrocious as ops with the uneven bullshit going on.
>>
File: fivegon.jpg (95 KB, 1284x768)
95 KB
95 KB JPG
>>678332
5-gons have been used quite successfully by some pretty noteworthy artists. Know Stahlberg did a lot of that stuff.
If you can make unconventional topology work for you there is nothing wrong with it, the real goal is to have smooth shading not making a pleasing wireframe.

Depending on how much curvature you aim to describe with your base geometry you'll end up with many uneven loops to solve into quads only.
Attempting to quad everything will lead to a highly variable density in areas with lot's of poles (places where more than 4 edges connects to the same vertex).
Poles will create pinching problems in shading, but poles also can't be avoided entirely, just minimized.
So an advanced answer is to use triangles (which subdivide predictably into 3quads), or 5-gons that subdivide predictably into 5quads.
(I don't like using 5-gons myself but it's def a valid strat)

>Pic related is what's going on and why it is a valid solution.

It's cute to see tr/3/pots criticize Op's topology when it's evident from the surface flow Op has a deeper understanding of this than most of you all.
>>
File: hand base subd compare.png (838 KB, 1863x853)
838 KB
838 KB PNG
>>678334
Blenderella's approach (the approach everyone seems to favor here) is tedious. i don't have the attention span to finesse 2000 vertices on the hand alone. I let the program resolve the n-gons, so I can work at 500 vertices and get farther faster.
>>
File: forallquadsNPCs.png (826 KB, 1028x1510)
826 KB
826 KB PNG
this approach is very fun and it is always amusing to trigger the all-quads-NPCs.
>>
File: 1555470250115.gif (486 KB, 182x190)
486 KB
486 KB GIF
>>678343
>trigger the all-quads-NPCs
>blendlet doing shit but feeling superior to those who actually know how to create correct topology
>>
>>678342
>get farther faster
yeah that's great but I'd recommend slowing down and focusing more on form and silhouette
>>
>>678334
Top tier explanation
>>
>>678298
What is this?
>>
File: garb top.jpg (330 KB, 2483x991)
330 KB
330 KB JPG
>>678334
ok let me debunk your shitty argument because your so fucking inexperienced in subdivision modelling its not even fucking funny and why your models look like fucking apes and retarded looking.
firstly you dont just throw triangles and n gons just because you can, thats not how it works and thats never how it will work. with any modelling work the simple workflow is to add definition to areas, usually by running a loop through it and in some cases you close that loop into a triangle because it doesnt need to continue(pic related) it just gives you more curvature where you need it. and thats not what you have going on in your shitty topology. the simple reason why your topology just doesnt work, is because you have RANDOM fucking extrude LOOPS aka "c-loops" because theyre the most common loop you use. its not used to define surfaces whats so fucking ever. i can tell you why youre using it because i did the same exact shit before, you saw an infographic image that describe these loops as "loop reduction" methods when in fact thats not how theyre used at all. just look at how dense the palm and back of the hand are, those are suppose to be less dense. and the most sinful thing about your topology is the random fucking broken loops that dont add anything to your model, it doesnt need the extra definition and it just makes it look even more ugly being left as an ngon, look at the image on left and look at where the n gons are located. it breaks in spots because the loops dont need to fucking continue, since it will make the mesh denser than it needs to be.
>>
File: mod_notes.5.png (160 KB, 656x409)
160 KB
160 KB PNG
>>678423
>>678334
also to add to it, pixar has a thing called saddle points that are n gons placed in areas to give the mesh better curvature, thats a GOOD way to use n gons, not the shitty mess and abusiveness of what youre doing.
>>
File: suzanne_+_yosho.png (966 KB, 1142x881)
966 KB
966 KB PNG
>>678427
you comparing one thing that has no relation with the other.
here is suzanne, suzanne is subsurf modeled bcoz she is supposed to be hundred of thousands polys. with subsurf modeling it can be done fast.

here is yosho, yosho will be shadeless, hand painted. yosho can be rendered in real time.
>>
>>678455
yosho will not be subdivided. like the anime girl in your pic
>>
>>678455
are you retarded? i wasnt comparing two different things you ignoramus, you are viewing your dumb ass model in insoline mode which is viewing the low poly version of your fucking model not the high poly, so i was comparing low poly version on both sides and your being the out most shitty one.
>>
>>678455
>>678423
further more look at your fucking pic, you have isoline view off and showing the higher poly version like >>678343, good job retard.
>>
>>678423
Take a deep breath and unrustle your jimmies anon, I'm not Op just an oldtime fellow tr/3/pot chiming in because this topic is my pet peeve.
It keeps on resurfacing because a lot of 3D newbies are told they should space out their edgeflow and quad their geometry - end of discussion.

They are not educated on exactly why they are to be doing this so they hold this simplified idea in their mind that becomes
this aggressive dogma because some 'guru' type character told them 'u art not to triangle, u must never N-gon'.

Since a lot of people gave them shit on messageboards for their sloppy topology when they started out many of them now sit hot on the trigger
waiting to take someones head off for malpracticing topology, a topic they themselves don't understand anywhere near as well as they think they do.

For anyone interested to gain further insight here's a link to a archived discussion of this matter over at polycount that raises many key points.
https://polycount.com/discussion/146695/polycount-to-professors-ngons-are-useful
>>
File: handscreen2.jpg (220 KB, 635x589)
220 KB
220 KB JPG
>>678461
This palm geometry from that thread by the aforementioned Steven Stahler is a good example of something that looks like a tangled mess in the basewire.
It's utilizing not only pentagons but even hexagons, but doing so in a very deliberate way so that under subdivision it'll generates specific topology in areas.

Some of it I'd like to call into question, but in other places you can def see how it can be successfully utilized to create complex curvature changes with very few verts.
This isn't something one should be attempting to do starting out since many important concepts are obscured by building geometry this way.
But it's something you should be aware of and may attempt to utilize in places once you're operating on a expert level.
>>
File: 32-gon.png (120 KB, 417x426)
120 KB
120 KB PNG
>>678461
>u must never N-gon
say that to my face and not online and see what happens
>>
File: 4rqrq4eqe3.jpg (71 KB, 920x1020)
71 KB
71 KB JPG
>>678463
but that is noting like what youre doing, that is in essence sculpting but with poly modelling which youre not doing, yours is too rounded and look like limbs of a fat girl and in some cases of that palm he fucked up areas and some areas dont make sense, so this is not a good example by any means, hes doing the same thing that jap guy is, just more ugly. but again this is not what youre doing.

>>678461
youre definitely right, cant help these retards, just hope they dont spread their shitty ideology els where, ESPECIALLY since they have no idea what theyre talking about.
>>
File: study.png (1.06 MB, 1792x804)
1.06 MB
1.06 MB PNG
>>678468
wow thats amazing. pls say who is he thats the level i want to achieve.
>>
>>678476
Curvy McCurvington

just offset your geo lol
>>
>>678468
Do you even read the posts you are responding to anon?

Not a very good look to be deflecting calling everyone and everything 'shitty retarded multi-fucks' left and right
ESPECIALLY not at the same time as you're toggling CAPS on people while engaging in DAMAGE CONTROL type behavior.
https://www.youtube.com/watch?v=ay9ZxsulN2c
>>
>>678480
is your brain damaged from all the damage control? or are you just illiterate?



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.