1153x1293

justfuckmyshitup.png

🧵 Untitled Thread

Anonymous No. 898420

>I'll give my client a topological disaster if I deem it so
Why are blendlets like this?

778x202

ngonsoup.png

Anonymous No. 898431

Russian 3d artist makes the switch to Blender creating weapons for Trace Studio and Call of Duty games thinks employers won't be looking at your wire frame so just never bother retopologizing anything in your portfolio. Meanwhile Red Storm entertainment cleans and retopos the hell out of CAD files.

https://www.artstation.com/mrgesy/profile

1280x720

8789315.jpg

Anonymous No. 898433

>but even with characters manual retopo isn't something that should be a thing because of machine learning and cyberdyne and shieet

Anonymous No. 898434

>>898420
While the points they raise are "true" to an extent, the point of "topo doesn't really matter if it doesn't deform", I feel like it's still a bad practice since having decent topo makes unwrapping quite simple. Do the work to save yourself the work later on. Or save the work of whoever is next on the pipeline so they can do THEIR job, not yours.

Anonymous No. 898448

Do blendlets not realize shit topo shades like shit? Nice triangle soup fag, look at them wobbly bevels.

Anonymous No. 898455

this is why these people don't have jobs.

Anonymous No. 898467

>will the object be rigged to deform
It doesn't have to be rigged for the mesh to break from using modeling deformers. And that ruskie sounds like a nightmare for those down the pipeline to work with. No wonder he doesn't have Maya listed on his art station.

Anonymous No. 898475

software schizos are seething because this guy got a job on infinity ward while they do nothing but sit on this board and shit at people who don't use autonigger tools

82x96

2022-05-20-165304....png

Anonymous No. 898476

Anonymous No. 898478

>>898476
I can't wait to learn how to make my own nigger!

Anonymous No. 898490

Are they talking about just the high poly mesh. It doesn't matter what topology that has as long as it shades well.

200x252

download.jpg

Anonymous No. 898491

>>898478
based

Anonymous No. 898493

>>898490
I think they are talking about game assets

Anonymous No. 898496

>>898431
>>898434
>>898455
>>898448
>>898420
>>898467
Tools like InstaLod eliminate the issue entirely.

Anonymous No. 898532

Deforming asset (animation, displacement) - has to be quads
Non-deforming - doesn't matter at all
Might want to try to go for quads wherever possible though

Anonymous No. 898594

>>898475
>job at Bobby Kotick's bioware-tier infinity ward
Guy works at an outsource company in Armenia. Issue is some he's saying this shit and someone new to 3d is thinking "oh I guess topology isn't a big deal and won't need quads unless I'm making Shrek because my bosss is a boomer." Go show some prof at gnomon who was a senior or better yet supervisor who worked on blockbuster. They'll shit their fucking pants. Also, he's just cooming over his guns while assuming everyone knows his specific workflow
>>898434
>Or save the work of whoever is next on the pipeline so they can do THEIR job, not yours.
Exactly this.

Anonymous No. 898597

>>898490
Found OP's thread. I'm guessing that's what the Hadriscus poster does since he retopos everything even though buildings especially stylized and for film aren't topologically complicated. One other guy just does high poly into CAD. Nobody mentioning poles which you be good in this discussion since you can get away with 6 if the connecting faces are flat like with default cylinder primitives. Rawalanche just wants a magic auto button for everything. and strictly auto uv unwraps to this day

>Unwrapping and UV layout is not an issue either. Unwrapping is nothing that should be done manually. You simply get your “triangle soup” model, select all faces, and do auto unwrap based on angle and then pack it using something like UVpackmaster addon, which gives you lot tighter pack than any human could possibly make in one click. That’s it.

https://blenderartists.org/t/what-happened-to-hardsurface-modelling/11

Anonymous No. 898969

>>898448
who gives a shit when i'm baking the normals off an hp model

Anonymous No. 898975

>>898476
Beat me to it.

Why is EVERYTHING these days advertised with an ugly nigger?

Anonymous No. 899926

"Good topology" is just whatever looks right in the end result. Yeah you should know how to do clean modeling for subdivision, but if you do subdivision modeling for every single highpoly you create you're wasting time. It's just one tool. Remeshing is another tool and choosing not to use it is just hobbling yourself. Have fun making one asset for three months while another guy is doing three highpolies a week and smoking you because he's comfortable with the entire toolset.

Anonymous No. 899928

>>898975

It's been like this forever. The mediocre looking female negro with an afro is literally everywhere in the computer industry. One of the oldest memes in existence.

Anonymous No. 899935

Good topology makes a great asset in a pipeline but it is creatively limiting. Not everything can be topologized properly. If you're aiming for a more stylized approach, good topology is a must because of unforgiving shading. Not to mention it's almost impossible to sculpt proper stylized characters. You will have to do subD modelling. Stylization is at its core hard surface modelling.

Bad topology, as in merging everything together, means more freedom to create. If you're going for something realistic it doesn't really matter if you're in a pipeline or not. There are a ton of workarounds to make then asset fit when you're not worried about polycount or subdivision.

Anonymous No. 899959

>>899935
>Not to mention it's almost impossible to sculpt proper stylized characters
Wow, you're a fucking idiot. Stop pretending you know what you're talking about.

Anonymous No. 899960

>>898431
Is this guy just your typical Russian maniac who fell through through the cracks or lying by omission? Wherever you're submitting a demo reel you're employers absolutely do care about your wireframe.

Pixologic is even telling you this
>Wireframes show potential employers that you understand important concepts like edge flow, the "Rule of Three" and how to generate low poly cage objects which are essential for gaming and animation.
>As a rule of thumb, you should show off your wireframes on the LOWEST SUBDIVISION LEVEL to enable employers to see the base structure of your mesh. However this will prevent them from seeing the more detailed forms that only appear at higher subdivision levels.

https://pixologic.com/interview/artist-in-action/demo-reel-with-zbrush/1/

>>899926
>"Good topology" is just whatever looks right in the end result.
That's up for riggers and animators to determine.

Anonymous No. 899961

>>899960
>That's up for riggers and animators to determine.
Deformation is obviously a factor in what looks right in the small fraction of assets that are going to be rigged and animated, so this is a meaningless attempt at a nitpick.

Anonymous No. 899975

>>899959
no he's absolutely right, that's why you zremesh as soon and as low as possible when you do stylized sculpting to get smooth surfaces.

In fact good stylized sculpting almost comes down to how little you can touch the mesh.

671x350

ngons.png

Anonymous No. 900007

How topology is assessed is dependent on context, balance, and what the model is used for. Your high poly model will generally be focused around supporting edge loops that control the high poly shape and prepare the mesh for baking in a way that doesn't result in artifacts before or after the bake. Low poly models focusing on optimizing performance in games are a balance between shape and poly count, trimming unnecessary/excess geometry for the desired shape. Regardless, one truth is that acceptable or even optimal topology is defined by what stage of the process your model is at and what you're planning on doing with it.
>but my creativity is impaired
That's what Zbrush's dynamesh and sculptris pro is for which you still need retopology for, but if we're all focusing only only on hard surface poly modeling let's take this premise into account: clean topology on your base mesh is easier to work with and edit making your life easier later on. Anyone who has put considerable practice into 3D find this out on their own. But let's say you made a topological mess of your hard surface? Are you going to retoplogize it or would you rather have just built up right the first time? What are you go to do about those edge loops that you determined you needed for your curved surface that are trailing to the other dense of the model in a dense cluster where they'll cause nothing but problems? Do you even want to use edge loops and control over where they go?

>>899960
Omission. He has to triangulate those n-gons before baking or exporting. Goes to show how being industry pro doesn't make you a good teacher. Just think how someone just learning the ropes of 3D with no grasp of the more technical details or nuances that comes from alot of practice and reflection. It's going to leave dumbass impressions and encourage bad habits for students or anti-intellectualism. Post like that are poison.

1280x720

maximumarmor.jpg

Anonymous No. 900025

On final note the dialogue on topology presented to people getting into 3D is becoming stupid and confused. Vague youtube videos waxing philosophy are partly to blame along with Blender communities. If you use quads, which there are proprietary software in film will accept nothing else, you're a dogmatic boomer using outdated workflow totally not cool like all the rebellious revolutionary zoomers rocking the future with triangle and n-gons.

But the pedagogy of monkey see monkey do but monkey not understand is what mainly enables this. You're told to do it a said way this because that's just the way it is, because it's better, or it's harmless. You don't know how to calibrate, optimize, nor how to use situationally the motions you're going through because you don't even understand what the means to the end is like someone swinging a plough around in a certain arc mindlessly. That's why I'm always thinking about the purpose of what I'm doing while learning, as if I were preparing to explain it to someone else. For example, tris could work for my model yet why do I start with quads? Because I plan to create isolated edge flow islands and work from there while having control of where my edge loops go. Why go for a tris instead of a quad later? because it's in a non deforming
area (or one that won't be seen in the situation) and the vertices between previously separated parts don't line up like buttons on a coat. No need to compromise the rest of the mesh when a tri can do the trick. N-gon on an armpit? Actually works out and you avoid a six pole here. Pixar explains
>Used sparsely, non-quads can be very useful to gather 3 or more diverging edge-loops. These are often encountered in highly deforming areas with curvature saddle-points (ex: arm-torso connection). The strategic placement of a pentagon in one of these critical spots ensures that the surface remains smooth, while allowing for complex topology to flow around.

Anonymous No. 900032

>>899959
NGMI

Anonymous No. 900918

>>900025
>which there are proprietary software in film will accept nothing else
Imagine believing this

Anonymous No. 900939

>>900918
I don't have to when I've talked directly with people who do this for a living where something like maya is so loaded with plugins and modified by software engineers it's not the same program anymore. It's only hard to imagine when you're strictly a Blender hobbyist, so you've never had to prepare a file for export in your life just so another software will accept the import. Yeah just imagine film studios that only want quads have algorithms that don't allow a mesh to be subdivided and preventing turds from being flushed down the pipeline and wasting everyone's time.