Guys, I have a huge problem. I need a methodology and a pipeline for creating 3d models of complex and simple shapes. It's like I've been standing still for 2 years and I'm not moving from a dead point. Please help me. Tell us how models were made for games such as GTA san andreas or mario 64 or cs go or some AAA projects. I really really need it. There are very few lessons on this topic on the Internet. I want to make props and objects and characters.Thanks for the help. Below are examples of things I can't do.
My goal is to make props but i don t know the pipeline in addition, you need to start with a blockout. Not every object can be created from simple forms
literally all objects can be made from simple forms.
ok - so in order for anyone to help we need to know how far you have got already.
if you have never opened up a 3d application you're in a very different position from someone who has been learning for some months and needs help on specific issues.
I don't understand how you've gone two years wanting to learn and not learned anything.
There's a HUGE amount of free resources online for this. Blender's tutorials alone should be enough to take anyone from beginner to at least moderate skill.
Even just looking at those images should tell you a lot.
How about starting by saying exactly what you do know?
This is a blockout of my model. I made it a hour. I don't understand the pipeline itself of creating complex shapes, such as cars or something like that. I can model simple things, that's how it is. please read more carefully))
have you ever googled "3d art for beginners"? Forget fancy AAA lingo like 'pipeline' lol you need to look up introductory tutorials. Everyone's been telling you to search for tutorials online. Please read more carefully
This site also has a wiki.. and a very famous thread about how to model some "not so easy parts" of a model after the first blockout gave you the general form.. ( even renaming the headline because.. users get too "demanding" ?? ).. it's all there.. since years.. ( and you do not even have to register to see it.. like in other fancy look XY Grrg sites.. )
Hey bud, this course really helped me grasp my understanding of sub-d modeling. I'd suggest getting the hang of sub-d before moving onto game resolution assets, because a lot of times those require sub-d modeling early in the pipeline. If its still the same, you can demo a Pluralsight account for a few days and blitz through this tutorial, but it shows you how you would model complex shapes. Can't recommend it enough.
This is a blockout of my model. I made it a hour. I don't understand the pipeline itself of creating complex shapes, such as cars or something like that. I can model simple things, that's how it is...
Well at least it's a reasonable start!
Also by the way was it modeled off of a
background reference image or concept art?
I’ll recommend finishing translating
each individual blocked piece, into a high resolution (sub divided high poly)
artifact free base mesh which in my opinion believe is the actual core issue,
preventing further progress.
So here's a few pointers how I’d personally approach
unifying complex hard surface shapes into one cohesive object, beginning with
this thing's main body:
Firstly,
fairly standard practice while modeling your high is to always keep in mind how
your low poly will 'look' in terms of retopologized efficiency
because whether portfolio or in-engine optimized game ready model, it's really the
only element a player or viewing audience will eventually see, once rendered
out.
Right
then for this example, I'll use an edge control/support loop workflow in
Blender, again whilst keeping an eye on creating an intended generated proxy, therefore
making the entire process less grinding than it should be i.e. by simply
deleting applicable loops.
Now
the primary cylindrical shape has vertically opposed varying sized intersecting
cylinders, located longitudinally along it’s length and since we’re mostly
dealing with curved surfaces, selecting optimal segment counts can be crucial. So
I chose 34 x 26 x 22 for both main and secondary cylinders respectively, although
that being said there’s no rule of thumb per se, other than maintaining a consistent
near enough segment match, as follows:
Next
we’ll run a Boolean operation via applying the modifier operand set too Union, thus
joining our shapes.
Typically
there’s some cleanup involved, welding stray vertices while we manually cut
support loops with the knife tool alongside applying the Mirror modifier,
making the job slightly easier.
Ok, error checking our work, applying a reflective material,
smooth shaded topology and Subdivision Surface modifier set too level 2.
Eyeballing the
reference, continuing to sketch plus edge extrude additional detail
Moving
on, basic reuse of existing geometry by duplication, in order to replicate other structural parts of this high poly mesh.
Anyhow, have a crack completing what you’d
initially started!
More relevant info:
Frank Polygon’s Sketchbook - practically a hard surface wizard plus helpfully he also authored an in-depth breakdown describing every stage of the current gen pipeline.
guys I understood everything, I found an awesome YouTube channel where there are answers to most of my questions! Thanks for the support! https://www.youtube.com/@OnMars3D it is him
Lol...guess couldn't let this one go, anyway had actually personally struck a chord remembering how it was when I had started out. Aside from the OP, many have bumped up against the frustrating challenge CG art can often represent...this 3D medium is no two ways about it really f**king hard! technically let alone artistically but all the same utterly rewarding, especially after long hours creating 'THE' thing you've focused a lot of time and effort tweaking into existence plus lets be entirely honest here - not always fun : / however in my opinion whether a potential career aspiration or simply an amateur pastime. I mean like there's this sort of addicted concentrated mindset, takes over control to further progress ones overall skill just by attempting to generate increasingly better more complex stuff, which is where I think initially a lot of people experience their first, bump in the road.
Nowadays as a novice, they're spoilt for choice in the sheer number of avenues too exploit learning how to create 3D content whether formal academia or online self-taught tutorial based education, practically the "world is indeed your oyster", although that all being said developing our object into something resembling a complete PBR textured asset rather than an uninspiring blinn-phong shaded model, specifically if we're utilizing a FOSS pipeline for those of us on a budget, does tend from the outset very likely laborious to achieve.
Therefore, in order too 'pay-forward' assorted learning's over the years, here's a straightforward approach I've been faffing about with since my earlier post upthread. A kind of pre-vis conceptualized staged process, implementing a rapid iterative paradigm tri-planar PBR material tiled photo sourced image workflow, enabled in the lookdev (material shading) viewport, without UVs or baked data just raw smoothed sub-divided meshes essentially driven through Blender's nodal system plus additionally computed via minimal processing overhead, as well i.e. a $1300USD off the shelf gaming laptop, Ryzen 7 series CPU - RTX3060 GPU - Memory 16GB, as follows:
Tutorial YT Series:- EEVEE and Cycles Materials System - authored by Mr Chipp Walters
Replies
ok - so in order for anyone to help we need to know how far you have got already.
if you have never opened up a 3d application you're in a very different position from someone who has been learning for some months and needs help on specific issues.
literally first thing I get when I search "modeling car in blender"
https://www.youtube.com/watch?v=VGPvxIrobFE
Error 1009
(((Well at least it's a reasonable start!
Also by the way was it modeled off of a background reference image or concept art?
I’ll recommend finishing translating each individual blocked piece, into a high resolution (sub divided high poly) artifact free base mesh which in my opinion believe is the actual core issue, preventing further progress.
So here's a few pointers how I’d personally approach unifying complex hard surface shapes into one cohesive object, beginning with this thing's main body:
Anyhow, have a crack completing what you’d initially started!
More relevant info:
Frank Polygon’s Sketchbook - practically a hard surface wizard plus helpfully he also authored an in-depth breakdown describing every stage of the current gen pipeline.
http://wiki.polycount.com/wiki/BaseMesh
And tons of example wireframes to examine and learn from
http://wiki.polycount.com/wiki/BodyTopology
http://wiki.polycount.com/wiki/Limb_Topology
https://www.youtube.com/@OnMars3D it is him