Are there Mathmathical or Algorithmic Ways to Decide on Most Efficient Ways to Model?

Support Modeling
1

Hello all, this is my first post here! I’m a professional graphic/motion designer who is new to 3D and Blender. My goal is to integrate professional level 3D into my work process as soon as possible.

[SUB]I use basic math and logic to help me be efficient at creating relevant and appealing designs. They are applied in areas of project management, design principles and process, and using software and some scripting for automation (I mainly use Photoshop, Illustrator, and After Effects). I’m currently experimenting with controlling repetition with variety for creating visually appealing designs efficiently.[/SUB]

I now have understanding of the basics of Blender, and idea of 3D workflow and how it fits into a complete design project workflow. I like to learn by becoming comfortable with each step in the workflow. At the moment, I’m comfortable with setting up reference images, navigating in 3D space, and transforming objects.

Now, I want of focus on modeling. I want to get to the point of being very efficient at modeling anything. These are the steps I’ve taken so far.

  1. I followed about 10 tutorial on modeling various things using Blender - While it was cool to see step by step modeling process in Blender, it was hard to understand WHY the modeler took each step. It’s hard to know if those steps were taken for a reason or it was just arbitrary. It’s hard to know what aspects will carry through for modeling other subjects.

  2. I read the book “Digital Modeling” - It takes you through various methods in modeling, and emphasizes that observation and knowing how to approach modeling is what separates pros from the amateurs. However, it does not really break down how to observe and how to use the observation to decide the most efficient steps, while it does show different methods applied to different subjects.

  3. I’ve trying reading math books on topology and computational and discrete geometry, and these materials flew over my head. They seem to assume knowledge of various forms of calculus, linear algebra, analysis, etc. which I don’t currently possess.

What can you recommend as my next step for learning how to be most efficient at modeling anything in Blender? Is there a way to bypass advanced math courses and know how to algorithmically decide on most efficient way to modeling any subject? I know efficiency also depends on personal habits, but I’m starting with a clean slate for the most part.

2

Moved from “General Forums > Blender and CG Discussions” to “Support > Modeling”

3

you can generate a sun flower with “Fermat’s spiral”.
a very basic primitive landscape with only a few polygons can be distorted into to a very realistic landscape using a “Fractal”.

4

It’s very much about having an artistic eye. Trying to reduce it all to mathematical proofs and theorems will drive you insane. If it looks good, it looks good. If you’re not happy with it, it doesn’t. Things like the golden ratio, the rule of thirds, and Fermat’s spiral aside (which you can certainly carry over from design work) it’s really about doing it long enough to gain an eye for what works and what doesn’t.

5

you need 3 things

1 - model practice
2 - pratice model
3 - practice both 1 and 2

learn all the different tools available and practice anything
that way you 'll learn how to do things with different tools
and begin to get the feelling of modelling organic or mechanical models

and no you cannot learn that in 3 weeks sorry there is a stiff curve to learn blender and all the tools
for modelling rigging animation physics material texturing lighting !

happy blendering

6

My advise would be to throw the math out the window and let experience be your teacher. The number of permutations to create any given object can be quite large. There is no correct or set rule for creation, and the more you practise, the more you will discover and be able to apply actions from one thing to another… I think you are really overthinking things here.

7

You could check out TopMod or Structure Synth.

8

Hello Fweeb, thanks for moving it to the right place. I read your book as well, and it was very helpful in understanding how Blender works.:eyebrowlift:

Thanks for the cool tips!

Thanks. I figured some of the replies would be along these lines, and I agree that the artist’s eyes are the ultimate guide. I do use the golden ratio, rule of thirds, and other methods(perspectives, making intervals vary in all the features such as space, value, color, etc to make things look interesting, etc.) to support my visual aesthetic sense and I feel these principles are very useful.

Thanks! A steep learning curve indeed! I don’t expect to master 3D in a few weeks, I just want to learn it in most efficient way as possible. Practice I shall. :yes:

So I have slept on this and came up with a “matrix” of modeling methods and some “metrics”. I’d love to hear some feedback.

I’ve tried to order them roughly by big general methods to more detailed level. Seems logical to see if the “larger” methods are appropriate before moving down the list per modeling subjects, and combining the methods appropriately as you progress.

Primitive Modeling - Seems to be the “largest” method, since you’re starting off with 3D objects and combining them and modifying the shapes.

Box Modeling - Also starting off with a 3D object, and then “growing” the form via extending and beveling faces(2D).

Sculpting - Seems like a different way of shaping a 3D object.

NURB/Spline Modeling - Modeling based on groups of 2D outlines.

Edge Extend Modeling - Based on 2D polygons, extending the 1D edges to grow more polygons.

Point by Point Modeling - Adding vertices one by one to make 1D edges and by extension 2D polygons.