Most efficient way to learn modeling?

Hello all!

I’ve been messing around with Blender in my free time, watching tons of Youtube videos and tutorials. I’ve learned a decent amount for the time I’ve spent with Blender but I feel like I’m trying to learn concepts that always need a firm understanding of some other basic concept, correct topology probably being the main one. I’m wondering from your perspective, what do you believe is the smartest way to go about learning blender modeling correctly and efficiently? I tend to just jump around, trying to gain any knowledge I can. I eventually come to a point where I believe I have a decent understanding of what I should be doing, but start modeling and always get to a crossroad where I can’t quite figure out what to do. From the feedback on other forums I usually get the answer of “Topology” being my main issue. I’ve been searching around for some Topology courses / classes online, I have no problem paying for them. But when doing so a new question always arises in my head “Should I be learning something BEFORE topology to completely have an understanding of whats going on.”

So my basic question is this: From everyone’s experience, what is the smartest and most efficient topics / concepts to understand and grasp before moving on? For example :
1.Learning Blenders UI
2.Topology
3.Etc.
4.Etc.

Is there common knowledge of this I just haven’t come across? I’d be willing to take a course / class online to better my skills and eventually become an actual Modeler.

Any information, links, courses / classes you can suggest would be greatly appreciated. Thank you for your time!

-Sean

You didn’t provide examples of your modeling, where it was problematic, what the goal was, so this is going to be very generic.

Yes the basics of the program are important https://www.youtube.com/playlist?list=PL0AfIdK08E7_PKsXOO_yuAql9b3hdcwPR

After that, if you’re completely new to polygonal modeling, a common path is to start with a modeling project. There are tutorials available that are structured as a course or as a larger project, cgcookie.com and cgmasters.net has those for example. Thing is, try as you might, you won’t find one big tutorial series that would teach you everything about modeling because there’s so much to it.


Modeling is about taking an existing design and turning it into a usable 3D model, or models, as modeling may have multiple outputs. A model is produced for a purpose: a still, animation, game engine use, 3D printing, or other. The list of requirements vary a lot depending on the end use, which means that the way the model is constructed and what kind of structure it has can be very different.

Then there are requirements from the pipeline. UV unwrapping is easier with a readable model structure. Subdivisions might need to be supported for texturing. If texturing uses baking, it could mean producing high and low polygon model. If the model deforms, the structure needs to be ready for rigging.

Modeling itself is done with the understanding of forms, proportions, what structures make up those forms, the structure requirements from the mentioned, and combining those with the tools you know to decide on a modeling workflow.

With experience, that’s the order of modeling: end use, pipeline, structure, simplified structure for the tools, and then you start modeling. No one learns that way though. Learning is about modeling the other way around to learn about the forms, proportions, structures, how the tools affect the structure, and then what the model would actually need to have included. Even the polygonal modeling basics might only be clear later when you really understand meshes, what implications it being about approximation has, or it being a surface type modeling paradigm, and how all that is related to mesh errors and problems, and also affects materials/lighting/rendering.

Topology is in all that. It’s about the structure flows, but it’s also about all the things that can affect those flows, like forms, deformations, and anatomy. A very extensive subject matter.

If you know what you’re going to be modeling for, that might help to focus your learning towards that goal in mind. Otherwise, a few more things:

  • Don’t model what you can’t see with your eyes. If you want to model your own design, you need to design it first. Use references
  • Don’t obsess about absolute precision. Polygonal modeling is done with straight edges and flat polygons which means it’s about approximation of curves and surfaces. You’re supposed to make an intelligent approximation, not a dimensionally accurate one.
  • Don’t try to swallow an ocean if you haven’t even tried a cup of water yet. Too difficult subject matter too soon can stop your learning
  • You don’t need to know every tool to model. Learn enough so you can make things and learn more as you go. The tools will change, your knowledge will change, and the workflows you’ve learned will also be in a constant change
  • Don’t expect to be done learning. That won’t happen if you’re doing it right.

Ask yourself if your model looks like what you want to it to look like. Watch tutorials of people modeling similar things, compare the tutorials against your own workflow to work out the kinks. Find out how you can make it simpler, faster and easier.

Basically just rinse and repeat.

I think the best way is to practice and so this mean create, follow tutorials, watch videos etc.

Each time you will potentially face a new problem. ( a panel in a curved surface, specifc jonction between pieces etc )… and this is when it happend that you can search, ask ( to you and other ): what i have do wrong or how could i model this for get the aspect i want it have.

It is really impossible to imagine learn all the solutions without having face the problem one time…

You start to understand you have miss some little things here and there, so now you have allready a base for the question. It is where start the learning… how i can model this the good way, what is a not so bad topology for this object … what i have do wrong the first time in situation that i can do better the next time etc… then you start to develop your workflow, your own methodology.