Please help me finally getting my feet wet with Subdivision Surface Modeling!
Hello everyone!
Man, Subdivision Surface Modeling is something I've been trying to learn for many years now. But I seem to learn in a very different fashion than many people. I can't really learn well with "just trying around", I need a structured approach with heavy focus on the fundamentals. I am just a Hobbyist and don't have too much time on my hand, so when I'm in the mood to practice more modeling, I'm finding myself just using methods that give me good results quickly (like Boolean + Remesh, Hard Edge Chamfering / Beveling and then yeeting a Subdivision Surface modifier on top of it). I mainly do hard Surface things so it always kind of worked. Kind of.
Once there were more organic shapes or even characters (even if highgly stylized) I really quickly approached my skill limit and stopped pursuing it / lost motivation. Decided to learn Subdivision Surface modeling for real, watched tutorials, followed people along modeling things, I got the basics, know some interesting topologies for some specific shapes, asked questions, realize that there are tons of other concepts I never heard about, and once I have my own model I have no idea how to go about it. My model I wanna practice with is different enough that I struggle with applying the concepts I have learned, I fail to see "situations" where a specific action should be applied, if that makes any sense. I noticed, throughout many many tutorials, that people are subconsciously counting vertices, considering pole placement, know tricks and tips and prerequesites they need to set up to achieve a certain shape and a smooth model, or know how and where to re-route loops -- stuff that gets entirely lost because all I see is that Dude #1 pulls that edge up and cuts that poly there. They show a reference picture and just "get started", no mention why they chose this amount of Segments, why they insert so many loops and what would happen if they didn't, why it's okay to place a pole here and not there, etc... I know how to make *that specific* model in the tutorial, but I don't know how to make my own models, or to solve issues that I have with my own models.
Too many courses and tutorials show me how to create one specific model, but what I would like to learn is the basics and "strategies", real in-depth, and then have multiple examples or even exercises that test me on these principles, before we move on. Oftentimes tutorials or even paid courses feels like "look, Subdivision Surface smooths out your model, put edges to make it sharp. Now let's go model a head!", others are like "Let's learn subdivision surface modeling! So let's work with lattices and modifiers and scripts to get that real real complex shape! Done!" They are so unfocused and, for me, is unfortunately a terrible experience.
So I found that just watching tutorials and working along doesn't seem to work well enough for me, as the concepts are too convoluted for me. I think I am this kind of person who needs a lot of hand-holding and a very structured approach that really nails in basics. Can anyone relate to my situation? What did *you* do to "git gud"? Which is why I am finally writing this here thread. I think I am a bit stuck with how to approach even learning Subdivision, despite all the content that's available. Now with having all that said, there are two thoughts left:
- Am I just too impatient? Maybe I need to follow not 10, not 100, but 1000 tutorials modeling something with Subdivision to at some point recognize patterns and "conventions" myself. Not my preferred method but if that's the way to go I at least have a path to follow. So far though it seems like it's not going anywhere.
- Any resources you can recommend that satisfies basics and are at least a little bit "hand hold-y"? Books, Courses, Youtube Channels, anything -- definitely ready to spend money!
- What did you do to get as good as you are? What helped you the most? What would you recommend a beginner who wants to get started?
Thanks for reading, everyone. I seem to be someone who needs hand holding or guided a way, otherwise I learn very ineffectively and lose motivation; might as well be in need of a mindset-shift! So I hope anyone has any recommendations or advice! :)
- Crimzan
Replies
subdivision modelling is the wrong approach for surfaces that require precision or complex compound curves.
assuming you've grasped the basics of how a subdiv operator interferes with your mesh perhaps you're just trying to make the wrong things?
I know that the subdivision surface modifier subdivides each face into 4 smaller faces, and smoothes the result, which is why you place edges along the perimeter of your details to get sharpness.
I'm not really looking to make anything mechanically accurate or complex in the way of it being precise. I want to make some Props, some hard surface things for game engines, but also would like to explore more organic shapes, like modeling simple stylized characters just to get better. Tutorials I have watched always seem to be using subdivision for those, so I think it's good to get into that!
I don't know if I expressed myself in a confusing way -- just to clarify, it's not that I don't like the objects that people are modeling in their tutorials, like a Head modeling tutorial for instance. It's that they seem to be applying techniques and methods that I am simply not aware of, and don't always talk about it. One example of what I mean when I say "methods I am not aware of" are for instance methods to terminate 4 to 2 faces, 6 to 4, etc, I just very recently learned about that -- I'm talking about techniques like that, techniques that people just seem to subconsciously apply to deal with a specific issue in their model. I am not talking about general modeling functions like insetting or extruding or something like this, in case it came across that way.
But once my topology differs from theirs in a tutorial because I want to try a different but still similar shape, the loops flow differently and I don't know anymore how to deal with an issue on MY model -- which I assume is due to a lack of fundamental understanding, not particularly of the modifier itself, but of things to keep in mind when using it.
And that kills my repeated attempts at getting better at this. I don't know how to properly approach learning it, especially with my unfortuantely very particular learning style, which is why I am looking for advice how I could approach learning it or for resources that go more in-depth about just the plain basics and prepare me for issues that I might have on my model, and how to overcome them.
If there are no such resources I think one way to go about it is brute-forcing it in by following simple guides "how to get shape XYZ" and trying to memorizing them, trying to subconsciously build a repertoire of common methods to fix things, even if this hasn't proven to be particularly effective for me thus far.
have you seen this thread?
https://polycount.com/discussion/56014/how-the-f-do-i-model-this-reply-for-help-with-specific-shapes-post-attempt-before-asking
^^ That thread is a gold mine.
See also: https://polycount.com/discussion/221392/sketchbook-frank-polygon/p1
and http://wiki.polycount.com/wiki/Subdivision_Surface_Modeling
Also a key thing is to post examples of your modeling works in progress, so others can provide feedback and ideas. Either make your own Topic, or post in the thread sprunghunt linked above. Oh, and make your models easy to view.
Good luck, and let us know how it goes!
Okay! This is probably my best bet if there are no resources or anything I suppose. Also Frank Polygon's guides are probably the ones I would like to see if I can work through them. I found his Sketchbook before but hadn't read it in its entirety, but they are very well explained, even if things rarely stick in my head. Guess that's how I'm going to go about it!
So, thank you for the links and the info. Guess I need to brute-force things a bit, I'll try working through the Sketchbook and trying to re-create the examples, and then once I work on and struggle with another Subdiv model asking for feedback in the thread sounds like a solid way to learn!
I just did some more research and I found "The Pushing Points Topology Workbook" by William Vaughan. It seems to have generally great ratings, maybe this is a real nice book to check out additionally to get a good understanding of the basics. I think I'm going to get that! Did anyone else read it?
I haven't but William does know his stuff.
Highly recommended, which initially proved helpful getting too grips with a selection of subd modeling principles/concepts.
Polygonal Modeling: Basic and Advanced Techniques by Mario Russo
I know how to make *that specific* model in the tutorial, but I don't know how to make my own models, or to solve issues that I have with my own models.
Tutorials is how we learned in th bad ol' dayz. Whatever I learned came from the internet and forums like polycount. The thing I had trouble with was creating game models and textures. For that I followed an online course but that was it. That's also probably why I ended up here. I think it's really helpful to be more open and flexible about the information you find. If you want to make a ray gun for example you would look for pipes, refineries, greebles tutorials anything to help you with the components or method, and apply those techniques to your situation.
I am this kind of person who needs a lot of hand-holding and a very structured approach that really nails in basics. Can anyone relate to my situation?
What you are learning is the principles of modelling, not answers to every question you have. Again, making a car, sure you would seek vehicle tutorials, but your design may be revolutionary, like no one has seen or thought of before, you may find a fighter jet or racing yacht tutorial useful and grab techniques you can use.
So you are making models because you love it. Just about everybody on this forum is in the same boat. They love it. So try out your methods and when you think you have hit the wall, post here and ask for help.
Don't forget to have fun.
Thank you very much for your response, it was very insightful!
From reading this, it seems that I'm not particularly allowing myself to fail or struggle -- instead, I seem to be desperately looking for the "ultimate hassle-free super easy course / resource" that goes through things in a much easier difficulty to avoid the fear of failure. What you're saying makes me think that I need to try out things more and play around, see what works, and even if something doesn't -- I still gained important expierence. This will prolly be a tough ride for me to stop, because I tend to lose the fun when I'm struggling / failing (which is why I go back to methods that I know) -- so maybe I need to re-wire my thought process here and see the "failure" as nothing bad and allow myself to learn, to try things out, and to not force myself to focus strictly on achieving a specific result. Gotta learn to have fun practicing! :D
Aaah this looks cool and seems to cover really interesting topics, albeit a bit old! Unfortunately it doesn't seem to be available anymore, but I'll see if I can find a copy in another store!
I see! I got my hands on it and I'm gonna be checking it out in the future as I go!
This thread is less of a "how do you learn subdivision modeling", but more of a "How does this Crimzan guy get over his own damn issues to sit down and practice" haha :D Your responses have been very helpful and it's been giving me ideas how I can start approaching things to shift my mindset. I think I'm especially going to maybe make a little collection links to Frank Polygon's posts, his guides are to the point and really small, so the projects / tasks don't end up being an absolutely massive multi-hour monster of a complex model.
Thanks for everyone who responded and who took time reading this thread! :)