6 key principles for 3D modeling

learning 3d modeling can be pretty challenging especially when there are a million different do's and dont's online and it's very easy to get lost in the weeds of tools and topology thankfully there are just a few different main concepts that when you really focus on them no matter what project you're working on or what skill level you're at will instantly help improve your results I'm Jonathan Lambeau with CG cookie calm and today we're gonna look at six key modeling principles and how to apply them first up is form the first thing that you want to think about when modeling is the overall shape of what it is that you're trying to make announced might seem obvious but it's often the most tricky part to get right now it might be a detail on a character or a part of a car but there are some shapes that are harder to make than others and the thing to remember here is that complex shapes are pretty much just combinations of simple shapes now of course it's not always that easy but even if you just block something out and you use spheres and cubes and just minor mesh edits and smash those together to make the shape of what it is they're trying to make that might not be usable in production but it'll at least help you get a good sense of the underlying structure of what it is that you're trying to make so if you get stuck try to do that first and then come back to the final like mesh topology edit next identify the most defining features first and then fill in the gaps later so whether that's a jaw line on a character and the ring around the eyes or some sharp edges or defining curves whatever is the most defining feature of your object make that first and then fill in everything later and that will help you make sure that you get the best form possible of course use a reference as you go to check your angles and your proportions and all of that stuff but be sure not to rely too much on modeling sheets because they can be really really helpful to start out with but what often happens especially with beginners is that you'll jump from side view to front view and back again maybe top view but when you pop out of that and rotate around in perspective things start to look a little bit weird and the reason for that is just relying too much on those orthographic perspectives and just lining up each individual vertex and being super careful with that but not really understanding the no shape of what it is that you're making and so if you find that to be the case where things look good from front inside view but then you go into orthographic and it just doesn't look appealing then ditch the modeling sheets for a little bit you can move them off to the side or use them in a different window but don't lay them directly on top of the model and this will help you get a much better understanding of the form of your object as you're making it now the last tip about form is that if you start with very little detail as little as possible whether you're poly modeling or sculpting and only add more detail once that initial shape looks as good as it can get that will help your models to look clean and crisp so if you find your models to look like lumpy or things get a little bit like mushy or there's just too much bumps or creases around things that shouldn't be there then you're likely adding too much detail too much topology too quickly so you've got your overall form down and now you're ready to push your model to the next level by adding some detail now it's really important to understand exactly what details to make and where to place them of course there's going to be some pretty hard limits if you have a polygon count limit or a certain rendering budget but even if you have a completely unlimited poly count say you're working for animation instead of a game and it can take as long as you want to render you still need to understand the levels of detail because that's gonna help your objects just read more clearly and just look better but it'll also help in the modeling process because then you can add details in a way that makes it a little bit easier instead of harder three artists Neal Blevins has a really legendary blog post all about primary secondary and tertiary shapes and I'll link to that in the description below it's well worth the read but in short you should have areas of very large details areas of medium details and areas of very small details and a good mix between them and of course areas where it's highly concentrated and areas where things are mostly left blank and given room to breathe the way you want to build this out when modeling is to work in passes and this is where the idea really comes into play because you want to make all of your big details first and once those all look good then move on to medium and then to small so this is hopefully sounds familiar but it's a good point to drive home because if you only focus on one area at a time and detail it and then move on to the next one and detail that all the way down and then move on to the next one and you know you just work kind of linearly instead of impasses then what happens is by the time you get to the end things don't really match as they should so by working all the way around your model just very quickly switching between different areas and only focusing on big first and then once the entire model has all of the big details move on to the medium until the entire model is done and then move on to the small it'll help things to look a lot more harmonious and if you get stuck on one area just jump to another area temporarily and then come back when you're ready to problem-solve again number three is scale try to model to real-world scale whenever possible of course the differences will be pretty subtle at first when it comes to how lights interact with the scene and how simulations run and how procedural textures are applied and all that stuff it will make a difference with those but most importantly it'll help you to be consistent between objects in between scenes so that you can easily work no matter what file you're in or no matter what you're appending or linking things to you'll help everything just to look just right the things that our pretty obvious of course are the length and width and height of things and I would definitely recommend using reference to check for those but another reason that you want to use reference is to check the thickness of stuff in the amount of bevels because you'd be pretty surprised at how we think something looks is different from how it actually does look and a good example of this is a table if you just kind of eyeball it and let's say you look at the dimensions online and you find the length width and height and you model it to exactly those dimensions it can still look off and in this example we've just made the legs a bit too thick and added a bit too much bevel and it just makes the whole thing look smaller than it should even though it's technically the right size so if you put this in a room it'll it'll just look off so you definitely want to use a reference to check all of these different things and make sure that they're all in proportion to each other and in proportion to the real world scale principle number four is adaptation by this I just mean keeping your object as easy to modify as possible now this could mean non-destructive modeling where you use modifiers and tools like that to make changes to your mesh that are not permanent so that you can always go back and change the original starting position or different forms without affecting all the details that you've added later on so for example like a solidify modifier makes it easy to take a very simple two-dimensional shape and give it a three-dimensional complexity without you having to then deal with all of those vertices in edit mode so non-destructive modeling is a great way to go but also just keeping the mesh as simple and less dense as possible will help keep your edits to be very very quick without having to do a bunch of selection gymnastics in order to get your desired result so of course keeping things simple keeping it non-destructive is one part of adaptability but it also needs to adapt to new positions when animated so this can mean anything from making sure that your topology supports whatever rig it's going to need to have whether that's a joint in the arm or something like that do you also make sure that your orientation send pivot points are properly set up and facing the right direction so that's very easy to animate principle number five is reuse I don't think anybody likes doing the same work twice and they certainly don't like doing it 300 times or a thousand times so reusing as much of your meshes and models as possible will help you save a ton of time so this could mean using the mirror modifier of course just to only model half of an object or it could mean using the array modifier things like that but could also just mean duplicating parts of your mesh and using them as a starting point for other parts of your mesh or other models so that you don't have to start from scratch every time now another good thing to do is to use alt D instead of shift D to make an instance instead of a duplicate so you can have multiples of the same object with different like scales and rotations but still all have linked geometry so that when you make a change to one it makes a change to another you could also use particles and do plaits there's lots of different options for this now if you don't want things to look obviously duplicated of course a good thing to do is to change the rotation and scale either in object mode if you're using objects or with particles and you can also vary the UVs a little bit if you're using the array modifier or just use the object or the random socket in the object info node in the materials to vary the materials either their color or their UV offset or something like that just so that it doesn't look quite so obviously duplicated now this can save you a lot of time and it will help your objects to look much more detailed without you having to do a ton of extra work number 6 is surface quality this is how your object looks when rendered now because of how rendering works how you model your object and the direction that the geometry flows in has a big impact on how the light interacts with the surface so be on the lookout for pinches and bumps and things like that that aren't really part of the actual model or should it be this is a sign that your topology isn't matching up with the form of the object so when subdivision surface modeling use a loop of quads or round any sharp edges or important details only use end Gon's on flat shaded surfaces avoid triangles at first until you have a good reason to use them and make sure that your normals are pointing in the right direction be really careful where you place poles which are vertices of five or more connecting edges if you're not familiar with those topics or you're skeptical as to why there are really good rules of thumb I have a super thorough course called the mesh modeling bootcamp that'll link to in the description below and that'll walk you through all of it with tons of examples if you're just starting out we also have fundamentals courses all about the basics of the interface and modeling sculpting lighting rendering all of that good stuff but whether you're a beginner or a pro keeping these six modeling principles in mind will help to improve your work and avoid some of the modeling headache that can typically happen when you try to make something new thanks for watching hope you found this helpful and go ahead and click Subscribe if you want to see more videos like this as looking I'll break up at night so much [Music]

Channel: CG Cookie

Views: 1,028,020

Keywords: blender modeling, blender 2.8 modeling, blender tutorial, 3d modeling, blender model, modeling tutorial, beginner, modeling, blender

Id: OVbIOHAI3iY

Length: 11min 12sec (672 seconds)

Published: Tue Jun 04 2019