While cleaning up my system I came across this old video by another modeler. There are many ways to introduce new flow and geometry into your mesh. Most likely, Triangles and nGons will be introduced when you use Knife, Loop Cut, and many other tools. However, if you start with a Cube and plan out your extrude(s) and spin edge carefully you can have a mesh that is all quads. This three operations: Extrude, Spin and Tweak seems to be sufficient in most cases. This type of mirror editing seems to work flawlessly in LightWave, but Blender? Nope. It doesn’t work perfectly and strange things do occur. As of this writing, Blender doesn’t do well with this Combo operation in a single-mirror editing mesh.
There are a lot to talk about down the road, but in the meantime, I want to showcase one of Xuan’s models. Visit the other post for some background.
Modeling what you see is what I called “Observational Modeling.” I call it such because I happen to do 2d and I see a correlation. Drawing what you see is “Observational Drawing.” From this, it’s not hard to see that there’s also Observational Sculpting. The key word is “observational.”
There are many ways to approach observational modeling. It’s mostly done in ZBrush sculpting (for characters/figures), and if in traditional software such as Blender, it’s done with the help of mirror and basic rig setup. But in Wings3D, it forces you to think creatively, and in this example, Xuan thought outside the box to accomplish this piece.
When Xuan sent me these shots and said that it’s done in Wings3D, and that what I’m seeing is only 95% completed, I had a hard time believing it. If it was ZBrush or sculpting then I can understand, but 100% Wings3D? No bones, no mirror, no blueprint, and just that toy as a reference? I was speechless. My first thought was “3D scan” but I can reassure you that it’s none of that. Of course, anything is possible if you put your mind to it, but if you’re going to do observational modeling (modeling what you see) of this level and details when it comes to figure/character, then there got to be a trick. And yes, there is. Xuan will share that later on.
How do you think he did it? I’ll show more shots once I have access to the 3d file.
On and off, I have been trying to solve this strange anomaly, and haven’t found a solution yet. When writing tutorials, I want to have things look consistent throughout, so this is a preparation for future modeling tutorials. I’m really scratching my head with this one. I played with all the options and can never get the Right image to look exactly like the one on the Left. Now, of course I can just take a screenshot of the Left, but that is inconvenient and has a few limitations.
Free model. Click on the image to download. Make sure to read the text file that comes with it. Free for personal and commercial use. No string attached.
It’s late here and I had another window opened. I meant to close that window but accidentally closed Blender without saving the work I put on modeling the gums for the teeth. Luckily “File -> Recover -> Auto Save…” solved the problem. Make sure to always have “Auto Save Temporary Files” checked in the Preferences.
This is one of the reasons why I mute the CMD+Q (for quit) so that I don’t accidentally close Blender. Just this time I haven’t muted it since I had no plan to model anything in Blender yet. Right now just experimenting.
This head could be done much quicker with Sculpting, but I decided to go with Box. There’s something about pushing points around with the Mouse, the thrill you get… With Sculpting you use the tablet, with pushing points, you use the Mouse.
It’s interesting what you’ll encounter as you move away from character modeling.
I wanted to straighten all the vertices diagonally. The most common way to go about doing this is “S” (for scale) and then Axis [X, Y or Z] and then “0.” But to achieve this diagonally, you need to reposition the Gizmos. After a bit of thinking, my approach is this: Set the Transformation orientation to “Normal.” And Transform Pivot Point to “Active Element.” Select all the edges that you want to straighten… make sure not to select the straight edge first. The edge that is straight should be the last to be selected, so that it becomes the active element, and all other edges will line up.
This will work out perfectly if your active element is actually “perfectly straight.” But what if it’s not? You can manually reposition it so that to the human eyes it looks straight, but you can’t be certain that it is.
Luckily, I had the outer ring untouched the whole time. I used that edge as the active element. And voila!
Just think ahead. Make a duplicate of an edge from the start and leave it there in case you might need to use it as an active element to straighten your vertices down the road. If you know of another way, I would love to hear it in the comment section.
LoopTools is a free addon that comes with Blender, disabled by default. I believe the reason why it’s not on by default is because if all addons were enabled, it would clutter the app. For example, I’m not into animation, so I don’t need all the extra features for animation in Blender enabled by default. So… if you’re a modeler, LoopTools is a very useful and handy addon to speed up your modeling. Once you get into shading, Node Wrangler will be your friend. Know what you do and enable the addons accordingly.
Here’s a LoopTools’ Curve feature. By selecting the first, mid and last points, I was able to create a curve that matched the blueprint.
One of my future projects is called “Observational Modeling.” A PDF with video demonstrations on modeling what you see, with an emphasis on observing and simplifying. Taking this idea from “Observational Drawing.”
Modeling things from your imagination, or just making things up as you go isn’t the best way to grow your modeling skill. It’s also hard for others (and you!) to judge your skill or progress. But if you model an object or things that exist in the real world, then others will have something to compare to. Think of it this way: If you model or sculpt a creature, you can get away with it with lots of details. But if you try to model or sculpt a well known person, no amount of details can cover up your inability to capture the likeness of that person.
If your goal is to become a character modeler, then by all means, start with character modeling. However, if your goal is to improve and grow as a modeler in general, then I recommend that you start with hard surface. As for me, I started off with creature/character, and now struggling a bit with hard surface. Both have its own challenges to overcome.
Here’s a model I did earlier. The reference looks simple. So I thought! But don’t underestimate. A reference of the thing you want to model might look simple, but you won’t know the challenges you’ll face until you begin modeling it in the 3d viewport. Since this is based on a reference, I can look at my model and see which area I need to work on/improve. If this was purely from imagination, there’s no way to know where I struggle or where the challenge lies. Hence, observational modeling is the way to learn, grow and improve.
The Mirror modifier can save a lot of time, but in Edit Mode, how do you model the top and not have to repeat it for the bottom? And is that even possible? If I repeat the same for the bottom it can be tedious. Perhaps there’s another way, but that would mean I have to start all over? I don’t have enough experience with hard surface to answer this. This is an example of one of the challenges I faced with something that looks simple.
I came across a Maya tutorial and was inspired to do a test model of a weight plate.
The math is basic. I’m planning to convert that tutorial into Blender for others to follow, but there’s one problem I’m trying to work out before I get to writing it. Two modifiers were used: Mirror and Array (in that specific order). The problem? Mirror has clipping to prevent vertices from crossing (blue), Array has merge but it doesn’t work like Mirror (Red). Not sure if it’s even possible to have Array clips like Mirror. It would definitely save a lot of time from having to go about fixing it in a round about way.
EDIT: After playing around with the Array modifier, it seems like clipping is not possible. Merge first and last is the closest one can get.
This is my first time doing Isometric. I have come to realize that this is a great way to practice for the following reasons:
- It doesn’t have to be 100 accurate. You can be casual and learn in the process of modeling. The image you see above is based on an actual basement. Even though it’s based on it, it doesn’t have to look exactly like it.
- You don’t have to model the entire environment. Just the corner and what you want to communicate.
- You can experiment with rendering or save that for later and just focus on the modeling.
- Modeling non-human forces you to know the tool. I find myself looking up for tutorials on curves and setting the Origin.
I’m more into character/creature modeling than anything else, so this is out of my comfort zone. However, Isometric is a great way to explore it! If you want to get started, the first thing you must learn is setting up the camera for Isometric. Also, knowing how to control the camera is good knowledge to have.
Last March I wrote a short post to talk about how sculpting the hand can be a challenge without the ability to quickly pose the sculpt. You can read it here. Today I found out that 2.81 has it and you have no idea how excited I am. Someone posted a video demonstrating that brush here. Earlier I reopened that unfinished hand sculpt file to test out the brush and it’s amazing.
This one below is another test to see how convenient it is with modeling. All I can say is WOW. You can tab in and out of “Edit Mode” to “Sculpt Mode.” Edit Mode to model, Sculpt Mode to use the Pose Brush without damaging or make any changes to your mesh poly count! I’m impressed.
This Pose Brush is making Box modeling fun again.
Back when I was starting out with 3D modeling, there wasn’t that much resources to help me in my endeavor. But those that are starting out today are very fortunate to have a lot of resources and aids to help them progress much faster in their studies.
Digital Emily is a project that is based on an actual person (3d scanned). Notice that the left and right side of the face are not identical (asymmetrical). Very often when we model from scratch, it’s symmetrical (mirrored) to speed up the process, but in real life very few people have symmetrical faces. Do an image search on “photo symmetrical celebrities” and you’ll see how strange and bizarre well known celebrities look when you turn their asymmetrical into symmetrical. This is something to consider when seeking to capture the likeness of a well known person.
This one is from MB-Lab, and it’s base is not meant to be hyper realistic. However, it’s a good base to start with! This is a plugin for Blender that can create a full figure, with many parameters to change the look of the character. It’s also poseable.
- Human Base Mesh by Vidar
- Base Meshes @ polycount
- Base Mesh (Male) by angelaxiotis
- Default Human – Base Mesh by Alin Bolcas
- Stylized Female Modular Base Mesh by Toto Ditrani
- Stylized Male Modular Base Mesh by Toto Ditrani
- Style Female Body Basemesh by Ahyo (Doomlord)
- Head – Stylized Female Base Mesh by Walter Leon
- Free ZBrush Female Base Mesh
- Female base mesh by cgfarmer
- Poseable Female Base Mesh by sculptor_dad
- Chibi Style Body Base Mesh by the_gakabox
- Male BaseMesh by Guillaume Mahieu
- Bust Base Mesh for Likeness Sculpting by EA Yee
- Base Mesh optimized for Facial Rigging by Loïc Pinsard
- Human Anatomy Study: ZTL + OBJ by Tom Newbury
This post will be updated with more links as I come across them.