Hi,
My Introduction to sculpting was with the "CgCookie Fundamentals of Digital Sculpting with Blender 2.8" course, since then i tried to sculpt a couple of things and it seems that the timelapse portion of that course and the simplicity of that character gave me the impression that DynTopo should be applied after the very first initial blockout which in my mind consists of JUST adding the main primitive meshes and then just play with the grab brush, i literally never used the Draw, Clay, Crease, etc.. brushes without Dynamic Topology, and just recently i discovered this video:
How to sculpt BETTER Base Mesh | Blender 2.8 Tutorial | VeryHotShark
where he talks exactly about this issue,
but what really made me post here is this video in the official Blender channel:
Speed Sculpting Live Session - Julien Kaspar
he went from this:
to this:
Without using Dynamic topology at all, but he used the Remesh feature a few times, which bring me to the final "query", can someone breaks down for me the Remesh feature plase, i know it basically subdivide the mesh to give you more details, but i don't quite get its workflow, nor why is is "better" than the SubSurf modifier for sculpting, the only difference that i noticed after doing some experiments:
Is thatthe Remesh never create N-gons, which is maybe ideal for sculpting ?
So yeah,
i would really appreciate it if someone explains this a bit better for me, cause all the tutorials i watched keep saying that its a "personal preference", but i honestly don't like that, being very "rigid" in art is not a good thing, but at the same time i believe that beginners should stick to a certain limit and standards to avoid bad habits, i think this is one of the main issues of online learning in general too, like if i was in a class with Kent am almost 100% sure that this issue would've been solved within the very first few sessions, a teacher would just point out and say "you're using DynTopo too early" and show you how to do it, but because you're learning alone at home, you are really prone to building bad habits and that's what am trying to avoid.
Thank you!
Hello alaslipknot! You pose a great question - one that I'm sure confuses a lot of people getting a grasp on digital sculpting. I will try to answer this as fully and clearly as I can.
First I have to start by saying it really does come down to personal artist / workflow preference. The question of "am I applying dyntopo too early" can only be answered by you. For me I usually start with dyntopo from the beginning and never use remesh, so technically you can't be switching to dyntopo too early. But as you can see Julien uses remesh exclusively in his video. It's entirely valid to use 100% dyntopo or 100% remesh or 50/50 of each.
BUT I won't leave you with that only. Instead I will try to outline the Pros and Cons of Remesh and Dyntopo for you to decide which fits your preferred workflow.
PROS
The remesh workflow is the newest workflow added with Blender 2.8x. It comes from Zbrush's feature that projects applies a universally consistent topology across an existing shape (remember the Spiderman web illustration). This is beneficial because it "resets" a shape with wacky topology that couldn't be sculpted well any further.
A lot of sculptors like this approach because they can boldly move around simple shapes, remesh, continue adjusting, remesh again, so on and so on. Remesh also allows separate meshes to be "welded" together with a simple click. This is how Julien places a bunch of individual simple shapes at the beginning, focuses on each piece by itself until they're all ready to be combined into a whole.
This workflow is a lot like how I learned to sculpt with physical clay: In the case of a humanoid sculpture I'd roll a log for the left leg, roll a log for the right leg, an elongated sphere for torso, logs for each arm, a ball for the head. This is the preferred method with clay because you can't generate volume out of thin air. Volume is bound by reality's physical limitations and thus clay needs to be added and subtracted accordingly.
Dynamic topology is a departure from reality because volume can be added and subtracted in a way that can't happen in the real world.
CONS
The biggest con of REMESH is also it's biggest pro: Universally consistent topology. This is only a pro at the beginning of the workflow and becomes and increasing problem with the more detail that is needed and if you want to preserve previous details. Here's the main two problems:
1. Remesh will diminish any strong details you've already sculpted. It's "destructive" in that way. Let's say at your tenth remesh you sculpt a smile that's 100% perfect on your character's face. Then you realize you need to remesh to achieve other smaller details. You click the remesh button and your lips look a little different than they did before because their shape was approximately projected to new topology.
2. As you want to add fine details (like wrinkles, cracks or skin pores, etc) you need to remesh your whole object at a high density, even if the detail is very small. This can result in a tremendously inefficient model that's too slow to work with. Practically speaking this commonly limits the remesh workflow to stylized characters/creatures with fewer fine details (like Julien's monkey). This also means remesh is bad for highly-detailed realistic sculptures.
PROS
Dynamic topology to accomplish details and shapes only where you need it, no more, no less. This method is more "magical" in the digital sense of being able to work in a way that's not possible in reality. You can generate digital clay out of nothing. In a way it's ultimate freedom for a digital sculptor with almost no limitations when used properly.
In the case of generating basic shapes to begin a sculpture, dyntopo allows us to easily pull any shape we want out of a simple sphere, generating new volume as we need it.
In the case of details you can easily add very fine *localized* details without affecting your entire mesh. This often leads to much more efficient sculptures and much more forgiving workflow.
CONS
Dyntopo tools are a bit weirder to get used to. If used improperly it's easy to feel like you're losing control of your sculpture, causing more damage than improvement. This is the #1 issue I hear from new and intermediate sculptors.
Also dyntopo is a bit more CPU-intensive. A 500K dyntopo sculpt will perform more slowly than a 500K remesh sculpt.
So in the end it comes down to what you are most comfortable using. For me, it's dyntopo for 99% sculpting (and multires for detail sculpting). Rarely I will use remesh not because it's a bad workflow. But rather because it doesn't improve anything about my dyntopo-based workflow (remember dyntopo has been around way longer and I'm way more experienced with it).
My suggestion is to practice using both and see which one suits you best. Both are equally valid approaches.
Thanks a lot Kent!
based on what you said, don't you think it would've been better for a Beginner to start without DynTopo ? considering that there is no way we're going to be creating super detailed AAA-quality models within the first few months ?
The way am looking at it, is kinda similar to choosing Blender (2.7x) or Maya if you are a complete beginner and just want to test 3D modeling, old Blender will just create an extra layer of difficulty (the "weird" UI compared to every other software).
The cons you stated for DynTopo are 100% real, most of the time i start feeling some "disconnection" between me and the level of control i have on the sculpture, and since like i said above, for early sculptures i don't think the cons of the Remesh (can't handle heavy objects) is going to affect me too much since i don't even have the skills to create those heavy detailed objects.
Anyways,
I'll start practicing more without DynTopo and see where that takes me.
Thanks again!
based on what you said, don't you think it would've been better for a Beginner to start without DynTopo ?
No I don't think so. I still think dyntopo is the better approach for the shark sculpture. Keep in mind it's primarily one large object, not a collection of individual pieces like Julien's monkey (biped). There's no point to individual focus on various appendages when there aren't appendages. Remesh would only serve to subdivide globally which isn't a benefit in this case. Instead it's better to add localized detail for tight details like the face on such a large, smooth, contiguous body.
In step with this, I also consider the large lump of a shark to be a simpler beginner project than a multi-appendage biped or creature; less parts to worry about.
Remember that workflows are largely relative. If dyntopo is more cumbersome to you then by all means stick with remesh! But I and a lot of beginners get along great with dyntopo.
I'll start practicing more without DynTopo and see where that takes me.
That's the spirit 👌
