Please correct me if I’m wrong…
But you work primarily with still images don’t you, where your 3d work neither needs to be deformed for animation, nor be ported to a game engine.
(Don’t confuse my question with a lack of interest in SDF modeling)
Please correct me if I’m wrong…
But you work primarily with still images don’t you, where your 3d work neither needs to be deformed for animation, nor be ported to a game engine.
(Don’t confuse my question with a lack of interest in SDF modeling)
Personally, I mostly do still images, yes. SDF is also suitable for animation though. An SDF editor like Unbound offers animation and interactive game dev functionality in SDF space. Sadly, MagicaCSG doesn’t offer animation yet, but as the closely related MagicaVoxel UI / UX does offer animation, I’m hopeful.
Another important factor not mentioned is the 2 distinct types of SDF/voxel (whatever we want to call it) modeling methods. This will radically determine what you can and can’t create.
There is the non-destructive primitive method and the SDF > triangle > SDF method that isn’t non-destructive, but has warping/smoothing/surface sculpting.
There are trade-offs with both methods Vs polymodeling. There are both strengths and weaknesses. Some shapes are very simple in voxels and other shapes are far easier to do in lowcage sub-D modeling, for example.
You obviously have a lot more control of your surface in lowcage sub-D, but again, some shapes that might take a bit of time and planning in sub-D might take a minute in voxels. It will all depend on what you are creating, and for what purpose.
At the moment, the post-voxel stage is the weak spot. If you want to further work the converted polysoup into a production asset of some sort you will have to plan accordingly. It’s great for still renders and 3D printing and even something like Nanite meshes in UE.
Probably the best post-voxel solution that I’ve tried (and I’ve tried many) is using Houdini.
Skull bomb 3D icon. ![]()
![]()
I’ve made use of MagicaCSG’s Avoid blend mode to embed the skull into the bomb. The skull is a grouped Boolean, to avoid each skull component being blended separately into the bomb.
Scene demo video:
It’s always fun to have look here… ![]()
I asked this a while ago, it’s nice that this would now being implemented.
Because be able to using Text will make MagicaCSG interesting for a lot of graphic designers who miss 3D Text in their graphics programs.
I think that something very much needed in Magica is a way of drawing shapes for extrusion. It is something that will make modelling complex shapes much more easy.
![]()
That’d be great indeed. I do know that Ephtracy, the MagicaCSG dev, is planning a way to sculpt in MagicaCSG, but it requires a rewrite of a lot of things, which I think has already started. Don’t know when the sculpting functionality will be introduced though.
I don’t know if this really fits, but I thought that MagicaCSG “Sweep”
mode could offer something in the direction of drawing shapes
Yes, and when the version that supports fonts is released, you can make use of glyphs / Wingdings to insert all kinds of shapes. ![]()
Ephtracy told me that the font SDF shapes are taxing in terms of necessary computing power though, in their current brute force implementation.
A few more preview images:
I know when I asked for a Text Tool on Patreon back then,
that was exactly why he objected. Anyway, I’m excited to see
how it turns out.
This is very impressive considering the limitations of live-bools. Very cool characters. ![]()
Ladies and gentlemen… The singing cactus! ![]()
![]()
![]()
Inspired by a HyunSong concept drawing.
WIPs:
3D scene video:

Another text teaser. ![]()
If I understood it correctly, there will be one fixed sans-type font with a number of characters, for the time being:
Re-rendered that stinging singer with Cycles…
Mikko Sinisalo created this head entirely in MagicaCSG:
There’s also a video showing the model inside MagicaCSG. You can find that in the MagicaVoxel Discord ➔ MagicaCSG channel. The MP4 video file was too large to post here.
He really is a dedicated MCSG user. This must be torture to do in live-blend prims.
Great work though.
I’m going to post this on the Modeler beta forums to showcase the power of swept profiles.