So far, the only similar reports I’ve received were caused by Avast incorrectly flagging meshoptimizer.dll as a threat. The DLL is built directly from the official meshoptimizer sources on GitHub, so unfortunately I’m not sure why some antivirus products are detecting it.
A few users were able to work around the issue by adding the DLL (or the SDF Modeler folder) as an exception in their antivirus software. If you’re using Avast or another security product, it might be worth checking whether the file has been quarantined or blocked.
At the moment there’s not much I can do on my end beyond investigating alternatives for future releases.
sounds like a technical issue when the file is empty. It would help if you could join our Discord server for troubleshooting. https://discord.gg/U7yxeqJBfW
There’s currently no mirror option for groups. As a workaround, you can duplicate the group and flip it along the desired axis, then use a negative coordinate to mirror it.
Alternatively, if you don’t need blending, you can duplicate the entire layer and flip the instance instead. That way, both versions will stay in sync when you make changes.
I’ve re-uploaded the latest release. There were some issues when parsing the system’s font directories that could prevent the application from opening. If you ran into this, please give it another try. Otherwise, feel free to reach out on Discord https://discord.gg/U7yxeqJBfW
Many people have asked if there's a way to support the project. If SDF Modeler has brought you some joy or helped you in your personal or professional work, you can now choose to pay whatever you want, as often as you like. Of course, it is and will remain completely free to download and use.
Thank you all for the support and kind words over the years!
Just to mention it here as well for reference. You can snap onto any surface by dragging the center of the transform gizmo while snapping mode is enabled.
you can adjust the resolution per layer. There’s also a whole section about layers in the manual that goes into more detail and has some helpful tips :)
OBJ export isn’t a great fit for the current state of the mesh export. The format doesn’t support vertex colors (unlike PLY), which means important data would be lost. Additionally, the generated meshes are still quite dense and not clean enough for most 3D applications without some form of post-processing.
STL also doesn’t support vertex colors, but in that case I assume most people are using it for 3D printing and applying color after printing.
So before adding more export formats, I’d prefer to further improve the mesh quality and overall usability of the export.
Glad to hear it feels intuitive and useful for you, that's exactly what I’m aiming for. And it's great to know it's become part of your regular workflow.
Thanks so much for the kind words. I’m really glad you’re enjoying it! Hopefully simpler and smoother workflows will emerge over time as things evolve :)
This might happen due to your GPU specs. SDF Modeler needs at least 2 GB of VRAM (preferably more). It would probably be easier to troubleshoot this on the SDF Modeler Discord: https://discord.gg/U7yxeqJBfW
Transparency is much cheaper for reference images since they’re rendered using a traditional raster pass. Making the shapes themselves transparent would be significantly more expensive with SDF Modeler’s custom renderer, which is why I’m currently trying to avoid that.
There is an option in the path tracer to render an object as transparent, but that’s probably not what you’re referring to. At the moment, there’s no way to make the viewport material itself transparent.
However, if your goal is to overlay reference images with semi-transparent shapes, you can place the reference image in front and adjust the image’s transparency instead.
At the moment, the best option is the "Right Mouse" navigation mode. It doesn't rely on the middle mouse button, so it should work reasonably well with a touchpad.
I do want to look into better touchpad support in the future, but it's a bit difficult to prioritize right now.
Select a point and press the + button to insert a new point after it or use the Duplicate button to copy the currently selected point (also available via Ctrl + D / Cmd + D).
Thanks a lot, really glad you like it! I'll probably look into support options for SDF Modeler in the future, but for now you're very welcome to join the Discord — https://discord.gg/U7yxeqJBfW
Always happy to have more people around for testing and feedback :)
Thanks a lot for the feedback! Rotation snapping is definitely planned, with angle increments to make clean ortho orientations quicker and easier. It will complement the reset rotation feature nicely.
Thanks! I’ll most likely open-source it if I ever grow tired of it, but for now I’m really enjoying working on it. I’m honestly pretty bad at working in teams, so for both your sanity and mine I prefer to work solo :D
I’ll add more modifiers as I go, but usually I’d just use a spline with two control points. If you set the curvature to 0, you’ll get a straight linear falloff between them.
I’m planning to add a template/prefab system in the future that will make it possible to reuse shapes across different models. That said, it will likely be independent of the grouping feature.
Thanks a lot for the kind words, glad you’re enjoying it! You’re absolutely right about non-uniform scaling on groups — that’s not supported at the moment because of how the system handles transformations internally. Once you mix non-uniform scaling with rotations in the hierarchy, the math turns into shears and other distortions, and the current SDF pipeline isn’t set up to handle that cleanly.
It’s definitely a technical limitation rather than an oversight. Appreciate you pointing it out, and who knows — maybe one day I’ll find a neat way around it.
Thanks a lot for the kind words! This request has actually come up quite a bit recently — the idea of entering a transformation state directly (without having to click on the gizmo first). I’ll see what I can do about bringing something like that in.
Ah, got it — thanks for clarifying! I was thinking in a different direction. That's not something SDF Modeler is really focusing on right now, since the main goal is interactive modeling and mesh export. But I do see the artistic use case you have in mind, and it could be interesting to revisit later once the core feature set is more solid.
A lot of people ask about some kind of “shader export” they could plug into Shadertoy or similar. The thing is, SDF Modeler doesn’t just brute-force raymarch — it uses a more sophisticated setup under the hood. That’s why it runs fast even with lots of shapes.
So simply exporting “passes” or a GLSL shader wouldn’t really reproduce what the renderer does here, and in practice would likely be far too slow.
