... More tutorials are available in the documentation.
By default, it works best with non-organic shapes, but itās possible to add custom nodes and extend the functionality or use your premade organic meshes as asset parts.
It can also be used to compose premade asset parts in random configurations.
Main features:
Possibility to create various types of asset generators
Runtime asset generation support (4.25+)
30 example asset generators included (+ Example how-to maps)
Intuitive node-based system
Easy spline shape-based mesh generation nodes
A large set of exposed parameters for mesh generation rules
Custom parameter system with formulas, bool conditions, and more
Procedural surface-aligning mesh decals with atlasing support
HDRI lighting setup with many customization options
Set of example materials & textures included
Angle-based mesh smoothing (+ All Smooth & All Flat options)
For now itās designed for in-editor only, but Iāll investigate this and maybe add runtime generation support with some future update, if nothing serious stops me. Technically it should be possible to generate it at runtime after altering some blueprints, but it might be tricky, e.g. generated asset is a Procedural Mesh Component (that you can convert to Static Mesh) that sits on (0,0,0) location by design. This also could be solved with some modifications, but there are many other possible roadblocks for this - e.g. how to control the generation parameters, how to drag the nodes around workspace at runtime, etc.
(I found it inconsistent, sometimes Google Docs wonāt display the Document Outline bar, but I think itās good now. Just wanted to make sure, since itās a lot easier to navigate using the outline)
This looks amazing. I would buy it, but not entirely sure Iām competent enough to use it. Could you perhaps make a āliteā version with a max of 3 parameters or something? I really love some of the examples youāve made though.
Thank you, thatās a good idea.
Removing most of the parameters would be hard to not cripple the functionality, but I could simplify the whole system by removing some of the more specialized features like different UV mapping methods, auto vert smoothing or surface-aligning mesh decals. Maybe concentrating only on simpler low poly flat shaded assets (e.g. like the included Bucket Generator) would help to create a Lite version, with decreased price.
The only thing that Iād need to figure out is to how to manage the āupgradeā process for customers. So if someone decides to upgrade from Lite version, then only the difference in price would be paid. But as far as I know, thereās no mechanism on the Marketplace for thatā¦
ā¦ So Iāll think about that.
For now please check the video tutorials and the documentation to see how the workflow looks like, that should give you some insight. I can also help you directly via support email.
Thank you, thatās very nice to hear! I hope itāll be useful for you. If you make any new asset with it, Iād be grateful if you could send me some screenshots of it - Iām just curious
Smart Spline Generator is a separate product not included in PAC, but itās interesting that you ask about this - Iāve thought about adding it (integrating with PAC) but it was too much.
PAC has some very similar logic for surface aligning decals (similar surface-align algorithm to the one in SSG, but a bit different, tailored for decal placement), but thatās the only thing that āleftā from my initial idea about SSG integration.
Iām not familiar with this VEA Games asset, but it looks like a different product - SSG is for surface-aligning asset generation like vines, cables, ropes, fences, etc., I donāt see any surface align feature in the Spline Mesh Solutions, but maybe Iām missing something. As far as I can see, both these tools are designed for something different - itās hard for me to do a proper comparison though, since I donāt own this asset.
Sure, Iād love to send some screenshots/movies of the stuff Iām doing, especially if it helps you promote PAC! Iāll try to do that for my AAA quality assets I woke up this morning really excited to use this, Iām really enjoying it so far. So many possibilities, I need to check in detail how things such as the potion label generation works.
Iād also say performance is really good for what itās doing - I added two highly detailed armchair static meshes (25k triangles each - for reference, some of the character models I have from the marketplace are like 13k) to various points of the potion generator just to see what happened - it worked pretty well (took around eight seconds to randomize). So you can get away with using two or three high poly meshes with materials, while also generating all the other random meshes, nodes and stuff, which is really good. Of course, when I added eight of those armchairs to the fence generator I got freezing at intermittent intervals for long periods - so I wouldnāt recommend that Also when it was frozen, it only reached 35% of CPU usage for some reason - maybe some room for optimisation there?
By the way, sometimes FinalAsset_BP just disappears from the viewport if its moved accidentally - is there a way to get it back without reopening the level? Can the FinalAsset_BP be used directly somehow or must we always click āCreate Static Meshā for placement in other projects? Iām guessing its better to keep the generator levels in this project (which is really nicely setup by the way). Now Iām suddenly very curious about āprocedural placementā haha.
Thanks for the info on the spline generator too, Iām really interested in trying the surface decal stuff next. Only scratched the surface so far!
Thank you for the info! Itās good to know that it can handle it, even if itās not designed for such heavy asset parts. Please keep in mind that vertex-heavy static mesh nodes can take a while to process, because it needs to iterate over each vert at one point. This might be improved in future if Epic provides a BP node for filling arrays with X amount of objects at once, Iāll probably make a request for that soon.
Iāve tried to lock the main FinalAsset actor from moving (you can see that it displays the red icon when hovering over the transform gizmo) but even if itās not moved, it still disappears for some reason if user tries to move it anyway. The best way to avoid that is to click on it only when converting to static mesh (thatās the only reason to click) and try to not touch the transform gizmo.
If this will keep happening to you, Iāll try to figure out some fix or workaround again.
Yes, itās better to export the generated meshes e.g. as .FBX or migrate them to your goal project as Static Mesh. Technically, the final mesh before converting could be used in game (itās a Procedural Mesh Component) and thatās one way to implement in-game asset generation in future, by the wayā¦ But saving it is a regular static mesh is more convenient, since you can save it to your project folder independently, or even export it to your 3D app for edits. Regular Static Meshes are also easier to integrate with other UE features.
Yes, procedural placement can be very powerful and time saving. To learn more about the PAC methods, itās worth to check the included āMaps/Simple Examplesā maps, along with the documentation. In case of any questions, feel free to ask here or on the support email and Iāll try to help.
Iām really glad that you like it!
Also when it was frozen, it only reached 35% of CPU usage for some reason - maybe some room for optimisation there?