a quick test with Geometry Nodes enjoy
This is a good example of how something still on early stages still manages to show great potential.
wow greatâŚ
Is there a build available somewhere? Iâd like to try it too.
It is a bit hidden. You have to go to the download section of blender.org and scroll all the way down to âGo experimantalâ.
I have it now thanks. I checked graphicall and I didnât think of checking the official blender builder
Itâs there only from yesterday I think
@dfelinto has opened a Geometry Nodes feedback thread on devtalk.
Itâs nice to see community feedback being encouraged this early in the project.
In my previous video I had to use performance heavy Booleans operation to change forest boundaries as well as to cut through forest. After an update to Geometry Nodes branch, itâs whole another story, I can do this purely just with Weight Masks now.
Oh great. I canât wait till we can manipulate vertex attributes/weights with nodes !
Youâre right, that is highly appreciated, simple setups are easy, but setups with tens of weight maps become untidy and hard to understand very quickly.
wow very impressive forest shaping! I like the âscrumâ aproach they have to geometry nodes⌠clear short term goals and targets is a good idea. the google summer of code also has this approach and gets good results. I have avoided particle systems/hair in hope that it all gets a preset based workflow. being able to use groups in geometry nodes and transform them on a per object basis seems really fun.
Really cool stuff, and this is just the beginning of Geometry Nodes.
Once particle nodes and friends get added, no one will want to go back to that dusty old clunker known as 2.90.
âOnce particle nodes and friends get added, no one will want to go back to that dusty old clunker known as 2.90.â
@Ace_Dragon Future Blender releases can be more than proud of you
A pretty neat example.
Itâll be fun to see all the crazy stuff the community does once this is more feature complete. Hopefully the asset manager will allow easily sharing node setups too (assuming the asset manager ever gets finished).
I dig your last video with the sliced cube (https://www.youtube.com/watch?v=WkyfLYtNgcs). I was thinking, having loops could greatly simplify that kind of setup. Of course you could always instance the slicing plane a number of times, rotate them randomly and do the boolean from that new object instead, but still. Loops would be nice. If you donât mind me asking @Jacques_Lucke is that something you plan to support ? I know that question was asked before but since the project has notably changed since thenâŚ
Cheers,
Hadrien
Very early attempt for non-destructive cell fracturing. Itâs not the best way possible and really sluggish for real time editing right now youtu.be/WkyfLYtNgcs
Oops you were faster sharing the link ⌠Iâm not sure how exactly loops would make the same result as boolean operation. As far as I get it, it would make new edges on objectâs mesh, but how the interior mesh would be created then?