Can somebody explain what advantage new splatter v2 provide?

For anyone looking for release info on that node, see here :

https://experienceleague.adobe.com/en/docs/substance-3d-designer/using/release-notes/version-16-0

and here :
https://experienceleague.adobe.com/en/docs/substance-3d-designer/using/substance-graphs/nodes-reference-for-substance-graphs/node-library/texture-generators/patterns/shape-splatter-v2
 

and a bit more write up here: https://www.cgchannel.com/2026/04/adobe-releases-substance-3d-designer-16-0/
 

Dave

Dave

Jooyoung Lee  Thanks for you reply.  Poison disk and texture space  are big indeed.   Too bad we have no ready made noise nodes using them.   I don’t think it recreates  old atlas scatter .But I hardly ever used it. It gave that slightly naive look  and bad aliasing  it’s very hard to weed out. So not sure.   Seems it rather use a new Grid atlas which with new greed atlas build node does look like replacing pattern input entirely. So I agree     it’s a nice feature   but i did similar one  decades ago  already.       

. 

This is just my opinion.

1. It’s a mix between Shape Splatter v1 and Atlas Scatter. While the Pattern Input-Grid Atlas is a core feature of the node system, I still don't understand why 'Pattern Input' and 'Grid Atlas' modes coexist. Since 'Pattern Input' is essentially based on the Grid Atlas, there is no reason for both modes to exist separately.
2. Heigh-Shape Extrude is added. Feels like more advanced version of ‘conform to bg’ option.
3. Poisson distribution is the big thing. 

but problem is also exists.

1. where is Pattern distribution map?
2. Normalized workflow is not work as expected when input is ‘Image Input’ or ‘Grid Atlas’

To be honest, the newly added SDF feature is disappointing. While providing a toolkit for modular raymarching implementation is quite impressive, many feel the learning curve is just too steep.

At this rate, Houdini’s Copernicus is much easier to use.