The cookie / boolean operations seem to have the same issues I have been seeing since PRISMS 4.2 (when I started). Which is to say, missing, or extra edges, and unusable output - especially with changing inputs. I assume that I am just ignorant of their correct current improved usage, or unaware of tricks to make it behave.

Has anyone been able to make the shatter SOp work reliably?

Thanks!

-caleb

it works very well in h10, if You scale up the object at SOP level, I was able to type in 10 in number of cuts, then i tought it would be better to use several shatter nodes with different seeds, but the second shatter exploded my model, so i don't have a clue.

Hey Caleb!

Yeah, there were some improvements made to the Shatter SOP since H9, but I know what you mean, it's not perfect and certain scenarios don't perform as well as one would like. There was hope for a while a new approach to the problem using a volume approach would make it into 10.0, but alas it was not to be. I'm hopeful they'll get to it for next release…it's a constant RFE…

Cheers,

J.C.

Hi Caleb,

Yes, unfortunately even simple cases routinely fail. I continue to not understand why this isn't written in C++ but regardless if you need a good Shatter, 3DSMax or Maya have plugins that do this very robustly and reliably for 1000's of cuts on complex objects.

Maybe for 10.1 we'll see something? It's only been 10 years without it, another few months won't kill us

Cheers,

Peter B

Well, I'm glad it's not just me, but at the same time, I was hoping it was just me. I'm currently engaged again in the folly of trying to sell my current studio on the advantages of Houdini as a part of the workflow. I have implemented a volume approach to fragmenting which Ben Schrijvers described at the GDC, and it worked flawlessly within a day (the fragging part). I haven't finished the UV transfer yet, but I have a few ideas in that regard.

Thank you Ben!

I'll try it out in v10, and see how it goes.

I like the volume approach. I've been tilting against 2D manifolds in favour of the more realistic solid field representations since I started at DD in ‘93. It’s nice to see it catching on! ;-)

Anyway, Thanks, guys - what an awesome forum to get such weighty feedback so readily! Houdini is doubly great for the community, and the product.

-caleb