Print

[WorldWideZ]

is there any best practice for making dynamic clothes for characters for the hero art pipeline? coming at it from a 3dsmax point of view , are cloth simulations a good idea (even possible?)

[PN-Dwight]

I remember from back in the old day that HeroEngine has that capability (although it is not "activated" for the users at this point).

You would require a source license to do so.

[WorldWideZ]

ok well scratch that till i find a huge bag of money from somewhere if its not a trade secret how do you achieve the dynamic look - im trying several methods but havent found anything that hangs right yet , although i guess i could weld the layers of mesh together or something to make it flow better , i had thought there was some more cleverness that i might be missing

[Legacy]

ok well scratch that till i find a huge bag of money from somewhere if its not a trade secret how do you achieve the dynamic look - im trying several methods but havent found anything that hangs right yet , although i guess i could weld the layers of mesh together or something to make it flow better , i had thought there was some more cleverness that i might be missing

Usually you just add some more bones (skirt, cape, ...) and animate them along. Then use normal Skin to create the illusion of dynamic cloth...

[Triplelexx]

I remember from back in the old day that HeroEngine has that capability (although it is not "activated" for the users at this point).

You would require a source license to do so.

I'm still not sure why cloth physics requires calculation on the server.
If you just wanted each viewer to run it's own local simulation of some cloth, I don't see why interaction with the server would be needed.

Explosions and such you would want accurate synchronized simulation I guess, but why cloth?
I am still looking at using bones personally, but having to jiggle stuff every animation by hand is painful.

[HE-BENNETT]

Regardless of whether it is do-able, to answer the original question, cloth simulations aren't really a good idea unless you're a big studio with lots of funding (and yes, a source license).  Getting them right can save animation time, IF it is necessary to have dynamic, flowing cloth.  But practically speaking, designing the characters around not having dynamic cloth is a VERY good idea and would probably take far less time then getting digital cloth behaving right.

In short, "NO CAPES!".

[grogbro]

I agree wholeheartedly that you don't need cloth physics in your MMO. Just use some bones and get the same effect and better performance.  I do understand the need to see if its capable within HE though. As it is I am used to working on MMO's without normal maps, so even that's a big thing I get to use, but cloth physics...thats another story.

Plus, why do your zombies need cloth physics??   

[HE-HERB]

Cloth has nothing to do with the server.

Cloth was deprecated on the client for three basic reasons outside our control.

1. Performance, even with hardware, was poor.  Especially, because you wanted everyone (or most people) to have simulated capes, hair, clothing, whatever.  And, the performance really only allowed about .... 3 capes.

2. The drivers for the hardware caused graphical performance issues, especially under the kind of scene load you'd have with an MMO where there's more than about 3 people with capes on.

3. The tools for creating the influences, beyond being not great, required Max or Maya.  Beyond the "not great" thing, we've been pulling functionality out and away from Max and Maya so they can eventually be removed as a dependency.

We are looking at options to bring usable cloth and hair simulation in.  And, we would do so today if the engine was used purely for 3 dudes with capes standing in a room, but MMO scenes are more complex than that.

Cheers