Cloth the future for skin and deform simulation?

General discussion about 3D DCC and other topics
Bullit
Moderator
Posts: 2621
Joined: 24 May 2012, 07:44

Cloth the future for skin and deform simulation?

Post by Bullit » 11 Jan 2014, 21:00

I am seeing more and more examples of using cloth simulation for character animation and deformation even of rigid materials like a car crash.

So is there a market for a "Realcloth"? <--just word playing with Realflow...

I think the 3d application that will have the most open and most complete cloth solution will be getting a good market and i don't see any application having a good cloth solution.

I worked recently with 3dsMax ones and while superior to Softimage with more options they still have a lot issues.

User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 19:11
Location: Zagreb, Croatia

Re: Cloth the future for skin and deform simulation?

Post by Mathaeus » 11 Jan 2014, 21:48

Bullit wrote: I worked recently with 3dsMax ones and while superior to Softimage with more options they still have a lot issues.
Are you trolling or what... What superior, where superior, who superior. Has presets, or like? Afaik there are two in box, about ten or more years old ClothFX and mCloth, adaption of Nvidia physX toy. Only one that looks impressive in Max world seems to be Qualoth plugin, but price is impressive too. Around 5 k.
When it comes to cloth simulator as modeler, I think every Max user and his neighbor, starts to talking about Marvelous Designer, nothing else.

Bullit
Moderator
Posts: 2621
Joined: 24 May 2012, 07:44

Re: Cloth the future for skin and deform simulation?

Post by Bullit » 12 Jan 2014, 20:09

If you can't read a critic to an aspect of Softimage without resorting to an ad hominem attack i'll have to ask you to not post in this thread Mathaeus.

User avatar
origin
Posts: 619
Joined: 09 Jun 2009, 09:59
Location: warsaw

Re: Cloth the future for skin and deform simulation?

Post by origin » 12 Jan 2014, 21:32

i think qualoth is maya only

User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 19:11
Location: Zagreb, Croatia

Re: Cloth the future for skin and deform simulation?

Post by Mathaeus » 13 Jan 2014, 00:18

Yeah it's Ez Cloth, price is also about 5k.
Back to cloth in Max, a far as I remember, Cloth FX was somewhere in class of old SI cloth - a lot of options, but engine was to weak to run the high resolutions. 128 * 128 grid, this was a challenge. And both were substantially weak, providing a low performance on newer machines, too. It's integrated into Max, 2004.
According to destiny of Reactor, which is integrated 2001 or so, but removed from Max 2012, people should not be surprised if Cloth FX will be removed too. So, most likely, future of simulation in Max, belongs to instant, low cost solutions. Hardly to imagine some significant, competitive usage.

luceric
Posts: 1252
Joined: 21 Jun 2009, 22:08

Re: Cloth the future for skin and deform simulation?

Post by luceric » 13 Jan 2014, 00:39

too much second hand information IMHO.
Bullit wrote:I worked recently with 3dsMax ones and while superior to Softimage with more options they still have a lot issues.
How did you conclude that the max cloth was superior to Syflex in XSI? Superior at what?

User avatar
origin
Posts: 619
Joined: 09 Jun 2009, 09:59
Location: warsaw

Re: Cloth the future for skin and deform simulation?

Post by origin » 13 Jan 2014, 01:01

Mathaeus wrote:Yeah it's Ez Cloth, price is also about 5k.
Back to cloth in Max, a far as I remember, Cloth FX was somewhere in class of old SI cloth - a lot of options, but engine was to weak to run the high resolutions. 128 * 128 grid, this was a challenge. And both were substantially weak, providing a low performance on newer machines, too. It's integrated into Max, 2004.
According to destiny of Reactor, which is integrated 2001 or so, but removed from Max 2012, people should not be surprised if Cloth FX will be removed too. So, most likely, future of simulation in Max, belongs to instant, low cost solutions. Hardly to imagine some significant, competitive usage.
Clothfx is good for simple stuff. I dont think they will remove it, afaik its another plugin they bought back then.
syflex is also old tech...with it's own issues.
tbh I prefere something thats slow but won't fail in middle of 2h sim (houdini cloth) than fast and unstable system (syflex)
I heard ncloth in maya is pretty good :)

Bullit
Moderator
Posts: 2621
Joined: 24 May 2012, 07:44

Re: Cloth the future for skin and deform simulation?

Post by Bullit » 13 Jan 2014, 02:01

How did you conclude that the max cloth was superior to Syflex in XSI? Superior at what?
Tearing and material flexibility in case of Cloth. It can go from paper to cloth to metal like crash deformation. It is slower than Syflex.
Tearing and speed in case of mCloth.


If there will be need for some ICE magic then it is where Softimage Syflex can take the upper hand.


I think while ago i said no cloth system appears to me good enough* and i maintain , imo there is still a jackpot to be won in this area specially if it is thought as more than cloth and integrated with all other simulation tools. But maybe we will have to for 5 years or so.



* adding a lot of bugs.

oglu
Posts: 16
Joined: 19 Aug 2011, 09:40

Re: Cloth the future for skin and deform simulation?

Post by oglu » 13 Jan 2014, 09:48

to get your cloth model i would start with marvelous designer...
a lot of character modlers are using it... they used it also on the hobbit to do the cloth...
http://www.marvelousdesigner.com/

EDIT BY OWEI: PLEASE USE URL TAGS!!!!

User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 19:11
Location: Zagreb, Croatia

Re: Cloth the future for skin and deform simulation?

Post by Mathaeus » 13 Jan 2014, 11:38

origin wrote: tbh I prefere something thats slow but won't fail in middle of 2h sim (houdini cloth) than fast and unstable system (syflex)
I heard ncloth in maya is pretty good :)
Well I think I feel a big facility attitude :) If I'm correct.. congratulations!.

IMHO that ncloth and built in muscles thing ( forgot the name of this software) sounds like best economic solution for mentioned task, at least theoretically.
Max with his disjointed system, what can (not) be connected, what doesn't work. Together with expectation of huge community (including possible boss), mainly based on plain believing. I don't want this, anymore.

EricTRocks
Moderator
Posts: 754
Joined: 25 Nov 2009, 00:41

Re: Cloth the future for skin and deform simulation?

Post by EricTRocks » 13 Jan 2014, 14:29

I've been having pretty decent results with Syflex in ICE recently. I know there are still some bugs and features we need ironed out but I feel pretty confident in Syflex within the ICE framework. I had to do a bunch of crowd cloth sims for White House Down and it was easy to setup and use. Then we just finished a yogurt commercial where I use Syflex in a skin binding setup around some ICE driven muscles.

Having only briefly looked into the Maya Muscle system it seems like while it is probably feature rich can be used on a high end feature rig, it's just sooooo much to setup and configure. The workflow isn't terribly user friendly from my point of view. Too many attributes and settings. I'd rather roll my own in ICE (which I'm doing) and add features as needed and be able to mainpulate all of the data rather than starting from a super heavy standard setup which it seems to be. I understand they need to cover all of the bases which a built in system though for an off the shelf product feature. Worth noting, I haven't used Maya muscle a lot so my views are from just brief research into the system. Would love to hear other experiences.

While Syflex may have been around for many many years, it's algorithms seem really efficient for what it can do. Being able to get at the internals to some of the parameters via ICE adds that much more to the value of the Syflex ICE nodes.
Eric Thivierge
Lead Kraken Developer, Fabric Engine
http://fabric-engine.github.io/Kraken

Bullit
Moderator
Posts: 2621
Joined: 24 May 2012, 07:44

Re: Cloth the future for skin and deform simulation?

Post by Bullit » 14 Jan 2014, 19:22

Thanks for feedback Eric