Re: transferring low res syflex to high res geo

2012-12-07 Thread olivier jeannel
Loved that one ! Would be worth uploading this to vimeo/Ice group. 
Excellent reference imho :)


Le 07/12/2012 05:36, Orlando Esponda a écrit :
Maybe I'm missing something here, because the solutions are rather 
complex  hehe  so please bear with me if I'm totally out of context.   
Here's a little video showing how my first suggestion works. The 
smoothing part, well, that's up to you, I was just playing around, but 
who knows, maybe it works for you...


http://www.screencast.com/t/jZIJ3L3lEDPc

Orlando.

On Thu, Dec 6, 2012 at 10:19 PM, Eric Thivierge ethivie...@gmail.com 
mailto:ethivie...@gmail.com wrote:


I never said it was going to be perfect or easy. :) Maybe
averaging the neighboring RefFrames would give a better result?



Eric Thivierge
http://www.ethivierge.com


On Fri, Dec 7, 2012 at 3:16 PM, Matt Lind
ml...@carbinestudios.com mailto:ml...@carbinestudios.com wrote:

They’ll stick, but they won’t necessarily have a desirable or
easily predictable orientation as the results will largely be
dependent on the structure of the mesh topology.




--
IMPRESSUM:
PiXABLE STUDIOS GmbH  Co.KG, Sitz: Dresden, Amtsgericht: Dresden, HRA 
6857,
Komplementärin: Lenhard  Barth Verwaltungsgesellschaft mbH, Sitz: 
Dresden,
Amtsgericht: Dresden, HRB 26501, Geschäftsführer: Frank Lenhard, Tino 
Barth


IMPRINT:
PiXABLE STUDIOS GmbH  Co.KG, Domicile: Dresden, Court of Registery: 
Dresden,

Company Registration Number: HRA 6857, General Partner: Lenhard  Barth
Verwaltungsgesellschaft mbH, Domicile: Dresden, Court of Registery: 
Dresden, Company
Registration Number: HRB 26501, Chief Executive Officers: Frank 
Lenhard, Tino Barth



--
Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte 
Informationen. Wenn Sie nicht
der richtige Adressat sind oder diese E-Mail irrtümlich erhalten 
haben, informieren Sie bitte
sofort den Absender und vernichten Sie diese Mail. Das unerlaubte 
Kopieren sowie die

unbefugte Weitergabe dieser Mail ist nicht gestattet.

This e-mail may contain confidential and/or privileged information. If 
you are not the intended
recipient (or have received this e-mail in error) please notify the 
sender immediately and destroy
this e-mail. Any unauthorized copying, disclosure or distribution of 
the material in this e-mail is
strictly forbidden. 




Re: transferring low res syflex to high res geo

2012-12-06 Thread Eric Thivierge
Use ICE to get the closest location on your low res (static, undeformed
copy) from the high res mesh and store the offset transform from that
location. Then use the reinterpret location onto the deformed low res geo.


Eric Thivierge
http://www.ethivierge.com


On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.com wrote:

 I'm curious how others handle layered or very high res geo clothing?  I
 have a pretty high res geo of layered clothing on a character.  I'm getting
 some decent results with just applying syflex to the geo, enabling self
 collision, boosting the subframe steps, etc. and letting 'er rip but its
 pretty slow.  I get fantastic results using one, lower res proxy mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping the
 high res too it but now I'm dealing with managing almost 2k worth of nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris



Re: transferring low res syflex to high res geo

2012-12-06 Thread Ed Manning
ICE?

I'm pretty sure there's either a sample scene or a tutorial floating around
online on how to deform one mesh onto another.

It's a little late in the day for me to actually work this out, but
basically, you'd have a highres mesh with an ICE tree getting position
deltas (change from last frame to current frame), interpolated, from the
corresponding locations on the low-res mesh.  Might want to have matching
UV's, but that might not be necessary.

I guess really it's a cage deformer running in ICE, but without any kind of
static weight binding from one mesh to the other.

Or I suppose you could try the old cage deformer, but that's horrible --
slow and inflexible.

Or extract a lot of curves from the low-res mesh, plot their shapes, and
envelope the high-res to them, if you need something like your current
workflow?

Or maybe I'm not understanding how complex the geo is, but maybe you could
cache the low-res simmed mesh, then extract the poly clusters that
correspond to the different parts of the high-res clothing.  Then either
leave the extractions live or if it's too slow, cache them. Then apply
however much subdivision and editing of the geo you need to the cached
extracted meshes.  Essentially leaving the modeling stack live on top of
the low res sim.

But that's getting kind of silly, I think.  ;-)


Ed


Re: transferring low res syflex to high res geo

2012-12-06 Thread Ed Manning
wouldn't the stored offsets be in world space?  Wouldn't you want them in
tangent space to not crumple weirdly?

On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Use ICE to get the closest location on your low res (static, undeformed
 copy) from the high res mesh and store the offset transform from that
 location. Then use the reinterpret location onto the deformed low res geo.

 
 Eric Thivierge
 http://www.ethivierge.com



 On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.com wrote:

 I'm curious how others handle layered or very high res geo clothing?  I
 have a pretty high res geo of layered clothing on a character.  I'm getting
 some decent results with just applying syflex to the geo, enabling self
 collision, boosting the subframe steps, etc. and letting 'er rip but its
 pretty slow.  I get fantastic results using one, lower res proxy mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping the
 high res too it but now I'm dealing with managing almost 2k worth of nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris





Re: transferring low res syflex to high res geo

2012-12-06 Thread Kris Rivel
Thanks for the suggestions.  Eric...your idea is what I had in mind but
having a hard time figuring out the right node usage once I extract closest
point.  How do I store an offset?  Cage is ridiculously slow so thats out.


On Thu, Dec 6, 2012 at 10:38 PM, Ed Manning etmth...@gmail.com wrote:

 wouldn't the stored offsets be in world space?  Wouldn't you want them in
 tangent space to not crumple weirdly?


 On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Use ICE to get the closest location on your low res (static, undeformed
 copy) from the high res mesh and store the offset transform from that
 location. Then use the reinterpret location onto the deformed low res geo.

 
 Eric Thivierge
 http://www.ethivierge.com



 On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.com wrote:

 I'm curious how others handle layered or very high res geo clothing?  I
 have a pretty high res geo of layered clothing on a character.  I'm getting
 some decent results with just applying syflex to the geo, enabling self
 collision, boosting the subframe steps, etc. and letting 'er rip but its
 pretty slow.  I get fantastic results using one, lower res proxy mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping the
 high res too it but now I'm dealing with managing almost 2k worth of nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris






Re: transferring low res syflex to high res geo

2012-12-06 Thread Eric Thivierge
Well yes you're going to have to figure out the transform space when
storing in the point's local space. Though I think you could hack your way
through it by ripping apart the stick to location node.


Eric Thivierge
http://www.ethivierge.com


On Fri, Dec 7, 2012 at 2:38 PM, Ed Manning etmth...@gmail.com wrote:

 wouldn't the stored offsets be in world space?  Wouldn't you want them in
 tangent space to not crumple weirdly?


 On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Use ICE to get the closest location on your low res (static, undeformed
 copy) from the high res mesh and store the offset transform from that
 location. Then use the reinterpret location onto the deformed low res geo.

 
 Eric Thivierge
 http://www.ethivierge.com



 On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.com wrote:

 I'm curious how others handle layered or very high res geo clothing?  I
 have a pretty high res geo of layered clothing on a character.  I'm getting
 some decent results with just applying syflex to the geo, enabling self
 collision, boosting the subframe steps, etc. and letting 'er rip but its
 pretty slow.  I get fantastic results using one, lower res proxy mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping the
 high res too it but now I'm dealing with managing almost 2k worth of nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris






Re: transferring low res syflex to high res geo

2012-12-06 Thread Eric Thivierge
Kris are you using the standard Cage deformer or the ICE based Hull
Deformer setup?


Eric Thivierge
http://www.ethivierge.com


On Fri, Dec 7, 2012 at 2:44 PM, Kris Rivel krisri...@gmail.com wrote:

 Thanks for the suggestions.  Eric...your idea is what I had in mind but
 having a hard time figuring out the right node usage once I extract closest
 point.  How do I store an offset?  Cage is ridiculously slow so thats out.


 On Thu, Dec 6, 2012 at 10:38 PM, Ed Manning etmth...@gmail.com wrote:

 wouldn't the stored offsets be in world space?  Wouldn't you want them in
 tangent space to not crumple weirdly?


 On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Use ICE to get the closest location on your low res (static, undeformed
 copy) from the high res mesh and store the offset transform from that
 location. Then use the reinterpret location onto the deformed low res geo.

 
 Eric Thivierge
 http://www.ethivierge.com



 On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.com wrote:

 I'm curious how others handle layered or very high res geo clothing?  I
 have a pretty high res geo of layered clothing on a character.  I'm getting
 some decent results with just applying syflex to the geo, enabling self
 collision, boosting the subframe steps, etc. and letting 'er rip but its
 pretty slow.  I get fantastic results using one, lower res proxy mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping the
 high res too it but now I'm dealing with managing almost 2k worth of nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris







Re: transferring low res syflex to high res geo

2012-12-06 Thread Kris Rivel
Thanks Matt...could give that a go too...although I think ICE syflex only
works on poly?  Eric...good suggestion for a starting point..will try that.
 I was referring to the default cage deformer...didn't think about hull
deformer.  I was looking at a few images and demos on syflex's website for
the Assassin's Creed trailer.  Pretty awesome setup they had with multiple
layers of clothing.  The ICE Syflex seems much more powerful at first and
is definitely faster I think but lacks some of the tools the addon version
has like weightmaps.  There's another one I'm looking for an answer too.
 Although I figured if I was able to blend a syflex mesh with a weighted
mesh using a weight map in ICE, I'd be set!


On Thu, Dec 6, 2012 at 10:46 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Kris are you using the standard Cage deformer or the ICE based Hull
 Deformer setup?

 
 Eric Thivierge
 http://www.ethivierge.com


 On Fri, Dec 7, 2012 at 2:44 PM, Kris Rivel krisri...@gmail.com wrote:

 Thanks for the suggestions.  Eric...your idea is what I had in mind but
 having a hard time figuring out the right node usage once I extract closest
 point.  How do I store an offset?  Cage is ridiculously slow so thats out.


 On Thu, Dec 6, 2012 at 10:38 PM, Ed Manning etmth...@gmail.com wrote:

 wouldn't the stored offsets be in world space?  Wouldn't you want them
 in tangent space to not crumple weirdly?


 On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Use ICE to get the closest location on your low res (static, undeformed
 copy) from the high res mesh and store the offset transform from that
 location. Then use the reinterpret location onto the deformed low res geo.

 
 Eric Thivierge
 http://www.ethivierge.com



 On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.com wrote:

 I'm curious how others handle layered or very high res geo clothing?
  I have a pretty high res geo of layered clothing on a character.  I'm
 getting some decent results with just applying syflex to the geo, enabling
 self collision, boosting the subframe steps, etc. and letting 'er rip but
 its pretty slow.  I get fantastic results using one, lower res proxy mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping 
 the
 high res too it but now I'm dealing with managing almost 2k worth of nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris








Re: transferring low res syflex to high res geo

2012-12-06 Thread Simon Anderson
We would take the high res geo, reduce it, then run a custom script that
generated nulls that were cluster constrained to the points on the geometry
and simulate it, then envelope those points onto out the high mesh. This
was one of our work arounds, depending on the shot and what was needed, but
this solution seemed to work pretty well, as long as your create the
geometry from a default flat position and then simulate into position.

we tried cage deformers on some things, but found the size of the cage
deformers to be huge compared to the enveloping.

I think each situation calls for different approach. Found that some times
you would get great results with one method and then the same method would
create horrid results.

On Fri, Dec 7, 2012 at 2:46 PM, Eric Thivierge ethivie...@gmail.com wrote:

 Kris are you using the standard Cage deformer or the ICE based Hull
 Deformer setup?

 
 Eric Thivierge
 http://www.ethivierge.com


 On Fri, Dec 7, 2012 at 2:44 PM, Kris Rivel krisri...@gmail.com wrote:

 Thanks for the suggestions.  Eric...your idea is what I had in mind but
 having a hard time figuring out the right node usage once I extract closest
 point.  How do I store an offset?  Cage is ridiculously slow so thats out.


 On Thu, Dec 6, 2012 at 10:38 PM, Ed Manning etmth...@gmail.com wrote:

 wouldn't the stored offsets be in world space?  Wouldn't you want them
 in tangent space to not crumple weirdly?


 On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Use ICE to get the closest location on your low res (static, undeformed
 copy) from the high res mesh and store the offset transform from that
 location. Then use the reinterpret location onto the deformed low res geo.

 
 Eric Thivierge
 http://www.ethivierge.com



 On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.com wrote:

 I'm curious how others handle layered or very high res geo clothing?
  I have a pretty high res geo of layered clothing on a character.  I'm
 getting some decent results with just applying syflex to the geo, enabling
 self collision, boosting the subframe steps, etc. and letting 'er rip but
 its pretty slow.  I get fantastic results using one, lower res proxy mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping 
 the
 high res too it but now I'm dealing with managing almost 2k worth of nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris








-- 
---
Simon Ben Anderson
blog: http://vinyldevelopment.wordpress.com/


Re: transferring low res syflex to high res geo

2012-12-06 Thread Ed Manning
that's why I was thinking UVs might help -- use them to interpolate the
spatial offsets?

On Thu, Dec 6, 2012 at 10:45 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Well yes you're going to have to figure out the transform space when
 storing in the point's local space. Though I think you could hack your way
 through it by ripping apart the stick to location node.

 
 Eric Thivierge
 http://www.ethivierge.com


 On Fri, Dec 7, 2012 at 2:38 PM, Ed Manning etmth...@gmail.com wrote:

 wouldn't the stored offsets be in world space?  Wouldn't you want them in
 tangent space to not crumple weirdly?


 On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Use ICE to get the closest location on your low res (static, undeformed
 copy) from the high res mesh and store the offset transform from that
 location. Then use the reinterpret location onto the deformed low res geo.

 
 Eric Thivierge
 http://www.ethivierge.com



 On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.com wrote:

 I'm curious how others handle layered or very high res geo clothing?  I
 have a pretty high res geo of layered clothing on a character.  I'm getting
 some decent results with just applying syflex to the geo, enabling self
 collision, boosting the subframe steps, etc. and letting 'er rip but its
 pretty slow.  I get fantastic results using one, lower res proxy mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping the
 high res too it but now I'm dealing with managing almost 2k worth of nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris







Re: transferring low res syflex to high res geo

2012-12-06 Thread Kris Rivel
Thanks Simon...I was using that method as well.  It works fine on some
other elements but for this one particular case, it was difficult due to
the layered fabric.  Will give that another go as well.  Didn't think about
doing a poly reduction, keeping it live and using that as my null deformer
group...might be more manageable.


On Thu, Dec 6, 2012 at 10:52 PM, Simon Anderson 
simonbenandersonl...@gmail.com wrote:

 We would take the high res geo, reduce it, then run a custom script that
 generated nulls that were cluster constrained to the points on the geometry
 and simulate it, then envelope those points onto out the high mesh. This
 was one of our work arounds, depending on the shot and what was needed, but
 this solution seemed to work pretty well, as long as your create the
 geometry from a default flat position and then simulate into position.

 we tried cage deformers on some things, but found the size of the cage
 deformers to be huge compared to the enveloping.

 I think each situation calls for different approach. Found that some times
 you would get great results with one method and then the same method would
 create horrid results.


 On Fri, Dec 7, 2012 at 2:46 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Kris are you using the standard Cage deformer or the ICE based Hull
 Deformer setup?

 
 Eric Thivierge
 http://www.ethivierge.com


 On Fri, Dec 7, 2012 at 2:44 PM, Kris Rivel krisri...@gmail.com wrote:

 Thanks for the suggestions.  Eric...your idea is what I had in mind but
 having a hard time figuring out the right node usage once I extract closest
 point.  How do I store an offset?  Cage is ridiculously slow so thats out.


 On Thu, Dec 6, 2012 at 10:38 PM, Ed Manning etmth...@gmail.com wrote:

 wouldn't the stored offsets be in world space?  Wouldn't you want them
 in tangent space to not crumple weirdly?


 On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge 
 ethivie...@gmail.comwrote:

 Use ICE to get the closest location on your low res (static,
 undeformed copy) from the high res mesh and store the offset transform 
 from
 that location. Then use the reinterpret location onto the deformed low res
 geo.

 
 Eric Thivierge
 http://www.ethivierge.com



 On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.comwrote:

 I'm curious how others handle layered or very high res geo clothing?
  I have a pretty high res geo of layered clothing on a character.  I'm
 getting some decent results with just applying syflex to the geo, 
 enabling
 self collision, boosting the subframe steps, etc. and letting 'er rip but
 its pretty slow.  I get fantastic results using one, lower res proxy 
 mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping 
 the
 high res too it but now I'm dealing with managing almost 2k worth of 
 nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris








 --
 ---
 Simon Ben Anderson
 blog: http://vinyldevelopment.wordpress.com/




Re: transferring low res syflex to high res geo

2012-12-06 Thread Ed Manning
Anyway, past my bedtime.  Kris, you're in much better hands than mine if
you listen to Eric, Matt et al.  Best of luck with it.

On Thu, Dec 6, 2012 at 10:55 PM, Ed Manning etmth...@gmail.com wrote:

 that's why I was thinking UVs might help -- use them to interpolate the
 spatial offsets?


 On Thu, Dec 6, 2012 at 10:45 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Well yes you're going to have to figure out the transform space when
 storing in the point's local space. Though I think you could hack your way
 through it by ripping apart the stick to location node.

 
 Eric Thivierge
 http://www.ethivierge.com


 On Fri, Dec 7, 2012 at 2:38 PM, Ed Manning etmth...@gmail.com wrote:

 wouldn't the stored offsets be in world space?  Wouldn't you want them
 in tangent space to not crumple weirdly?


 On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge ethivie...@gmail.comwrote:

 Use ICE to get the closest location on your low res (static, undeformed
 copy) from the high res mesh and store the offset transform from that
 location. Then use the reinterpret location onto the deformed low res geo.

 
 Eric Thivierge
 http://www.ethivierge.com



 On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel krisri...@gmail.com wrote:

 I'm curious how others handle layered or very high res geo clothing?
  I have a pretty high res geo of layered clothing on a character.  I'm
 getting some decent results with just applying syflex to the geo, enabling
 self collision, boosting the subframe steps, etc. and letting 'er rip but
 its pretty slow.  I get fantastic results using one, lower res proxy mesh.
  Whats the best way to transfer or copy a low res sim to a high res mesh?
  I was creating nulls on each point on the lower res geo and enveloping 
 the
 high res too it but now I'm dealing with managing almost 2k worth of nulls
 and any envelope corrections are impossible.  Any other suggestions other
 than brute force?

 Kris








RE: transferring low res syflex to high res geo

2012-12-06 Thread Matt Lind
You'd have context switching issues doing that as each vertex has multiple 
samples (UVs).

Matt



From: softimage-boun...@listproc.autodesk.com 
[mailto:softimage-boun...@listproc.autodesk.com] On Behalf Of Ed Manning
Sent: Thursday, December 06, 2012 7:56 PM
To: softimage@listproc.autodesk.com
Subject: Re: transferring low res syflex to high res geo

that's why I was thinking UVs might help -- use them to interpolate the spatial 
offsets?
On Thu, Dec 6, 2012 at 10:45 PM, Eric Thivierge 
ethivie...@gmail.commailto:ethivie...@gmail.com wrote:
Well yes you're going to have to figure out the transform space when storing in 
the point's local space. Though I think you could hack your way through it by 
ripping apart the stick to location node.


Eric Thivierge
http://www.ethivierge.com

On Fri, Dec 7, 2012 at 2:38 PM, Ed Manning 
etmth...@gmail.commailto:etmth...@gmail.com wrote:
wouldn't the stored offsets be in world space?  Wouldn't you want them in 
tangent space to not crumple weirdly?

On Thu, Dec 6, 2012 at 10:31 PM, Eric Thivierge 
ethivie...@gmail.commailto:ethivie...@gmail.com wrote:
Use ICE to get the closest location on your low res (static, undeformed copy) 
from the high res mesh and store the offset transform from that location. Then 
use the reinterpret location onto the deformed low res geo.


Eric Thivierge
http://www.ethivierge.com


On Fri, Dec 7, 2012 at 1:58 PM, Kris Rivel 
krisri...@gmail.commailto:krisri...@gmail.com wrote:
I'm curious how others handle layered or very high res geo clothing?  I have a 
pretty high res geo of layered clothing on a character.  I'm getting some 
decent results with just applying syflex to the geo, enabling self collision, 
boosting the subframe steps, etc. and letting 'er rip but its pretty slow.  I 
get fantastic results using one, lower res proxy mesh.  Whats the best way to 
transfer or copy a low res sim to a high res mesh?  I was creating nulls on 
each point on the lower res geo and enveloping the high res too it but now I'm 
dealing with managing almost 2k worth of nulls and any envelope corrections are 
impossible.  Any other suggestions other than brute force?

Kris






RE: transferring low res syflex to high res geo

2012-12-06 Thread Matt Lind
EDIT:

Surface constrain the nulls to the NURBS surface, not to the control points.  
This would allow you to drive the orientation of the nulls as well as 
translation.

Should be able to do the same trick with a low resolution polygon mesh as the 
simulated surface if NURBS doesn't work.

Matt



From: softimage-boun...@listproc.autodesk.com 
[mailto:softimage-boun...@listproc.autodesk.com] On Behalf Of Matt Lind
Sent: Thursday, December 06, 2012 7:47 PM
To: softimage@listproc.autodesk.com
Subject: RE: transferring low res syflex to high res geo

I wrote an article on this technique many years ago for HDRI 3D magazine:

Assuming Syflex works on NURBS surfaces and your cloth is a planar polygon mesh 
deformed with wrinkles and whatever.  I would surface constrain the mesh to a 
NURBS surface of similar size/shape, and run Syflex on the NURBS surface.  When 
you're testing the simulation, you can hide the mesh to get the performance.  
When the simulation is to your liking, unhide the mesh to check the results.  
When you have something you really like, you can optionally plot the results as 
shape animation onto the polygon mesh.

If you prefer, you can also plot out as envelope animation by making a bunch of 
nulls to act as deformers of your polygon mesh, and then constrain them to the 
control points of the NURBS surface.  Plot transforms for the nulls whenever 
you feel the results are to your liking.  If you want multiple 'takes' to 
choose from, you can plot to a mixer source instead of directly to the 
Transform parameters as FCurves.  You can then play with the mixer sources to 
further massage into a result you like.

This would give you a lot of freedom and quick iteration.

Matt

PS - if Syflex doesn't work for NURBS, you can use a low res mesh as the 
working copy.




From: 
softimage-boun...@listproc.autodesk.commailto:softimage-boun...@listproc.autodesk.com
 [mailto:softimage-boun...@listproc.autodesk.com] On Behalf Of Kris Rivel
Sent: Thursday, December 06, 2012 6:59 PM
To: Softimage List
Subject: transferring low res syflex to high res geo

I'm curious how others handle layered or very high res geo clothing?  I have a 
pretty high res geo of layered clothing on a character.  I'm getting some 
decent results with just applying syflex to the geo, enabling self collision, 
boosting the subframe steps, etc. and letting 'er rip but its pretty slow.  I 
get fantastic results using one, lower res proxy mesh.  Whats the best way to 
transfer or copy a low res sim to a high res mesh?  I was creating nulls on 
each point on the lower res geo and enveloping the high res too it but now I'm 
dealing with managing almost 2k worth of nulls and any envelope corrections are 
impossible.  Any other suggestions other than brute force?

Kris


Re: transferring low res syflex to high res geo

2012-12-06 Thread Eric Thivierge
Locations have the PolyReferenceFrame or PointReferenceFrame that you could
use which I think the Stick to Location method / Hull Deformer uses...


Eric Thivierge
http://www.ethivierge.com


On Fri, Dec 7, 2012 at 2:55 PM, Ed Manning etmth...@gmail.com wrote:

 that's why I was thinking UVs might help -- use them to interpolate the
 spatial offsets?


RE: transferring low res syflex to high res geo

2012-12-06 Thread Matt Lind
They'll stick, but they won't necessarily have a desirable or easily 
predictable orientation as the results will largely be dependent on the 
structure of the mesh topology.

Matt



From: softimage-boun...@listproc.autodesk.com 
[mailto:softimage-boun...@listproc.autodesk.com] On Behalf Of Eric Thivierge
Sent: Thursday, December 06, 2012 8:06 PM
To: softimage@listproc.autodesk.com
Subject: Re: transferring low res syflex to high res geo

Locations have the PolyReferenceFrame or PointReferenceFrame that you could use 
which I think the Stick to Location method / Hull Deformer uses...


Eric Thivierge
http://www.ethivierge.com

On Fri, Dec 7, 2012 at 2:55 PM, Ed Manning 
etmth...@gmail.commailto:etmth...@gmail.com wrote:
that's why I was thinking UVs might help -- use them to interpolate the spatial 
offsets?



Re: transferring low res syflex to high res geo

2012-12-06 Thread Raffaele Fragapane
With the right topology and density, they will.
We bound millions of points worth of creatures between dragons and dinos
that way across multiple resolutions ;)

On Fri, Dec 7, 2012 at 3:16 PM, Matt Lind ml...@carbinestudios.com wrote:

 They’ll stick, but they won’t necessarily have a desirable or easily
 predictable orientation as the results will largely be dependent on the
 structure of the mesh topology.

 ** **

 Matt





RE: transferring low res syflex to high res geo

2012-12-06 Thread Matt Lind
As a user my main worry would be editing one of the mesh topologies would 
unexpectedly cause some nulls to change orientation because the reference frame 
changes.

A safer approach would be to create a smooth and continuous coordinate space 
and align the nulls using that data. If the mesh topology is edited, the nulls 
won't have the rug unexpectedly pulled out from underneath.  Nvidia has sample 
code to compute such a space for normal maps (NvMeshMender?).  The code could 
be modified to output a coordinate space as a weightmap, vertex color, UVspace, 
userdatablob, or whatever suits the purpose.

Matt





From: softimage-boun...@listproc.autodesk.com 
[mailto:softimage-boun...@listproc.autodesk.com] On Behalf Of Eric Thivierge
Sent: Thursday, December 06, 2012 8:20 PM
To: softimage@listproc.autodesk.com
Subject: Re: transferring low res syflex to high res geo

I never said it was going to be perfect or easy. :) Maybe averaging the 
neighboring RefFrames would give a better result?


Eric Thivierge
http://www.ethivierge.com

On Fri, Dec 7, 2012 at 3:16 PM, Matt Lind 
ml...@carbinestudios.commailto:ml...@carbinestudios.com wrote:
They'll stick, but they won't necessarily have a desirable or easily 
predictable orientation as the results will largely be dependent on the 
structure of the mesh topology.



Re: transferring low res syflex to high res geo

2012-12-06 Thread Orlando Esponda
Maybe I'm missing something here, because the solutions are rather complex
hehe  so please bear with me if I'm totally out of context.   Here's a
little video showing how my first suggestion works. The smoothing part,
well, that's up to you, I was just playing around, but who knows, maybe it
works for you...

http://www.screencast.com/t/jZIJ3L3lEDPc

Orlando.

On Thu, Dec 6, 2012 at 10:19 PM, Eric Thivierge ethivie...@gmail.comwrote:

 I never said it was going to be perfect or easy. :) Maybe averaging the
 neighboring RefFrames would give a better result?


 
 Eric Thivierge
 http://www.ethivierge.com


 On Fri, Dec 7, 2012 at 3:16 PM, Matt Lind ml...@carbinestudios.comwrote:

 They’ll stick, but they won’t necessarily have a desirable or easily
 predictable orientation as the results will largely be dependent on the
 structure of the mesh topology.




-- 
--
IMPRESSUM:
PiXABLE STUDIOS GmbH  Co.KG, Sitz: Dresden, Amtsgericht: Dresden, HRA 6857,
Komplementärin: Lenhard  Barth Verwaltungsgesellschaft mbH, Sitz: Dresden,
Amtsgericht: Dresden, HRB 26501, Geschäftsführer: Frank Lenhard, Tino Barth

IMPRINT:
PiXABLE STUDIOS GmbH  Co.KG, Domicile: Dresden, Court of Registery: 
Dresden,
Company Registration Number: HRA 6857, General Partner: Lenhard  Barth
Verwaltungsgesellschaft mbH, Domicile: Dresden, Court of Registery: 
Dresden, Company
Registration Number: HRB 26501, Chief Executive Officers: Frank Lenhard, 
Tino Barth


--
Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte 
Informationen. Wenn Sie nicht
der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, 
informieren Sie bitte
sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren 
sowie die
unbefugte Weitergabe dieser Mail ist nicht gestattet.

This e-mail may contain confidential and/or privileged information. If you 
are not the intended
recipient (or have received this e-mail in error) please notify the sender 
immediately and destroy
this e-mail. Any unauthorized copying, disclosure or distribution of the 
material in this e-mail is
strictly forbidden. 


Re: transferring low res syflex to high res geo

2012-12-06 Thread Raffaele Fragapane
Why are we even talking at all when Leonard pointed out there's a factory
node doing exactly what is being asked for? :p


Re: transferring low res syflex to high res geo

2012-12-06 Thread Leonard Koch
That is not a factory node.
I have built similar systems in the past.
So it didn't take me too long to make a proper compound for this.


On Fri, Dec 7, 2012 at 6:01 AM, Raffaele Fragapane 
raffsxsil...@googlemail.com wrote:

 Why are we even talking at all when Leonard pointed out there's a factory
 node doing exactly what is being asked for? :p




RE: transferring low res syflex to high res geo

2012-12-06 Thread Matt Lind
You suggested using a reference frame.  I merely pointed out a potential 
trouble spot.


Matt





From: softimage-boun...@listproc.autodesk.com 
[mailto:softimage-boun...@listproc.autodesk.com] On Behalf Of Eric Thivierge
Sent: Thursday, December 06, 2012 8:58 PM
To: softimage@listproc.autodesk.com
Subject: Re: transferring low res syflex to high res geo

Why are we even talking nulls when Kris already said it was getting too 
cumbersome to use nulls?


Eric Thivierge
http://www.ethivierge.com

On Fri, Dec 7, 2012 at 3:33 PM, Matt Lind 
ml...@carbinestudios.commailto:ml...@carbinestudios.com wrote:
As a user my main worry would be editing one of the mesh topologies would 
unexpectedly cause some nulls to change orientation...



Re: transferring low res syflex to high res geo

2012-12-06 Thread Raffaele Fragapane
Oh ok, thought I was just behind on features (haven't used syflex since
Sucker Punch).
Good stuff :)

On Fri, Dec 7, 2012 at 4:04 PM, Leonard Koch leonardkoch...@gmail.comwrote:

 That is not a factory node.
 I have built similar systems in the past.
 So it didn't take me too long to make a proper compound for this.


 On Fri, Dec 7, 2012 at 6:01 AM, Raffaele Fragapane 
 raffsxsil...@googlemail.com wrote:

 Why are we even talking at all when Leonard pointed out there's a factory
 node doing exactly what is being asked for? :p





-- 
Our users will know fear and cower before our software! Ship it! Ship it
and let them flee like the dogs they are!


Re: transferring low res syflex to high res geo

2012-12-06 Thread Ben Davis
And merry Christmas to you too Leonard, I have a scenario that this
compound will find usage in right this morning.
Thanks for sharing!

Ben

--
Benjamin Clifford Davis

www.moondog-animation.com

office:   +33 9 50 04 76 15
mobile: +33 6 88 48 54 50

6 bis avenue des Iles
74000 Annecy
FRANCE



On Fri, Dec 7, 2012 at 6:19 AM, Raffaele Fragapane 
raffsxsil...@googlemail.com wrote:

 Oh ok, thought I was just behind on features (haven't used syflex since
 Sucker Punch).
 Good stuff :)


 On Fri, Dec 7, 2012 at 4:04 PM, Leonard Koch leonardkoch...@gmail.comwrote:

 That is not a factory node.
 I have built similar systems in the past.
 So it didn't take me too long to make a proper compound for this.


 On Fri, Dec 7, 2012 at 6:01 AM, Raffaele Fragapane 
 raffsxsil...@googlemail.com wrote:

 Why are we even talking at all when Leonard pointed out there's a
 factory node doing exactly what is being asked for? :p





 --
 Our users will know fear and cower before our software! Ship it! Ship it
 and let them flee like the dogs they are!