ICE strand orientation trouble - again :/

[Morten Bartholdy]

I have a nice setup which creates lovely grass deforming geometry instances on strands so I can have turbulent wind moving the strands around. I am using Emit Strands for this, but I seem to be unable to randomize the instance y-orientation using a Randomize Rotation by Cone - they are unaffected by this. If I create a different setup using Emit from Geometry and Create Strands on that, I can control the y-rotation, but then the instances are upside down, and my non simulated turbulence does not work.

 

How can I get to control instance y-rotation when using Emit Strands?

 

MOrten

 

[Morten Bartholdy]

Here are a couple of images showing what I am doing:

 

http://dl.dropbox.com/u/763668/ICE/Strands_InstanceGeo_Rotation_01.jpg

 

http://dl.dropbox.com/u/763668/ICE/Strands_InstanceGeo_Rotation_ICE_Tree_01.jpg

 

I just need to randomize of the geometrys y-rotation, and according to the docs and pointers given here earlier this should do the trick, but no go. Is there a magic step I might have missed to get it working?

 

/Morten

[Stefano Jannuzzo]

You posted an Arnold render. You mean that you can't achieve the same result with mr, or that the Arnold result itself is wrong?
Stefano

[Christian Keller]

try the loft ... checkbox in your create strands compound.
and if you use the sitoa strand compound, there´s also one for the
rotations.
not sitting in front of xsi now, but sure you´ll figure it out within
one minute ;)

-- 
christian keller
visual effects|direction

m +49 179 69 36 248
f +49 40 386 835 33
chri...@me.com

[Arda Sernaz]

Yes i tried a feather instance on strands but i couldn't set the orientation...

-- 
Ardahan SERNAZ
VFX TD
www.lhvfx.com

[Sandy Sutherland]

I can send you a scene I developed based on a technique I came across somewhere to do awesome simulated strands, by creating particle systems to loft the strands between and doing the sim on the particles - this allow you to set orientation on the particles and then the strands - I was using it to do grass and bushes being flattened by a vehicle and I had the bushes/grass randomly oriented as I was rendering a field using strand deforming.

S.

_____________________________
Sandy Sutherland
Technical Supervisor
sandy.su...@triggerfish.co.za
_____________________________

---

From: softimag...@listproc.autodesk.com [softimag...@listproc.autodesk.com] on behalf of Arda Sernaz [ar...@lhvfx.com]
Sent: 16 May 2012 14:16
To: soft...@listproc.autodesk.com
Subject: Re: ICE strand orientation trouble - again :/ now with images

[Morten Bartholdy]

I would be very interested in that one Sandy :)

 

Morten

[Arda Sernaz]

Yeah! me too :)

[Rob Chapman]

'a technique I came across somewhere' .... oh! , sounds a bit like this http://sites.google.com/site/tekanoice/tutorials/ice-flowers-part-i

;)

heres the scene file from that tutorial
http://sites.google.com/site/tekanoice/tutorials/ice-flowers-part-ii/strands_deforming_flowers_tutorial_send_v01.zip?revision=1

best regards,

Rob

[olivier jeannel]

Ahem, I'd be curious too ^^

[Arda Sernaz]

Hi Rob,

Thanks for sharing, i was also looking for this tutorial. But the main problem is the instances orientation when stranddeform applied onto them.I am sending an image below .

http://dl.dropbox.com/u/40801938/error.jpg

[Sandy Sutherland]

https://docs.google.com/open?id=0B8MfxqvWZh32N1dJV2x2RTIwMGc

Here ya go - it is that system that Rob mentions, I added more particle segments to allow more accurate collision - but otherwise the same thing - you can add forces such as wind into the mix on the particle setups - it is a really nice method to do it!

Cheers

S.

_____________________________
Sandy Sutherland
Technical Supervisor
sandy.su...@triggerfish.co.za
_____________________________

---

From: softimag...@listproc.autodesk.com [softimag...@listproc.autodesk.com] on behalf of olivier jeannel [olivier...@noos.fr]
Sent: 16 May 2012 14:59

[Rob Chapman]

Sandy, just so you know because it isnt that obvious, it was me that wrote that tutorial :)  also Arda, I imagine the orientation problem you are having is because they are all on a sphere rather than a grid - are you taking the spheres normal into account regarding the orientation?  And as for OP, Morten am not sure why your orientation is having the effect it is, perhaps because you are setting the orientation only on emit then doing something else that effects it. try putting the orientation back on as the last step in the ice tree.

best regards

Rob 'tekano' chapman

[Sandy Sutherland]

Yep - remember it now Rob - sorry I did not remember until I had dug my scenes out and saw how they worked - was done a while ago in R&D on some stuff that might have made it into the movie - will most likely not because of time constraints.
It is really good stuff - a nice way to do it!

S.

_____________________________
Sandy Sutherland
Technical Supervisor
sandy.su...@triggerfish.co.za
_____________________________

---

From: softimag...@listproc.autodesk.com [softimag...@listproc.autodesk.com] on behalf of Rob Chapman [tekan...@gmail.com]
Sent: 16 May 2012 15:42

[Sandy Sutherland]

Yep -that is it Rob - that system works a treat for this kind of stuff!

S.

_____________________________
Sandy Sutherland
Technical Supervisor
sandy.su...@triggerfish.co.za
_____________________________

---

From: softimag...@listproc.autodesk.com [softimag...@listproc.autodesk.com] on behalf of Rob Chapman [tekan...@gmail.com]
Sent: 16 May 2012 14:44

To: soft...@listproc.autodesk.com
Subject: Re: ICE strand orientation trouble - again :/ now with images

[Arda Sernaz]

Rob,

i set both particle and strand orientation but nothing changed after , i used align to surface node. in opengl view instances oriented correctly but when i render it is different. 

[Rob Chapman]

Arda, so when it renders they are facing the wrong way? as in they are all pointing inwards rather than outwards?  just revere the instance and freeze - it will look wrong in openGL but render correctly.

[Arda Sernaz]

Rob, the thing is, in some areas they  oriented right so changing the instance rotation wouldn't give a solution.

Thanks  for the fast replies bye the way :) 

[Morten Bartholdy]

Rob, it was Arnold specific. Fixed in a newer version, so upgrading fixed it for me. Happy now :)

BTW I pored over your flower strands stuff to learn, thanks for putting it up there!

 

MB

[Sandy Sutherland]

I would head in Rob's first suggestion - get the data from the emit location in terms of normal etc... and use that to set orientation of the strands.

S.

_____________________________
Sandy Sutherland
Technical Supervisor
sandy.su...@triggerfish.co.za
_____________________________

From: softimag...@listproc.autodesk.com [softimag...@listproc.autodesk.com] on behalf of Arda Sernaz [ar...@lhvfx.com]

Sent: 16 May 2012 16:41

[Rob Chapman]

zactly!  like this here (zipped scene attached)  also, credit where credit is due - I originally ripped the strands from simulated particle technique from a compound that mr Vince Fortin so kindly posted to this list on how to do a nice plasma ball effect :)

best regards

Rob

[Arda Sernaz]

Maybe the problem is melena..