Kristinka Hair Tests

Comments on work in progress
User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 21:11
Location: Zagreb, Croatia
Contact:

Re: Kristinka Hair Tests

Post by Mathaeus » 13 May 2010, 21:18

Guys,

I need to say that all images I've posted, were rendered with XSI 7.01. Also, I didn't used mi_sss_skin shader, anywhere. Subsurface is done with some kind of ICE vertex color 'shaders', a bit improved version of these.

If anyone is interested, I'll post last version - these 'shaders' are maybe harder to figure out, there is no 'real' connection to lights or FG - but output is a much lighter for Mental Ray, just vertex colors are rendered.

Cheers

User avatar
rray
Moderator
Posts: 1774
Joined: 26 Sep 2009, 15:51
Location: Bonn, Germany
Contact:

Re: Kristinka Hair Tests

Post by rray » 07 Aug 2010, 15:47

Image
softimage resources section updated Jan 5th 2024

j3st3r
Posts: 121
Joined: 11 Jun 2009, 09:13

Re: Kristinka Hair Tests

Post by j3st3r » 01 Nov 2010, 14:57

rray, can you share your setup for short hair? I wanna make a short haired guy, but I really don't know how to start... :(

User avatar
rray
Moderator
Posts: 1774
Joined: 26 Sep 2009, 15:51
Location: Bonn, Germany
Contact:

Re: Kristinka Hair Tests

Post by rray » 01 Nov 2010, 15:03

You can find the scene with the short curly hair here.
softimage resources section updated Jan 5th 2024

j3st3r
Posts: 121
Joined: 11 Jun 2009, 09:13

Re: Kristinka Hair Tests

Post by j3st3r » 01 Nov 2010, 20:08

Thanks Reinhard,

Actually now I'm in deeeeeep trouble, I can't create decent short hair with Kristinka...anybody has some idea to help me?

User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 21:11
Location: Zagreb, Croatia
Contact:

Re: Kristinka Hair Tests

Post by Mathaeus » 01 Nov 2010, 21:07

Jester,

what short hair.... if you can show some picture.... There are so many hairdos, it seems to be really hard to find an unique workflow for all.

Cheers

j3st3r
Posts: 121
Joined: 11 Jun 2009, 09:13

Re: Kristinka Hair Tests

Post by j3st3r » 01 Nov 2010, 21:13


User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 21:11
Location: Zagreb, Croatia
Contact:

Re: Kristinka Hair Tests

Post by Mathaeus » 01 Nov 2010, 22:29

if you want to do that in procedural way .... (I believe most of people would prefer brushes from standard XSI hair, for short hair setups), probably you can try this one

Setup is a mod of 'fur from nurbs' example, with improved 'vector field' compounds. Hair's orientation is created by nulls. For now there are three options: null's Y axis as direction, direction to null itself, circle around null's Y axis. Scaling of nulls modifies the influence.
Or another one ( kH2 Vector Field by Curve) that takes direction from curve.

Nodes are new, very different than ones from last release (incompatible with 'fur' nodes...) .

I created these last Vector field nodes for fur and feathers, if this fits for hair....it would be nice.

Cheers

j3st3r
Posts: 121
Joined: 11 Jun 2009, 09:13

Re: Kristinka Hair Tests

Post by j3st3r » 02 Nov 2010, 06:33

Thanks Anto, I'll give it a try!

Cheers


Jester

shadow_ex_
Posts: 302
Joined: 16 Feb 2010, 14:18

Re: Kristinka Hair Tests

Post by shadow_ex_ » 02 Nov 2010, 11:00

great work on both models and hair :D ;) <------
i will check the thread later more carefully b-( ^^ |
|
edit: checked.goes back to the first line --------------------|

j3st3r
Posts: 121
Joined: 11 Jun 2009, 09:13

Re: Kristinka Hair Tests

Post by j3st3r » 03 Nov 2010, 20:32

So it seems that I got the meaning. But, if I connect the Kh2 Strand Length node to the guide ICE tree, it renders only a small portion of the hair. When I disconnect, the hair is rendered properly. Any idea?

User avatar
rray
Moderator
Posts: 1774
Joined: 26 Sep 2009, 15:51
Location: Bonn, Germany
Contact:

Re: Kristinka Hair Tests

Post by rray » 03 Nov 2010, 23:00

Nice work on these new nodes, Anto.
Jester, I've never had that effect you describe - it could theoretically be that since the guides are referenced to build interpolated hair, the strand length node could have some unwanted side effects. If that is so, it probably shouldn't be used on guides. It's always difficult to debug having only a written description (lots of room for interpretation ;-).
softimage resources section updated Jan 5th 2024

User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 21:11
Location: Zagreb, Croatia
Contact:

Re: Kristinka Hair Tests

Post by Mathaeus » 03 Nov 2010, 23:12

Actually, it can be that some attributes, used by hair filler point cloud, aren't interpolated properly by kH2 Strand Length node (as it does complete 're construction' of everything) - but this should be visible in view ports.

And yes, 'safe' usage of kH2 Strand Length would be at the end of hair filler's ICE tree - 'hair' is interpolated between three closest guides, if these three are varying too much, interpolation will become wild - but this also should be visible in view ports.

btw, it seems to be a new issue :)

j3st3r
Posts: 121
Joined: 11 Jun 2009, 09:13

Re: Kristinka Hair Tests

Post by j3st3r » 04 Nov 2010, 06:59

Meanwhile I resolved the issue, I was driving the parameter with the weightmap but it made a mess, so I drive the length with the weight on another node. Now, the question is how can I set the thickness of the rendered hair?

izze
Posts: 176
Joined: 24 Sep 2009, 20:02

Re: Kristinka Hair Tests

Post by izze » 04 Nov 2010, 18:32

Use KH2 Strand fCurve in SI Units on the size port in the emitter compound.

User avatar
Mathaeus
Posts: 1778
Joined: 08 Jun 2009, 21:11
Location: Zagreb, Croatia
Contact:

Re: Kristinka Hair Tests

Post by Mathaeus » 04 Nov 2010, 22:31

j3st3r wrote:Meanwhile I resolved the issue, I was driving the parameter with the weightmap but it made a mess, so I drive the length with the weight on another node.
I was unable to repro this, but anyway, kh2 Strand Length node, as it is for now, allows to set length over 'ratio' 1 ('ratio' is relative strand length). Probably you won't do that manually, but a mix of some nodes, plugged into 'factor', can do that. It's 'no good' for rendering, because of multiple strand positions on the same place. From my (empirical) knowledge with older XSI versions, this can cause NaNs with Hair Geo shader, or memory leak with pHair_TK shader. Standard XSI Hair shader seems to have some mechanism to avoid these problems (at least I think so...)

Here is a kh2 Strand Length node that doesn't allow the strand positions to 'group' at the end of strand.

That's all what I found with this node, that can cause problems. Maybe... on very short length, it can also set the strand positions on top of each other. Other than that, as RRay said: 'written description = lots of room for interpretation ' :).

Cheers

Post Reply

Who is online

Users browsing this forum: No registered users and 38 guests