Re: Dirtmap and XSI5

Date : Fri, 21 Oct 2005 18:27:08 -0300
To : <XSI(at)Softimage.COM>
From : "Pablo Tufaro" <pablo.xsi(at)gmail.com>
Subject : Re: Dirtmap and XSI5
Since I had this stranges straight lines I´ve never used XSI´s AO again. Nothing beats Dirtmap for me.
 
P.
 
 
 
 
 
 
 
 
----- Original Message -----
Sent: Friday, October 21, 2005 5:31 PM
Subject: Re: Dirtmap and XSI5

Sometimes I just need to be kicked.

Thanks Peter




On 10/21/05, peterb <peter_b(at)skynet.be> wrote:
 
I just did a few quick tests and dirtmap for the most part is a tiny bit faster. I did a shaded scene with dirtmap and XSI_AO plugged into the diffuse. And an actual AO pass with both. I kept the scene itself really simple. Same shader on all objects. It was a little difficult to get the shaders to look exactly the same though.

Dirtmap - shaded / 8.31s  |  AO / 5.07s
XSI AO - shaded / 9.21s  |  AO / 5.96s

What was a little weird to me though was that if I plugged them into a mix_2_colors node to use the weight to toggle which one is being used the render times doubled and made the times closer together.  Almost identical times.
makes sense doesnt it?
the mixer is not an on/off switch. It mixes results of both inputs.
In your case, both the dirtmap and XSI AO got calculated, which explains the doubled rendertime, and the identical times regardless of which one you'd see.
 
rather use a mix8colors and turn on or off  color1.
 
 


On 10/21/05, guillaume laforge <guillaume.laforge.3d(at)gmail.com> wrote:
I tried your setup Alan, thanks to share it :-)
 
The first thing I tried is the occlusion area shadow like in Renderman. Unfortunately there are artifacts like with dirtmap :-(.
 
Too bad :-/
 
Guillaume
 


 
2005/10/21, Alan Jones <skyphyr(at)gmail.com>:
If you want to get a basis direction with the built in one - plug XSIAmbientOcclusion into a bumpmap node's input. Put the basis vector into the bump input.

Cheers,

Alan.


On 10/21/05, guillaume laforge < guillaume.laforge.3d(at)gmail.com > wrote:
I don't think the XSIAmbientOcclusion shader is slower. In fact I think they are very similar. With same number of samples, same spread and same attenuation it is the same look and same render time. To compare the "far attenuation" parameter of the dirtmap you must set the "Maximum distance" of the XSIOcclusion shader to the same value divided by 2.
There is no "basis vector" parameter in the XSIocclusion shader but you can drive the occlusion with an incidence light pluged in the Dark color.
 
But with mr 3.4 the final gathering is a very good option for ambient occlusion ;-).
 
Cheers,
 
Guillaume Laforge
CG artist | Jr Td
 

 
2005/10/21, Stefan Andersson <sanders3d(at)gmail.com>:
kewl, that was what I needed to hear :)

/stefan


On 10/21/05, Andi Farhall < andi(at)clear.ltd.uk> wrote:
> Dirtmap works well here in 5. I agree about the built-in one
>
>
>
> -----Original Message-----
> From: Stefan Andersson [mailto: sanders3d(at)gmail.com]
> Sent: Friday, October 21, 2005 1:07 PM
> To: XSI(at)Softimage.COM
> Subject: Dirtmap and XSI5
>
> Anyone used Dirtmap with XSI 5? I haven't tried yet but was wondering
> if anyone has tested it yet? The builtin ambient occlusion shader
> doens't really do the same job, and it's slower :)
>

---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi






--
-gc



--
-gc

Search the XSI List archives here or use the advanced search form to search across mailing lists. Searching help is available.
This site supposedly brought to you by Benjamin Grosser and the Imaging Technology Group.