Surfaces Roughness and simulation parameters

Dear List,

I was trying to simulate the visual appearance of brushed alluminum, but
I am struggling with the roughness of the materials.

I am using metal and metal2 at different roughness settings, but I
always seem to get a pattern in the roughness of the material, I was
wondering if any other rpict parameters affect the material behaviour.

My current suspects are:

-ad Set the number of ambient divisions to N
-as Set the number of ambient super-samples to N
-ms Set the medium sampling distance to sampdist

does anybody has some insight in this?
Thanks,

Giovanni

////////////////////////////////////////////////
Giovanni Betti

Hi Giovanni,

Set "-u+" to turn on random Monte Carlo sampling if the pattern bothers you. You'll get a noisier image, but without the brushed look of the low discrepancy sequence rpict normally uses.

-Greg

···

From: "Giovanni Betti" <[email protected]>
Date: July 12, 2011 11:43:02 AM PDT

Dear List,

I was trying to simulate the visual appearance of brushed alluminum, but
I am struggling with the roughness of the materials.

I am using metal and metal2 at different roughness settings, but I
always seem to get a pattern in the roughness of the material, I was
wondering if any other rpict parameters affect the material behaviour.

My current suspects are:

-ad Set the number of ambient divisions to N
-as Set the number of ambient super-samples to N
-ms Set the medium sampling distance to sampdist

does anybody has some insight in this?
Thanks,

Giovanni

Hi Giovanni,

I think that what you need to look at is actually specular jitter and threshold (-sj and -st), -ms is for something unrelated assuming I remember correctly.

Another factor that can contribute to the appearance of a "pattern" (sort of striated) in material with specular and roughness components is that fact that Radiance use pseudo random sampling. One way to make this go away (albeit at some expense in terms of compute time as well as requiring higher image oversampling) is to use the -u switch for uncorrelated monte carlo sampling.

Best,

-Jack

···

--
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction

On 7/12/2011 2:43 PM, Giovanni Betti wrote:

Dear List,

I was trying to simulate the visual appearance of brushed alluminum, but
I am struggling with the roughness of the materials.

I am using metal and metal2 at different roughness settings, but I
always seem to get a pattern in the roughness of the material, I was
wondering if any other rpict parameters affect the material behaviour.

My current suspects are:

-ad Set the number of ambient divisions to N
-as Set the number of ambient super-samples to N
-ms Set the medium sampling distance to sampdist

does anybody has some insight in this?
Thanks,

Giovanni

////////////////////////////////////////////////
Giovanni Betti

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

Thanks Jack and Greg,

I am trying this right now,
Will let you know how it works, once the rendering completes!

Thanks,

G

···

-----Original Message-----
From: Greg Ward [mailto:[email protected]]
Sent: 12 July 2011 19:54
To: Radiance general discussion
Subject: Re: [Radiance-general] Surfaces Roughness and simulation
parameters

Hi Giovanni,

Set "-u+" to turn on random Monte Carlo sampling if the pattern bothers
you. You'll get a noisier image, but without the brushed look of the
low discrepancy sequence rpict normally uses.

-Greg

From: "Giovanni Betti" <[email protected]>
Date: July 12, 2011 11:43:02 AM PDT

Dear List,

I was trying to simulate the visual appearance of brushed alluminum,

but

I am struggling with the roughness of the materials.

I am using metal and metal2 at different roughness settings, but I
always seem to get a pattern in the roughness of the material, I was
wondering if any other rpict parameters affect the material behaviour.

My current suspects are:

-ad Set the number of ambient divisions to N
-as Set the number of ambient super-samples to N
-ms Set the medium sampling distance to sampdist

does anybody has some insight in this?
Thanks,

Giovanni

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general