Re: Glossy Reflection Bug?

[Reinhard Claus <rclaus(at)gmail.com>]

>> Any known reason for leaving that out of the regular phong, by the way?

It would make sense. It might be that the intention was not to cram too many items into one page.


>> I also would expect the reflected sphere on the floor to be less blurred

>> than it is now. (but thats subjective)

>> the blur would have to be related to the depth/distance.


that's interesting - I remember a thread here about depth-of-field: it
came out that for computing the dof radius, the total raylength must be
taken into account - not the direct/eye distance. This would matter
here because the reflected environment has a longer total raylength
than the local sphere/grid eye ray. 

I don't know though if would be the same issue with glossy reflections. It seems wrong in the picture, although intuitively, the
physical principles behind dof/glossy should be similar!?


Reinhard

2005/10/27, Arvid Björn <arvid.bjorn(at)palladiumfilm.com>:

Any known reason for leaving that out of the regular phong, by the way?


Reinhard Claus wrote:

> Hi there,
>
> It makes sense not to calculate glossy reflections of glossy
> reflections but hard reflections of glossies like in your case should
> be ok. It seems xsi has a common raydepth counter for both glossy and
> hard reflections and optimizes by only calulating glossies at raydepth #1.
> Ideally the two reflection types should be counted separately.
>
> For working around this you could dump phong refl and build your own
> using the reflection_diffuse shader. This has a setting for glossy depth
>
> Reinhard
>
> 2005/10/26, Michael Klein < forum(at)virtualrepublic.org
> <mailto:forum(at)virtualrepublic.org>>:
>
>     Hi,
>
>     I have an interesting problem.
>
>     Watch my picture at:
>     http://www.virtualrepublic.org/xsi/glossy_bug.jpg
>
>     You will recognize that the glossy reflection in the floor plane will
>     reflected normal in the sphere.
>     This seems wrong to me. I expect a reflected floor with a blurred
>     environment in the sphere.
>     Actually I use HDRi Environment but this occurs also with a normal
>     textured
>     environment sphere.
>
>     Any ideas if this is a missing feature in the raytracing code?
>     How is glossy computed? By a shader alghorythm at the end?
>
>     Regards,
>     President Michael Klein
>     virtual republic
>
>     ---
>     Unsubscribe? Mail Majordomo(at)Softimage.COM
>     <mailto: Majordomo(at)Softimage.COM> with the following text in body:
>     unsubscribe xsi
>
>
>
>
> 2005/10/26, Michael Klein <forum(at)virtualrepublic.org
> <mailto:forum(at)virtualrepublic.org>>:
>
>     Hi,
>
>     I have an interesting problem.
>
>     Watch my picture at:
>     http://www.virtualrepublic.org/xsi/glossy_bug.jpg
>
>     You will recognize that the glossy reflection in the floor plane will
>     reflected normal in the sphere.
>     This seems wrong to me. I expect a reflected floor with a blurred
>     environment in the sphere.
>     Actually I use HDRi Environment but this occurs also with a normal
>     textured
>     environment sphere.
>
>     Any ideas if this is a missing feature in the raytracing code?
>     How is glossy computed? By a shader alghorythm at the end?
>
>     Regards,
>     President Michael Klein
>     virtual republic
>
>     ---
>     Unsubscribe? Mail Majordomo(at)Softimage.COM
>     <mailto:Majordomo(at)Softimage.COM > with the following text in body:
>     unsubscribe xsi
>
>
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi

2005/10/27, Michael Klein <forum(at)virtualrepublic.org>:

I do not agree ... visually it's wrong!

---

From: owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM] On Behalf Of Reinhard Claus
Sent: Wednesday, October 26, 2005 11:30 PM
To: XSI(at)Softimage.COM
Subject: Re: Glossy Reflection Bug?

Hi there,

It makes sense not to calculate glossy reflections of glossy reflections but hard reflections of glossies like in your case should be ok. It seems xsi has a common raydepth counter for both glossy and hard reflections and optimizes by only calulating glossies at raydepth #1.
Ideally the two reflection types should be counted separately.

For working around this you could dump phong refl and build your own using the reflection_diffuse shader. This has a setting for glossy depth

Reinhard

2005/10/26, Michael Klein <forum(at)virtualrepublic.org>:

Hi,

I have an interesting problem.

Watch my picture at:
http://www.virtualrepublic.org/xsi/glossy_bug.jpg

You will recognize that the glossy reflection in the floor plane will
reflected normal in the sphere.
This seems wrong to me. I expect a reflected floor with a blurred
environment in the sphere.
Actually I use HDRi Environment but this occurs also with a normal textured
environment sphere.

Any ideas if this is a missing feature in the raytracing code?
How is glossy computed? By a shader alghorythm at the end?

Regards,
President Michael Klein
virtual republic

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

2005/10/26, Michael Klein <forum(at)virtualrepublic.org>:

Hi,

I have an interesting problem.

Watch my picture at:
http://www.virtualrepublic.org/xsi/glossy_bug.jpg

You will recognize that the glossy reflection in the floor plane will
reflected normal in the sphere.
This seems wrong to me. I expect a reflected floor with a blurred
environment in the sphere.
Actually I use HDRi Environment but this occurs also with a normal textured
environment sphere.

Any ideas if this is a missing feature in the raytracing code?
How is glossy computed? By a shader alghorythm at the end?

Regards,
President Michael Klein
virtual republic

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