[Radiance-general] perfect specular reflection in rtrace

[Antoine Bugeat]

Hello all,


Thanks again for your quick reply. Now I can model my street walls as a
perfect specular materials ("mirror" materials) and obtain the results I
expected.

I come back with an other problem:

The main purpose of my work is to compute the impact of the specular
reflections in a street canyon.
So, I would like to separate the light that comes from the reflections of
the walls (only specular) and the light that comes directly from the sky
(defined by "glow" material and "skyfunc"). I wanted to do a simulation
with 0 reflection and an other simulation with 5 (or more) reflections. The
results obtain with 0 reflection is directly the direct sky component and
the substraction of 5 reflections - 0 reflections is the "environment" part.

I did the same process with a diffuse environment and it worked well. I
changed the -ab parameter:
-ab 1 to have the direct sky component (1 because of "glow" material);
and -ab 5 to have the entire reflections in the environment.

*But in the specular environment (with "mirror" wall) there is no
difference between my "0 reflection" case and my "5 reflections" case*
I tried to change the parameters -ab -dr and -lr and it seems that all the
reflections is computed whatever the parameters.
>From what I know, -ab is for "diffuse inter-reflections", -dr is for
"direct reflections", and -lr I don't know...

Obviously, I can replace the "0 reflection specular" case with the "0
refelction diffuse" case because it should be exactly the same, but I would
like to understand where this problem comes from.

I do not know if I was clear and precise in my explanation. I am available
if you have questions.


Regards,

Antoine

2017-10-13 19:04 GMT+02:00 Greg Ward <gregoryjward at gmail.com>:

> Hi Antoine,
>
> Using the mirror material should work, but I forgot to mention that it's a
> one-sided effect.  You need to be looking at the front of your surface
> rather than the backside, which means the vertices need to be ordered
> counter-clockwise from the side you're looking at.
>
> I didn't check initially, but the surface is the wrong way around from
> your selected evaluation point.  Your sun is also on the wrong side.  If
> you reverse the order of your polygon vertices, then it should work as
> expected.
>
> The photon-mapping approach, which is now built into Radiance, should also
> work.  It is a more complex solution, but necessary when the reflecting
> surfaces are curved.
>
> Cheers,
> -Greg
>
> *From: *Antoine Bugeat <bugeatantoine at gmail.com>
>
> *Date: *October 13, 2017 8:02:20 AM PDT
>
>
> Hello all,
>
> First, thank you very much for your quick answers.
>
> If I understand correctly, among the possible options for solving my
> problem, the simplest is to replace the perfect specular "plastic" by a
> "mirror" material to have a "virtual source" calculation (I agree that 100%
> of reflection is totally not real but it's just for this case to understand
> the calculation).
>
> void mirror M_Spec
> 0
> 0
> 3 1 1 1
>
> But I still get 0...
>
> As a reminder, I use these radiance parameters (the same as DAYSIM except
> that I increase the -dr):
> -I -aa 0.1 -ab 5 -ad 1000 -ar 300 -as 20 -dj 0 -dr 5 -ds 0.2 -dp 512 -lr 6
> -lw 0.004 -st 0.1 -oov
>
> In parallel, I found a paper written by Roland Schregle (Lucerne
> University 2015) with an extension of Radiance about photon mapping for
> caustics reflection. It looks pretty complex and I think (and I hope) that
> I can solve my problem without using this.
>
> Best Regards,
>
> Antoine
>
>
>
> 2017-10-12 22:45 GMT+02:00 Jan Wienold <jan.wienold at epfl.ch>:
>
>> Hi Antoine,
>> specular reflections are not traced at all to light sources in the
>> ambient calculation (which you invoke by -I). Only to glow material. Thats
>> why also the reflections of the sun on glass-material is not captured by
>> rtrace -I.
>> So either you use mirror material and a -dr >2 or you calculate a
>> 180degree fisheye on you point and calculate the illuminace from the image.
>> The view rays for the image are traced via specular surfaces towards light
>> sources.
>> Jan
>>
>>
>> On 12 October 2017 19:22:28 CEST, Antoine Bugeat <bugeatantoine at gmail.com>
>> wrote:
>>>
>>> Hello all,
>>>
>>> I'm a civil engineering student and I'm trying to compute the impact of
>>> specular reflections in a street "canyon".
>>> My problem is:
>>> I use rtrace function with the basic parameters, the same as DAYSIM uses
>>> (-I -aa 0.1 -ab 5 -ad 1000 -ar 300 -as 20 -dj 0 -dr 5 -ds 0.2 -dp 512
>>> -lr 6 -lw 0.004 -st 0.1 -oov)
>>> My sky and sun description come from gendaylit.
>>> My wall is defined by a perfect specular plastic and the ground is
>>> defined by a diffuse material.
>>>
>>> When I do the simulation, I get illogical results as if the specular
>>> reflection of the walls had no impact.
>>>
>>> So, I tried to simulate a really simple scene with only one *specular
>>> surface*:
>>>
>>> void plastic M_Spec
>>> 0
>>> 0
>>> 5 0 0 0 1 0
>>>
>>> M_Spec polygon 1
>>> 0
>>> 0
>>> 12 -5.0 0.0 -5.0
>>> -5.0 0.0 5.0
>>> 5.0 0.0 5.0
>>> 5.0 0.0 -5.0
>>>
>>> My *source* is described as a sun perfectly perpendicular to the
>>> surface:
>>>
>>> void light solar
>>> 0
>>> 0
>>> 3 1000 1000 1000
>>>
>>> solar source sun
>>> 0
>>> 0
>>> 4 0.0 -1.0 0.0 0.5
>>>
>>> And the *sensor* is between the surface and the source and is regarding
>>> right to the surface:
>>>
>>> 0.0 -5.0 0.0 0.0 1.0 0.0
>>>
>>> And in this case, I get 0 after my rtrace simulation.
>>> I tried to vary some parameters of rtrace (direct and specular
>>> parameters) but it change nothing I still get 0.
>>>
>>> Thank you for your help.
>>>
>>> Antoine Bugeat
>>>
>>>
> _______________________________________________
> Radiance-general mailing list
> Radiance-general at radiance-online.org
> https://www.radiance-online.org/mailman/listinfo/radiance-general
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.radiance-online.org/pipermail/radiance-general/attachments/20171024/46781040/attachment-0001.html>