I like Thomas' original suggestion of mapping an image to the window and using an illum with the appropriate sky distribution. If you simply use the fisheye.cal mapping I recommended earlier and apply it to a fisheye capture out the window, all should be well for any perspective view. However, you will need a fairly high resolution fisheye image to make this work. The description goes something like this:
# Define sky
!gensky 5 27 12
# Define high frequency view out window
void colorpict west_view
9 red green blue outsideHF.pic fisheye.cal fish_u fish_v -rz 90
0
# Apply view to sky only for West elevation
skyfunc mixfunc full_sky_pat
4 west_view void "if(Dx,0,-Dx^.4)" .
0
# Apply this to the sky
full_sky_pat glow sky_glow
0
4 1 1 1 0
sky_glow source sky
0
4 0 0 1 360
# Define window glazing
void glass window_glass
0
3 .92 .92 .92
# Window angle affects transmittance, so use approximation to this
skyfunc brightfunc win_func
2 winxmit winxmit.cal
0
# Define the window as an illum minus the exterior pattern
win_func illum window_illum
1 window_glass
0
3 1 1 1
I tried it, and it seems to work OK.
-Greg
···
From: Thomas Bleicher <[email protected]>
Date: May 27, 2008 1:48:40 PM PDT
On 26 May 2008, at 19:11, steve michel wrote:
It occured to me that since cyl.cal already exist, could mkillum the sky/ground light distribution be applied to the cylinder of a panaromic view? That would eliminate having to match the viewpoint angle with the picture. As you no doubt see, I want to avoid 'forking' renders and get the accuracy and presentation in one render.
According to the man page mkillum does not work for cylinders (it works for
spheres, though, but there is no point in creating a sphere to simulate the
sky).
Perhaps you could use Radzillas 'ghost' material. IIRC that was a material
that was visible but did not influence the lighting calculations. If this is true
you can create a cylinder around your scene without compromising the sky
distribution.
Regards,
Thomas
Yes the parallax issue of having to match the camera angle with the render view:
The fisheye pic seems like an elegant solution.
Ill work with that.. I just wanted to find a good method 'on paper' first before committing.
Plenty of fisheye pics out there to try this out with..
thanks again Thomas, Greg
Steve
···
Subject: Re: [Radiance-general] Re: panoramic images [was: applied material ...]> Date: Tue, 27 May 2008 20:12:23 -0700> To: [email protected]> From: [email protected]> > I like Thomas' original suggestion of mapping an image to the window > and using an illum with the appropriate sky distribution. If you > simply use the fisheye.cal mapping I recommended earlier and apply it > to a fisheye capture out the window, all should be well for any > perspective view. However, you will need a fairly high resolution > fisheye image to make this work. The description goes something like > this:> > # Define sky> !gensky 5 27 12> > # Define high frequency view out window> void colorpict west_view> 9 red green blue outsideHF.pic fisheye.cal fish_u fish_v -rz 90> 0> 0> > # Apply view to sky only for West elevation> skyfunc mixfunc full_sky_pat> 4 west_view void "if(Dx,0,-Dx^.4)" .> 0> 0> > # Apply this to the sky> full_sky_pat glow sky_glow> 0> 0> 4 1 1 1 0> > sky_glow source sky> 0> 0> 4 0 0 1 360> > # Define window glazing> void glass window_glass> 0> 0> 3 .92 .92 .92> > # Window angle affects transmittance, so use approximation to this> skyfunc brightfunc win_func> 2 winxmit winxmit.cal> 0> 0> > # Define the window as an illum minus the exterior pattern> win_func illum window_illum> 1 window_glass> 0> 3 1 1 1> > I tried it, and it seems to work OK.> > -Greg> > > From: Thomas Bleicher <[email protected]>> > Date: May 27, 2008 1:48:40 PM PDT> >> > On 26 May 2008, at 19:11, steve michel wrote:> >> >> It occured to me that since cyl.cal already exist, could mkillum > >> the sky/ground light distribution be applied to the cylinder of a > >> panaromic view? That would eliminate having to match the viewpoint > >> angle with the picture. As you no doubt see, I want to avoid > >> 'forking' renders and get the accuracy and presentation in one > >> render.> >> > According to the man page mkillum does not work for cylinders (it > > works for> > spheres, though, but there is no point in creating a sphere to > > simulate the> > sky).> >> > Perhaps you could use Radzillas 'ghost' material. IIRC that was a > > material> > that was visible but did not influence the lighting calculations. > > If this is true> > you can create a cylinder around your scene without compromising > > the sky> > distribution.> >> > Regards,> > Thomas> > _______________________________________________> Radiance-general mailing list> [email protected]> http://www.radiance-online.org/mailman/listinfo/radiance-general
_________________________________________________________________
Try Chicktionary, a game that tests how many words you can form from the letters given. Find this and more puzzles at Live Search Games!
http://g.msn.ca/ca55/207