Hi all,
I’m trying to extract a concentrated light source from HDR picture of the sky using mksource.
I run the following script:
oconv materials.mat geometry.rad lightprobe.rad > scene.oct
mksource -d 1000000 -t 2 -a 5 scene.oct > srcs.rad
oconv -i scene.oct srcs.rad ground.rad > finalscene.oct
rpict -vth -vv 180 -vh 180 -vp 1.8 3 1.2 -vd 1 1 0 finalscene.oct > finalsceneview.hdr
ra_tiff finalsceneview.hdr finalsceneview.tif
My results are weird (see https://www.dropbox.com/s/2u4vnsi4ynw0ncz/finalsceneview02.tif)
Do you have any idea what is wrong? a problem with the capture of the sun? or rather a matter of parameters ?
Is it normal that mksource finds several sources ? (see https://www.dropbox.com/s/n7y97ppm1ip1gky/srcs.rad)
Thanks in advance for your help !
Beatriz Piderit M.
Arquitecta UBB, MA, PhD UC Louvain
Departamento Diseño y Teoría de la Arquitectura
Universidad del Bio-Bio Concepción, Chile
Hi Maria,
Thank you for following good list etiquette and posting your image on a website. You may want to check to be sure that you are subscribed with the HDRI mailing list <http://www.radiance-online.org/community/mailing-lists>, as I had to manually pass your post onto the list. We don't seem to have you in our system, at least not at this address.
I can't say seeing your light probe which is just shown as white out the window, but mksource will tend to create multiple sources for long, skinny regions. Reducing the -d setting will reduce the maximum number of sources created in this way. Your threshold setting also seems to be quite low, and you might consider letting mksource determine this threshold automatically.
The general idea with mksource is to replace only the very brightest regions in a distant environment map with direct sources. The rest will be handled well enough in the indirect calculation. Which brings me to my last point -- are you giving mksource the room description as well as the exterior environment? It completely ignores the local geometry, using only your environment map (light probe) to decide where to place sources. Perhaps you could post that?
Best,
-Greg
···
From: "María Beatriz Piderit" <[email protected]>
Date: January 20, 2014 10:36:53 AM PST
Hi all,
I’m trying to extract a concentrated light source from HDR picture of the sky using mksource.
I run the following script:
oconv materials.mat geometry.rad lightprobe.rad > scene.oct
mksource -d 1000000 -t 2 -a 5 scene.oct > srcs.rad
oconv -i scene.oct srcs.rad ground.rad > finalscene.oct
rpict -vth -vv 180 -vh 180 -vp 1.8 3 1.2 -vd 1 1 0 finalscene.oct > finalsceneview.hdr
ra_tiff finalsceneview.hdr finalsceneview.tif
My results are weird (see https://www.dropbox.com/s/2u4vnsi4ynw0ncz/finalsceneview02.tif)
Do you have any idea what is wrong? a problem with the capture of the sun? or rather a matter of parameters ?
Is it normal that mksource finds several sources ? (see https://www.dropbox.com/s/n7y97ppm1ip1gky/srcs.rad)
Thanks in advance for your help !
Beatriz Piderit M.
Arquitecta UBB, MA, PhD UC Louvain
Departamento Diseño y Teoría de la Arquitectura
Universidad del Bio-Bio Concepción, Chile
Hi Greg,
Thanks for your response.
I've also tried to run the simulation using mksource default parameters, I
obtained this picture but it seems that there is too many sources identified
(see finalsceneview02.tif in
https://www.dropbox.com/sh/6ra7xtansk8rdz0/z0aG1qlTHr ). I thought that in
increasing -d, the problem disappears. In this simulation, I gave the room
description but the ground is not specified because I was not sure on how to
describe it in IBL.
In the following simulation I added the ground in describing it using
gensky... I am not sure that it is a good idea (see finalsceneview02.tif in
https://www.dropbox.com/sh/4gu3onc08d933ea/xIwc2E9Kc6 ).
In these two folders you also will find the local geometry, the light probe
image...
Thanks,
Beatriz
···
El 20-01-2014, a las 16:28, Gregory J. Ward <[email protected]> escribió:
Hi Maria,
Thank you for following good list etiquette and posting your image on a website. You may want to check to be sure that you are subscribed with the HDRI mailing list <http://www.radiance-online.org/community/mailing-lists>, as I had to manually pass your post onto the list. We don't seem to have you in our system, at least not at this address.
I can't say seeing your light probe which is just shown as white out the window, but mksource will tend to create multiple sources for long, skinny regions. Reducing the -d setting will reduce the maximum number of sources created in this way. Your threshold setting also seems to be quite low, and you might consider letting mksource determine this threshold automatically.
The general idea with mksource is to replace only the very brightest regions in a distant environment map with direct sources. The rest will be handled well enough in the indirect calculation. Which brings me to my last point -- are you giving mksource the room description as well as the exterior environment? It completely ignores the local geometry, using only your environment map (light probe) to decide where to place sources. Perhaps you could post that?
Best,
-Greg
From: "María Beatriz Piderit" <[email protected]>
Date: January 20, 2014 10:36:53 AM PST
Hi all,
I’m trying to extract a concentrated light source from HDR picture of the sky using mksource.
I run the following script:
oconv materials.mat geometry.rad lightprobe.rad > scene.oct
mksource -d 1000000 -t 2 -a 5 scene.oct > srcs.rad
oconv -i scene.oct srcs.rad ground.rad > finalscene.oct
rpict -vth -vv 180 -vh 180 -vp 1.8 3 1.2 -vd 1 1 0 finalscene.oct > finalsceneview.hdr
ra_tiff finalsceneview.hdr finalsceneview.tif
My results are weird (see https://www.dropbox.com/s/2u4vnsi4ynw0ncz/finalsceneview02.tif)
Do you have any idea what is wrong? a problem with the capture of the sun? or rather a matter of parameters ?
Is it normal that mksource finds several sources ? (see https://www.dropbox.com/s/n7y97ppm1ip1gky/srcs.rad)
Thanks in advance for your help !
Beatriz Piderit M.
Arquitecta UBB, MA, PhD UC Louvain
Departamento Diseño y Teoría de la Arquitectura
Universidad del Bio-Bio Concepción, Chile