warning - outgoing BSDF direction 142 collects 107.4% of light

radiance-general
1

Hi Greg,

Yes, you understand my email correctly.
I have just find that the material is probably not only the reason. I have
the window split into parts parts and only BSDF for bottom part (down)
generate this warning. Here is the command for both parts of window, and
only difference is the bounding box, which I assume is correct.

### this one produce the warning
genBSDF -c 1000 -n 8 -r @BSDFparameters.opt +forward -backward -geom -dim 0
1.98 0 1.13 -0.01 0 \
geometry/materialNew.mat geometry/30_geometry.rad >
results/BSDFdown_30_new.xml

genBSDF -c 1000 -n 8 -r @BSDFparameters.opt +forward -backward -geom -dim 0
1.98 1.13 2.26 -0.01 0 \
geometry/materialNew.mat geometry/30_geometry.rad >
results/BSDFup_30_new.xml

I am actually using dctimestepcpu instead of dctimestep because it is much
faster, but I guess it is updated too.

I have got this message before but with dctimestep and not dctimestepcpu. I
fixed that in that time but I am not sure what was wrong, I guess it was
something with wrong geometry and there were many more dirrection which
collected higher %.

My question could be also understand as: Would this error also appear when
there is light concentration in one spot after it is reflected?

Thank you
David

···

On 23 May 2011 08:32, Greg Ward <[email protected]> wrote:

Hi David,

If I understand your e-mail correctly, you are seeing this error using the
NEW material, not the OLD one.

I cannot reproduce your error. I assume from your earlier e-mails that you
are using the latest HEAD of Radiance. You should make sure that everything
is up-to-date and try it again. In particular, we fixed some problems last
December in the normalization of results. The dctimestep program has also
been improved and corrected.

-Greg

*From: *David Appelfeld <[email protected]>

*Date: *May 23, 2011 7:51:13 AM PDT

*
*

Dear Radiance users,
I am getting warning message

warning - outgoing BSDF direction 142 collects 107.4% of light

when multiplying matrixes by dctimestep. I generated BSDF matrix by
genBSDF.
I guess that I can still use it but is this message saying that it could be
wrong or is it more just like a statement that there is literally more light
coming to direction 142.
I generated more of BSDF matrexis and I kept same geometry and changed only
one material which has higher reflectance. I guess that it can be reason but
I would like to be sure that the BSDF matrix is still reliable.

Here is the material description for OLD and NEW (with warning message when
used)
NEW material is generated by "optics2rad"
OLD material is generated by script "glaze" and "optics2glazedb"

Thank you very much I very appreciate your help
David

void BRTDfunc OLD
10
    sr_frit_r sr_frit_g sr_frit_b
    st_frit_r st_frit_g st_frit_b
    0 0 0
    glaze1.cal
0
11
0.0564
0.0492
0.051
0.126
0.126
0.126
0.054
0.054
0.054
    1 .6

void BRTDfunc NEW
10
    rR_clear rG_clear rB_clear
    0.678*tR_clear 0.655*tG_clear 0.576*tB_clear
     0 0 0
     window.cal
0
15 0 0 0 0 0 0 0 0 0
    0.267 0.321 0.395
    0.24 0.306 0.368

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

2

Hi Greg,

I will try to run the multiplication again with dctimestep and I will
compare them.

Thank you
David

···

On 23 May 2011 10:08, Gregory J. Ward <[email protected]> wrote:

Hi David,

I don't get the warning using your XML file. I suspect there is something
amiss with dctimestepcpu, which I don't know anything about. Maybe it was
derived from earlier code that summed up the BSDF improperly?

Cheers,
-Greg

*From: *David Appelfeld <[email protected]>

*Date: *May 23, 2011 9:50:43 AM PDT

*
*

Hi Greg

here is the file. This matrix is only the difference compare to other
calculations.

I really appreciate your help with my problems with Radiance.

Thank you
David

On 23 May 2011 09:35, Greg Ward <[email protected]> wrote:

Hi David,

I don't really know the source of this warning. Concentration of light
shouldn't affect genBSDF, as it measures everything at infinity. Why don't
you e-mail me your problem XML file in an off-list message, and I'll see if
I can spot anything.

-Greg

> From: David Appelfeld <[email protected]>
> Date: May 23, 2011 8:53:33 AM PDT
>
> Hi Greg,
>
> Yes, you understand my email correctly.
> I have just find that the material is probably not only the reason. I
have the window split into parts parts and only BSDF for bottom part (down)
generate this warning. Here is the command for both parts of window, and
only difference is the bounding box, which I assume is correct.
>
> ### this one produce the warning
> genBSDF -c 1000 -n 8 -r @BSDFparameters.opt +forward -backward -geom
-dim 0 1.98 0 1.13 -0.01 0 \
> geometry/materialNew.mat geometry/30_geometry.rad >
results/BSDFdown_30_new.xml
>
> genBSDF -c 1000 -n 8 -r @BSDFparameters.opt +forward -backward -geom
-dim 0 1.98 1.13 2.26 -0.01 0 \
> geometry/materialNew.mat geometry/30_geometry.rad >
results/BSDFup_30_new.xml
>
>
>
> I am actually using dctimestepcpu instead of dctimestep because it is
much faster, but I guess it is updated too.
>
> I have got this message before but with dctimestep and not
dctimestepcpu. I fixed that in that time but I am not sure what was wrong, I
guess it was something with wrong geometry and there were many more
dirrection which collected higher %.
>
> My question could be also understand as: Would this error also appear
when there is light concentration in one spot after it is reflected?
>
> Thank you
> David
>
>
> On 23 May 2011 08:32, Greg Ward <[email protected]> wrote:
> Hi David,
>
> If I understand your e-mail correctly, you are seeing this error using
the NEW material, not the OLD one.
>
> I cannot reproduce your error. I assume from your earlier e-mails that
you are using the latest HEAD of Radiance. You should make sure that
everything is up-to-date and try it again. In particular, we fixed some
problems last December in the normalization of results. The dctimestep
program has also been improved and corrected.
>
> -Greg