Query regarding rcontrib

radiance-general
1

I have an application which currently uses rtrace to generate illuminance data for sensors, and I'm trying to change this to use rcontrib.

When I have sensors on the WP pointing up (see simple model http://upload.iesve.com/downloads/rcontrib/model.png) I get results that I'm happy with (results for rtrace - http://upload.iesve.com/downloads/rcontrib/sensor1_rtraceCIE.png and results for rcontrib - http://upload.iesve.com/downloads/rcontrib/sensor1_rcontribCIE.png).

When I move the sensors to the ceiling and pointing down (see revised model http://upload.iesve.com/downloads/rcontrib/model2.png) I get no results for rcontib (results for rtrace - http://upload.iesve.com/downloads/rcontrib/sensor1D_rtraceCIE.png and results for rcontrib - http://upload.iesve.com/downloads/rcontrib/sensor1D_rcontribCIE.png).

I've attached the data files (http://upload.iesve.com/downloads/rcontrib/files.zip) - the file rcontrib.bat is what I'm running.

[IES]<http://www.iesve.com/>

Don Stearn
Software Engineer

T:

+44 (0) 141 945 8500

F:

+44 (0) 141 945 8501

http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**

Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP

Email Disclaimer<http://www.iesve.com/disclaimer.html>

2

Hi Don,
Two things:
1 - You should use the same material for the sky and ground so that the
ground reflected component is included in patch 0. I typically use a single
360 degree source to encompass by sky and ground.
2 - Your zip file didn't include test_Model1.opt, used by the rcontrib
command, but make sure that -ab is 2 or greater to
get inter-reflected contribution from the sky on a ceiling illuminance
sensor (if -ab isn't sepcified in the opt file rcontrib defaults to -ab 1).

Let me know if these two changes don't give better results and I'll
investigate further.

Andy

···

On Tue, Jul 23, 2013 at 4:02 AM, Don Stearn <[email protected]> wrote:

I have an application which currently uses rtrace to generate
illuminance data for sensors, and I’m trying to change this to use
rcontrib.****

** **

When I have sensors on the WP pointing up (see simple model
http://upload.iesve.com/downloads/rcontrib/model.png\) I get results that
I’m happy with (results for rtrace -
http://upload.iesve.com/downloads/rcontrib/sensor1_rtraceCIE.png and
results for rcontrib -
http://upload.iesve.com/downloads/rcontrib/sensor1_rcontribCIE.png\).****

** **

When I move the sensors to the ceiling and pointing down (see revised
model http://upload.iesve.com/downloads/rcontrib/model2.png\) I get no
results for rcontib (results for rtrace -
http://upload.iesve.com/downloads/rcontrib/sensor1D_rtraceCIE.png and
results for rcontrib -
http://upload.iesve.com/downloads/rcontrib/sensor1D_rcontribCIE.png\).****

** **

I’ve attached the data files (
http://upload.iesve.com/downloads/rcontrib/files.zip\) – the file
rcontrib.bat is what I’m running.****

** **

** **

[image: IES] <http://www.iesve.com/&gt;\*\*\*\*

*Don Stearn*
*Software Engineer* ****

T:****

+44 (0) 141 945 8500****

F:****

+44 (0) 141 945 8501****

http://www.iesve.com ****

**Design, Simulate + Innovate with the *<Virtual Environment>*** ****

Integrated Environmental Solutions Limited. Registered in Scotland No.
SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow
G20 0SP****

Email Disclaimer <http://www.iesve.com/disclaimer.html&gt; ****

** **

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

3

Hi Don,

Two things:
1 - You should use the same material for the sky and ground so that the
ground reflected component is included in patch 0. I typically use a single
360 degree source to encompass by sky and ground.

I completely agree with Andy. Put this in your white_sky.sky:
void glow sky_glow
0
4 1 1 1 0

sky_glow source sky
0
4 0 0 1 360

2 - Your zip file didn't include test_Model1.opt, used by the rcontrib
command, but make sure that -ab is 2 or greater to
get inter-reflected contribution from the sky on a ceiling illuminance
sensor (if -ab isn't sepcified in the opt file rcontrib defaults to -ab 1).

@Andy
The file names are all messed up by Windows. It's test_Model2.opt that you're looking for. -ab is actually set to 5.

@Don,
you're overwriting rcontrib's default for -lr:
$ rcontrib -defaults
...
-lr -10 # limit reflection (Russian roulette)
...

Your setting it to -lr 8
Note the difference between positive and negative integers. I have to admit that this is one of the more obscure options that I don't tend to temper with. See man rpict. I ran your model with the -lr 8 removed from the opt file. Completely different outcome.

Also, your -ad is not very high. Unless you nudge it up a lot, there is no point in using 4 Reinhart subdivisions of the Tregenza patches. An -ad of 4096 is decent for a normal rtrace daylight simulation. For rcontrib, it should be much higher.

I don't know how you post-process the irradiance vector, since this script(?) is not included in the Zip file. However, you're calling rcontrib with the -V+ option. This gives you the *absolute* contribution from the sky patches, ie the irradiance at the sensor points.

One tends to use -V- here to get the daylight coefficients, ie the *relative* contribution from the sky patches. Once they have been calculated, the DC are multiplied with the sky vector to get actual irradiance values.

Regards

Axel

4

rcontrib.bat says '@TEST_M~1.OPT' and this references 'test_Model2.opt' ?
The caps and ~ aside, I don't know how a 1 becomes a 2. Windows never
ceases to amaze me...

Andy

···

On Tue, Jul 23, 2013 at 2:43 PM, Axel Jacobs <[email protected]> wrote:

@Andy
The file names are all messed up by Windows. It's test_Model2.opt that
you're looking for. -ab is actually set to 5.

5

Hi Andy,

rcontrib.bat says '@TEST_M~1.OPT' and this references 'test_Model2.opt' ?
  The caps and ~ aside, I don't know how a 1 becomes a 2. Windows never
ceases to amaze me...

I think in Windows DOS-shell speak, TEST_M~1.OPT means something like

"the first file in this directory that starts with 'TEST_M' and has a base name (ie excl 3-letter extension) that is longer than 8 chars."

Since there is no 'test_Model1.opt' in this directory, 'test_Model2.opt' gets the '~1' treatment. Weird, I agree.

UPPER/lower case is pretty arbitrary under Windows, although you can't create 'test.txt' if 'Test.TXT' or 'TEST.txt' already exist.

What do you make of the -lr option? This is a little too obscure for me...

Cheers

Axel

6

Thanks for the explanation. I thought the ~ was like a * character on
unix, but it is subtly different.

···

On Tue, Jul 23, 2013 at 4:50 PM, Axel Jacobs <[email protected]> wrote:

What do you make of the -lr option? This is a little too obscure for me...

I don't see how the -lr being positive instead of negative could cause all
the results to be zero, but there is probably a connection that eludes me.

Andy

7

Andy, Axel,

Thanks for your replies - very helpful.

I've taken the "-lr 8" parameter out, changed the "-V+" to "-V-" and modified the "white_sky.sky" file.
I'm now getting results.

I'm using the following (shown for 15th June at 11:00am) to get the data -

gensky 6 15 11.00 -c -a 51.48 -o 0.45 -m 0.00 | genskyvec.exe -m 4 > day.dat
dctimestep.exe sky.dat day.dat > output.dat

donS

[IES]<http://www.iesve.com/>

Don Stearn
Software Engineer

T:

+44 (0) 141 945 8500

F:

+44 (0) 141 945 8501

http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**

Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP

Email Disclaimer<http://www.iesve.com/disclaimer.html>