[Radiance-general] Dimming the artificial luminaire

[Gregory J. Ward]

Hi Prudhvitej,

The rcontrib calculation proceeds differently from rtrace, since it cannot use the indirect irradiance cache.  Therefore, the -aa setting is overridden in rcontrib to "-aa 0" and the rest of your parameters should change to compensate.  Try the following settings instead of the -a* settings you have for rcontrib:

	-ab 2 -ad 50000 -lw 1e-5

The -lw setting is important to change so that all the initial samples get sent.  Your results may still change a little from run to run, but they should resemble the rtrace results more closely.

Best,
-Greg

> From: Prudhvitej Immadi <tej.immadi at samsung.com>
> Subject: [Radiance-general] Fwd: Re:  Dimming the artificial luminaire
> Date: May 16, 2016 12:54:41 AM PDT
> 
> Revisting the thread after long,
> Thanks for the awesome advice in the previous queries. They were very helpful but I'm stuck in a new problem.
>  
> I wanted to use rtcontrib to find out individual contributions of each luminaire at each point in the plane.
> So, I used
> rcontrib.exe -V- -I -h -ab 2 -aa 0.1 -ad 1024 -M file .oct <input.dat >output.dat
> where input.dat has location(x,y,z) in the form x y z 0 0 1
> since I'm trying to find illuminance when measured in (0,0,1) direction.
>  
> The values  of illuminance I get from rcontrib and the values I get from rtrace and surprisngly rcontrib finishes it's calculations in far less time than rtrace.
> More interesting part is that the output from rcontrib changes everytime I run it though the change is very small. I'm curious why it should change ?
>  
> Am I doing something wrong ? Can you guide me from here ? I prefer to get values  similar to rtrace.
>  
> I've searched the forum for similar issues and I found out that rtcontrib uses pure monte carlo and rtrace uses ambient interpolation
> 
> How do I decide on which values are correct/reliable ?
>  
> Thanks a lot,
> Prudhvitej Immadi.
> ------- Original Message -------
> Sender : Gregory J. Ward<gregoryjward at gmail.com>
> Date : Apr 26, 2016 12:02 (GMT+09:00)
> Title : Re: [Radiance-general] Dimming the artificial luminaire
>  
> You will need to use a spreadsheet or custom program to add together illuminance values rather than pcomb, which is for images.  As for computing the dimming levels to achieve the desired illuminance on the work plane, that's a linear optimization problem that shouldn't be too hard to solve using least-squares or similar, but I don't have a ready-made tool to help with that.  Again, Excel or Matlab might be useful for that.
> 
> Cheers,
> -Greg
> 
>> From: Prudhvitej Immadi <tej.immadi at samsung.com>
>> Subject: Re: Re: [Radiance-general] Dimming the artificial luminaire
>> Date: April 25, 2016 7:49:44 PM PDT
>> 
>> Hi Greg,
>> Thanks a lot for the helpful response. I've been using rtrace -I for my illuminance calculations than radiance picures and can pcomb be used here ?
>> Also, my bigger concern was that I have to do complex mathematics to find out my dimming factors to get the required illuminance value since I need to ensure minimum certain illuminance at all points in the work plane. Is there a way that I can reverse-engineer this, i.e., find out dimming coefficients for a certain level of illuminance rather than finding out illuminance for certain dimming coefficients.
>>  
>> Thank you,
>> Prudhvitej Immadi.
>>  
>> ------- Original Message -------
>> Sender : Greg Ward<gregoryjward at gmail.com>
>> Date : Apr 26, 2016 02:06 (GMT+09:00)
>> Title : Re: [Radiance-general] Dimming the artificial luminaire
>>  
>> Hi Prudhvi,
>> 
>> You are correct that changing the luminaire output with ies2rad would be a very inefficient way to go about things.
>> 
>> The best method is to add together illuminance or image results for your separately controllable luminaires using the appropriate dimming factors.  If all of your luminaires are controlled by a single sensor, then this is equivalent to adding a scaled "electric lighting only" component to a "daylight only" component.  This can be accomplished with pcomb like so:
>> 
>> pcomb -o daylight.hdr -s 0.7 -o electric.hdr > total.hdr
>> 
>> In this example, I assume that your dimming level is set to 70% output.  If you have three separately controlled luminaire banks, your command might look like this:
>> 
>> pcomb -o daylight.hdr -s 0.25 -o bank1.hdr -s 0.5 -o bank2.hdr -s 0.65 -o bank3.hdr > total.hdr
>> 
>> Here, I have set bank1 to 25%, bank2 to 50%, and bank3 to 65% output.  To compute each of these banks, you will need to render your scene without daylight and with the other banks turned off, somehow.
>> 
>> If you are computing illuminances rather that images, you can use a spreadsheet or similar to apply your dimming factors to the calculated illuminance components in a similar fashion.
>> 
>> If you have individually controlled luminaires, it might be more efficient to use rcontrib to compute the contribution of each luminaire without having to perform multiple rendering calculations.  This is a bit more involved, but you might be able to figure it out from the examples given in my talk from the 2012 Radiance workshop:
>> 
>> http://www.radiance-online.org/community/workshops/2012-copenhagen/Day1/Ward/Rtcontrib_tut1.pdf
>> 
>> I hope this is enough to get you started.
>> 
>> Best,
>> -Greg
>> 
>>> From: Prudhvitej Immadi <tej.immadi at samsung.com>
>>> Subject: [Radiance-general] Dimming the artificial luminaire
>>> Date: April 25, 2016 1:24:23 AM PDT
>>> Hi,
>>> I'm trying to simulate an open office environment using radiance. These days, there are luminaires which have photosensors installed available in the market. They dim automatically to the 500lux illuminance level (Standard for office spaces) when daylight is available. 
>>> Is it possible to simulate such a scenario using radiance. My aim is to find the dimming level when the 500lux is achieved on the work plane. 
>>> 1. How is dimming the luminaries achieved ?
>>> From what I gathered, I can have a multiplication factor in ies2rad. I think this is ineffecient because I will need to generate an octree again. I couldn't find a better alternative.
>>> 2. Is there a way to automatically dim the luminaire to meet a certain criteria on illuminance ?
>>> 
>>> I'm relatively new to radiance and pardon me if this is a trivial issue. 
>>> 
>>> Any help would be greatly appreciated.
>>> Thank you,
>>> Prudhvi Tej.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.radiance-online.org/pipermail/radiance-general/attachments/20160516/eb2cc824/attachment-0001.html>