# color calculations in mgfilt

**URL:** <https://discourse.radiance-online.org/t/color-calculations-in-mgfilt/3215>\
**Category:** radiance-general\
**Created:** [November 15, 2013, 11:27pm UTC](https://discourse.radiance-online.org/t/color-calculations-in-mgfilt/3215 "2013-11-15T23:27:49Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Lars\_Grobe](https://discourse.radiance-online.org/user_avatar/discourse.radiance-online.org/lars_grobe/32/469_2.png) [@Lars\_Grobe](https://discourse.radiance-online.org/u/Lars_Grobe)\
**Post date:** [November 15, 2013, 11:27pm UTC](https://discourse.radiance-online.org/t/color-calculations-in-mgfilt/3215/1 "2013-11-15T23:27:49Z")

</div>

Hi,

mgfilt is not really part of the core Radiance package, but typically  
used with Radiance. So I hope this is not entirely off-topic.

I have been using mgfilt from time to time as a convenient tool for  
getting color coordinates from measured reflection spectra. However some  
results appeared a bit strange, and thus I did a comparision with what I  
get from mgfilt and the results of using the CIE formulae with rcalc  
instead. To make it short, the results I got from mgfilt were quite off,  
and I am wondering wether I did something wrong or the accuracy  
implemented in mgfilt is really that much worse.

I uploaded a short report on the comparison here:

[http://static.larsgrobe.de/spec2Yxy-20131115.pdf](http://static.larsgrobe.de/spec2Yxy-20131115.pdf)

Cheers, Lars.

---

<div class="post-metadata">

**Author:** ![Greg\_Ward1](https://discourse.radiance-online.org/letter_avatar_proxy/v4/letter/g/3be4f8/32.png) [@Greg\_Ward1](https://discourse.radiance-online.org/u/Greg_Ward1)\
**Post date:** [November 16, 2013, 2:49am UTC](https://discourse.radiance-online.org/t/color-calculations-in-mgfilt/3215/2 "2013-11-16T02:49:03Z")

</div>

Hi Lars,

I'm pretty sure mgfilt does its calculation with an equal-energy white, not D65 or any of the other standard illuminants. Try repeating it with EE white to see what you get.

Cheers,  
-Greg

> **···**
>
> Sent from my iPad
> 
> > On Nov 15, 2013, at 3:27 PM, "Lars O. Grobe" \<grobe@gmx.net\> wrote:
> > 
> > Hi,
> > 
> > mgfilt is not really part of the core Radiance package, but typically  
> > used with Radiance. So I hope this is not entirely off-topic.
> > 
> > I have been using mgfilt from time to time as a convenient tool for  
> > getting color coordinates from measured reflection spectra. However some  
> > results appeared a bit strange, and thus I did a comparision with what I  
> > get from mgfilt and the results of using the CIE formulae with rcalc  
> > instead. To make it short, the results I got from mgfilt were quite off,  
> > and I am wondering wether I did something wrong or the accuracy  
> > implemented in mgfilt is really that much worse.
> > 
> > I uploaded a short report on the comparison here:
> > 
> > [http://static.larsgrobe.de/spec2Yxy-20131115.pdf](http://static.larsgrobe.de/spec2Yxy-20131115.pdf)
> > 
> > Cheers, Lars.
