I'm still playing around with Lightsys, and the more I learn, the more 
questions I have.  :-)

I did some tests with some of the demo scenes for Lightsys 4, where I 
replaced the reflective spectral data rgb values with plain-old rgb values. 
For the items I tested with, I didn't notice any big differences, but 
perhaps there are small differences which are more subtle.

My question(s) at this point:
What are the benefits of using the reflective spectral data (e.g. 
RS_White_Paint_1) over normal pigments?  Is it simply that the RS values 
will look better under a much wider range of lighting?  Is it ok to mix 
normal rgb pigments (e.g. rgb <1,0.8, 0.6>) along with the RS values?

Lightsys is an amazing set of includes/macros which everyone should probably 
be using, but at this point, I don't understand it very well.  Some day, it 
would be nice to have a good tutorial "Lightsys for Dummies" or something. 
Perhaps I'll begin writing it up, if I ever get to the point that I truly 
understand enough of it.

As always, any insight would be appreciated.

-- 
Jeremy
www.beantoad.com

Post a reply to this message

Jeremy M. Praay wrote:

> I'm still playing around with Lightsys, and the more I learn, the more 
> questions I have.  :-)

   You're not the only one! :) Seriously, I only had the initial idea and
zero knowledge about all the matters involved. Ive and others helped a
lot, and the set of macros can do a lot of things, but I really don't
understand all of them.

> My question(s) at this point:
> What are the benefits of using the reflective spectral data (e.g. 
> RS_White_Paint_1) over normal pigments?  Is it simply that the RS values 
> will look better under a much wider range of lighting?  Is it ok to mix 
> normal rgb pigments (e.g. rgb <1,0.8, 0.6>) along with the RS values?

   Technically, the advantage is that you have sampled colors, wich
should look realistic per se. Also, you have "abstract" spectrums that
you can "visualize" under any color space and with any white point, so
you can comunicate colors accuratelly, but this is not of much use with
POV-Ray.

   Practically, the important point is that you don't have to bother
making colors for pigments of known materials. I find it specially
useful with colors wich are difficult to get right, like the green
pigments of plants. It accomplishes the same goal as sampling the colors
from a photo, but with the advantage that the colors are not "changed"
by the lighting on the photo, the camera white balance, etc...

   Mixing with non-sampled colors is not a problem, but you should
adjust them to look similar to the sampled ones in brightness and
saturation, so it's better to use one of both types only.

> Lightsys is an amazing set of includes/macros which everyone should probably 
> be using, but at this point, I don't understand it very well.  Some day, it 
> would be nice to have a good tutorial "Lightsys for Dummies" or something. 
> Perhaps I'll begin writing it up, if I ever get to the point that I truly 
> understand enough of it.

   Yeah... I've that tutorial on the to-do list, who know for how many
years... as I mentioned, the main problem is that I don not fully
understand all the functionality (appart from lazzines, that is).

> As always, any insight would be appreciated.

   Hope that helped as a start...

--
Jaime

Post a reply to this message

"jaime" <jai### [at] ignoranciaorg> wrote in message 
news:422f4a28$1@news.povray.org...
<snip>
>
>   Practically, the important point is that you don't have to bother
> making colors for pigments of known materials. I find it specially
> useful with colors wich are difficult to get right, like the green
> pigments of plants. It accomplishes the same goal as sampling the colors
> from a photo, but with the advantage that the colors are not "changed"
> by the lighting on the photo, the camera white balance, etc...
>
>   Mixing with non-sampled colors is not a problem, but you should
> adjust them to look similar to the sampled ones in brightness and
> saturation, so it's better to use one of both types only.
>

Ok, so there wasn't a lot that I wasn't understanding in this regard.  The 
RS samples can be good to reference against, and they can be used to simply 
avoid having to spend a lot of time figuring out a certain color.  That 
makes sense to me.

I wasn't sure if I was really doing something "wrong" by using RGB values. 
There are so many other things that we guess at (like diffuse, brilliance, 
specular, etc.), that the RS is only one small piece.

>> Lightsys is an amazing set of includes/macros which everyone should 
>> probably be using, but at this point, I don't understand it very well. 
>> Some day, it would be nice to have a good tutorial "Lightsys for Dummies" 
>> or something. Perhaps I'll begin writing it up, if I ever get to the 
>> point that I truly understand enough of it.
>
>   Yeah... I've that tutorial on the to-do list, who know for how many
> years... as I mentioned, the main problem is that I don not fully
> understand all the functionality (appart from lazzines, that is).

It might be something that we could get started on the wiki, also.  Just 
another thought...

>
>> As always, any insight would be appreciated.
>
>   Hope that helped as a start...
>

It does!  I'm having a lot of fun messing around with lemon walls 
(demo_indoor1) among other things, like varying the Daylight/Sun 
temperature.  I've also learned a few other tricks from that demo which are 
quite useful, like putting media inside the window.

Thanks!  :-)

-- 
Jeremy M. Praay
www.beantoad.com

Post a reply to this message

In article <422f4268$1@news.povray.org>,
 "Jeremy M. Praay" <jer### [at] questsoftwarecom> wrote:

> What are the benefits of using the reflective spectral data (e.g. 
> RS_White_Paint_1) over normal pigments?  Is it simply that the RS values 
> will look better under a much wider range of lighting?  Is it ok to mix 
> normal rgb pigments (e.g. rgb <1,0.8, 0.6>) along with the RS values?

I'm not sure what Lightsys uses, but it sounds like it simply uses a 
greater number of samples for color. Real world color is a mess of 
wavelengths from a continuous spectrum. Not all light sources emit 
evenly through the entire visible spectrum, and most materials reflect 
very unevenly, with complicated variations of reflection as a function 
of wavelength. The same object can have very different appearances under 
two lights that both appear to be "white". The effect is especially 
important in highly monochromatic light, such as that from low-pressure 
sodium street lights. 3 channels simply aren't enough to simulate this 
kind of effect. The effects can be subtle or glaringly obvious, 
depending on the situation.

-- 
Christopher James Huff <cja### [at] gmailcom>
POV-Ray TAG: <chr### [at] tagpovrayorg>
http://tag.povray.org/

Post a reply to this message

"Christopher James Huff" <cja### [at] gmailcom> wrote in message 
news:cjameshuff-87CB6C.19220409032005@news.povray.org...
> In article <422f4268$1@news.povray.org>,
> "Jeremy M. Praay" <jer### [at] questsoftwarecom> wrote:
>
>> What are the benefits of using the reflective spectral data (e.g.
>> RS_White_Paint_1) over normal pigments?  Is it simply that the RS values
>> will look better under a much wider range of lighting?  Is it ok to mix
>> normal rgb pigments (e.g. rgb <1,0.8, 0.6>) along with the RS values?
>
> I'm not sure what Lightsys uses, but it sounds like it simply uses a
> greater number of samples for color. Real world color is a mess of
> wavelengths from a continuous spectrum. Not all light sources emit
> evenly through the entire visible spectrum, and most materials reflect
> very unevenly, with complicated variations of reflection as a function
> of wavelength. The same object can have very different appearances under
> two lights that both appear to be "white". The effect is especially
> important in highly monochromatic light, such as that from low-pressure
> sodium street lights. 3 channels simply aren't enough to simulate this
> kind of effect. The effects can be subtle or glaringly obvious,
> depending on the situation.
>


Thanks.  That's along the lines of what I was thinking.  As a real-world 
example, where I work, I sit under some strange yellow flourescent lights. 
Some of the papers strewn around my desk look white, others look very yellow 
or grey.  If I pick up those same papers and take them into my boss's office 
(different type of flourescent lights), they all look white.

Like I said to Jaime, I would guess that the reflectance values in POV-Ray 
probably are less significant than some of the other things we guess at in 
the finish statement (or at least I guess at them), like specular, phong, 
reflection, diffuse, brilliance, etc.  I think it would be nice if POV-Ray 
had more materials predefined using more physically accurate data.  Lightsys 
is a great step in that direction.  Does anyone know where to get 
information on those finish values?

-- 
Jeremy
www.beantoad.com

Post a reply to this message

Jeremy M. Praay wrote:
> Thanks.  That's along the lines of what I was thinking.  As a real-world 
> example, where I work, I sit under some strange yellow flourescent lights. 
> Some of the papers strewn around my desk look white, others look very yellow 
> or grey.  If I pick up those same papers and take them into my boss's office 
> (different type of flourescent lights), they all look white.

   That is metamerism, IIRC... and that's not possible without the 
raytracer handling wavelengths internally (that is, Lightsys can't 
simulate it).

> Like I said to Jaime, I would guess that the reflectance values in POV-Ray 
> probably are less significant than some of the other things we guess at in 
> the finish statement (or at least I guess at them), like specular, phong, 
> reflection, diffuse, brilliance, etc.  I think it would be nice if POV-Ray 
> had more materials predefined using more physically accurate data.  Lightsys 
> is a great step in that direction.  Does anyone know where to get 
> information on those finish values?

   The only pointer I got is reflection+diffuse=1. I really don't know 
where it comes from, but I've seen it mentioned several times by people 
which seems to know what they are talking about... and indeed seems 
logical and gives good results. For the rest, we still have to use the 
good and old "trial & error", I suppose.

--
Jaime

Post a reply to this message

"Jaime Vives Piqueres" <jai### [at] ignoranciaorg> wrote in message 
news:4232c60d$1@news.povray.org...
>   The only pointer I got is reflection+diffuse=1. I really don't know 
> where it comes from, but I've seen it mentioned several times by people 
> which seems to know what they are talking about... and indeed seems 
> logical and gives good results. For the rest, we still have to use the 
> good and old "trial & error", I suppose.
>


I don't think I've heard that before, or else if I did, I never paid 
attention.  I've never quite been sure what to use for diffuse, so I've 
generally just left it at the default.  I'm going to have to modify some of 
my scenes/objects and re-render them with that in mind.

Thanks!  :-)

-- 
Jeremy M. Praay
www.beantoad.com

Post a reply to this message

the set of macros can do a lot of things, but I really don't
understand all of them

I like what I've been able to do with lightsys so far, but that has only
been playing with the settings some and hacking .obj meshes into the demo
scenes. I'd like to at least understand what macros you do understand and
what they can do. :O)

Post a reply to this message

fls13 wrote:
> the set of macros can do a lot of things, but I really don't
> understand all of them

   You're not alone... and I'm not kidding. I was going only for a 
simple  automatic lighting system, but finally Ive implemented an entire 
"color management" system, which Lightsys uses to convert spectrums into 
rgb colors (which is only a little part of what the includes can do).

   I'm afraid that to use all the set of macros properly, you need to 
have first some good knowledge about professional/scientific color 
management (which I don't have). Anyhow, for regular scenes you can do 
as me: just use the lightsys macros with the predefined spectrums, and 
play a bit with the brightness and the white point to get the feel you want.

--
Jaime

Post a reply to this message

Anyhow, for regular scenes you can do
> as me: just use the lightsys macros with the predefined spectrums, and
> play a bit with the brightness and the white point to get the feel you want.

Ah, and how do you do that? :O) I'd be happy to write a tutorial on that
alone if I understood how. I set up a page with some figure meshes hacked
into the demos and the camera moved about.
http://www.geocities.com/fls13ec/PovTest-Lightsys4.html

For example, what do the .inc files actually include? I see in
lightsys_constants.inc file certain brands of light bulbs are assigned a
certain amount of lumens. Then there is a relationship with a spline that
is defined in either espd_cie_standard.inc or espd_lightsys.inc. At least I
think. How do you insert in a scene a GE 100 watt lightbulb in a particular
location, like a lamp? Or how about sunlight streaming through a window
into a room?

Post a reply to this message