I see that quite many post radiosity images rendered without
assumed_gamma 1.0. It is required for the radiosity calculations
to simulate a physical situation. Non-physical situations however
couldn't care less about the gamma :)

See news://news.povray.org/39423CFE.5565DBF%40kivisalo.net

_______________________________________________________________________
Kari Kivisalo                                   http://www.kivisalo.net

Post a reply to this message

Are you saying it would be "bad" to use something like 10? :)

Grim

"Kari Kivisalo" <kar### [at] kivisalonet> wrote in message
news:39B5A4FB.BC7F0C94@kivisalo.net...
>
> I see that quite many post radiosity images rendered without
> assumed_gamma 1.0.

Post a reply to this message

using assumed_gamma 1.0 does bad things to my scenes... Is there anyone here
that *doesn't* use it, and can reinforced my theory that it isn't necessary?

"Kari Kivisalo" <kar### [at] kivisalonet> wrote in message
news:39B5A4FB.BC7F0C94@kivisalo.net...
>
> I see that quite many post radiosity images rendered without
> assumed_gamma 1.0. It is required for the radiosity calculations
> to simulate a physical situation. Non-physical situations however
> couldn't care less about the gamma :)
>
> See news://news.povray.org/39423CFE.5565DBF%40kivisalo.net
>
> _______________________________________________________________________
> Kari Kivisalo                                   http://www.kivisalo.net

Post a reply to this message

I don't normally use it. It takes a post from Kari to get me thinking about
it. I had forgotten about it until he posted the famous Cornell box scene.

Post a reply to this message

Ian Witham wrote:
> 
> using assumed_gamma 1.0 does bad things to my scenes... Is there anyone here
> that *doesn't* use it, and can reinforced my theory that it isn't necessary?

It isn't necessary if you don't wan't realistic radiosity. It is necessary
for realistic radiosity. Good images can be produced without but it will
be much harder to predict what the scene will look like and will force
unrealistic texture properties.

_______________________________________________________________________
Kari Kivisalo                                          www.kivisalo.net

Post a reply to this message

"Tony[B]" wrote:
> 
> I don't normally use it. It takes a post from Kari to get me thinking about
> it.

All povray lighting calculations are done using normal linear intensites.
Monitor output intensity follows roughly function out = in^gamma. If
a pixel in a scene should be at 50% intensity the actual value that
is fed to the monitor should be 50%^(1/gamma) = 70% if the monitor gamma = 2.
In other words an image using linear intensities must be gamma corrected
so that the monitor displays the image correctly. Display_Gamma in master
povray.ini is the gamma. global_settings{assumed_gamma 1.0} instructs
pov to apply the correction to the image file.

I don't know how this can be explained more clearly. Maybe like this:
Povray thinks that when it sets a pixel at 20% intensity you will also
see it at 20% intensity. Povray is wrong! You will see the pixel at
20%^gamma = 4% intensity. Clearly not good. Use gamma correction and you will
see what povray wanted you to see.


RTFM & Gamma FAQ folks. 

_______________________________________________________________________
Kari Kivisalo                                   http://www.kivisalo.net

Post a reply to this message

Kari Kivisalo wrote:
> 
> I don't know how this can be explained more clearly. Maybe like this:
> Povray thinks that when it sets a pixel at 20% intensity you will also
> see it at 20% intensity. Povray is wrong! You will see the pixel at
> 20%^gamma = 4% intensity. Clearly not good. Use gamma correction and you will
> see what povray wanted you to see.

Gamma's a bit of a bitch to explain, particularly since most people just
assume that all displays work the same; even if they see the screen
being lightened and darkened and someone tells them it's because of
gamma, they still don't have a clue what's really happening (although
after that demonstration, they might think they do, and assume it's the
same as brightness).

I'm still not even sure *I* understand it completely, and I've been
trying to. Here's a site that discusses gamma and its effects at some
length, though:

http://www.aim-dtp.net/aim/index.htm

After reading through that (and experimenting with what it says for
myself), I'm convinced that gamma (1) is evil, (2) sucks, and (3)
eventually needs to be standardized at 1.0, not 2.2 as in sRGB (and
THAT'S a whole rant in itself).

-Xplo

Post a reply to this message

Xplo Eristotle wrote:
>
> http://www.aim-dtp.net/aim/index.htm

That site is by a famous net kook so the information presented there
is questionable at least. His initial assumptions about linear/non-linear
coding are wrong which affects all his conclusions. The widely accepted
as correct presentation about gamma and color is

http://www.inforamp.net/~poynton/Poynton-color.html

Poyntons comments about the aim site author:
http://www.inforamp.net/~poynton/notes/Timo/Concerning_Timo.html


Finally, if you don't know who to trust see what the authors of POV-Ray
STRONGLY RECOMMEND. An exerpt from The Holy Manual of POV-Ray:
 ___________________________________________________________________
|                                                                   |
| If you are creating new scenes or rendering old scenes, it is     |
| strongly recommended that you put in an appropriate assumed_gamma |
| global setting. For new scenes, you should use an assumed gamma   |
| value of 1.0 as this models how light appears in the real world   |
| more realistically.                                               |
|___________________________________________________________________|


_______________________________________________________________________
Kari Kivisalo                                   http://www.kivisalo.net

Post a reply to this message

Kari Kivisalo wrote:
> 
> Xplo Eristotle wrote:
> >
> > http://www.aim-dtp.net/aim/index.htm
> 
> That site is by a famous net kook so the information presented there
> is questionable at least.

Hmm, interesting. Still, kook or not, much of what he says seems to hold
true in testing.

> The widely accepted
> as correct presentation about gamma and color is
> 
> http://www.inforamp.net/~poynton/Poynton-color.html

I'll have to have a look.

-Xplo

Post a reply to this message

Xplo Eristotle wrote:
>
> Still, kook or not, much of what he says seems to hold
> true in testing.

I agree about image processing operations should be applied
to linear samples but his "linear" is not linear in physical sense.
Long time I too thaught that when a gamma correction is removed
from an image you get linear samples. Not so if the image came
from any kind of capturing device. POV-Ray samples are always linear
though :)

_______________________________________________________________________
Kari Kivisalo                                          www.kivisalo.net

Post a reply to this message