Got a wild hair tonight and tried a negative value in the
assumed_gamma global setting. Interesting effect. The first
image is how it would normaly appear using assumed gamma 2.2.
The second is assumed gamma -1.0. There must be an application
for this but at the moment I can't think of one....


-- 
Ken Tyler - 1400+ POV-Ray, Graphics, 3D Rendering, and Raytracing Links:
http://home.pacbell.net/tylereng/index.html http://www.povray.org/links/

Post a reply to this message

Attachments:
Download 'plus-22.jpg' (8 KB) Download 'minus-1.jpg' (24 KB)

Preview of image 'plus-22.jpg'

Preview of image 'minus-1.jpg'

Maybe, if you are going to print some blacklight images....

Grim

Post a reply to this message

I assume it won't work with AA ...


--

 ,',    Jan Walzer      \V/  http://wa.lzer.net     ,',
',','   student of      >|<  mailto:jan### [at] lzernet ',','
  '   ComputerScience   /A\  +49-177-7403863         '

Ken <tyl### [at] pacbellnet> schrieb in im Newsbeitrag:
3A0E4914.9E48C88E@pacbell.net...
>
> Got a wild hair tonight and tried a negative value in the
> assumed_gamma global setting. Interesting effect. The first
> image is how it would normaly appear using assumed gamma 2.2.
> The second is assumed gamma -1.0. There must be an application
> for this but at the moment I can't think of one....
>
>
> --
> Ken Tyler - 1400+ POV-Ray, Graphics, 3D Rendering, and
Raytracing Links:
> http://home.pacbell.net/tylereng/index.html
http://www.povray.org/links/


-----------------------------------------------------------------
---------------






-----------------------------------------------------------------
---------------

Post a reply to this message

In article <3a0ec1d6@news.povray.org>, "Jan Walzer" <jan### [at] lzernet> 
wrote:

> I assume it won't work with AA ...

Gamma correction has nothing to do with antialiasing...

-- 
Christopher James Huff
Personal: chr### [at] maccom, http://homepage.mac.com/chrishuff/
TAG: chr### [at] tagpovrayorg, http://tag.povray.org/

<><

Post a reply to this message

I don't know the source of POV, but

Looking at the picture I think the effect comes from some
clipping-effects of the different rgb-values or some
non-continueities (you know what I mean) of the Gamma-function
with negative values ... (How is gamma defined for negative
values ???)

I think, the question is: When is it applied? Before or after AA
??

--

 ,',    Jan Walzer      \V/  http://wa.lzer.net     ,',
',','   student of      >|<  mailto:jan### [at] lzernet ',','
  '   ComputerScience   /A\  +49-177-7403863         '

Chris Huff <chr### [at] maccom> schrieb in im Newsbeitrag:
chrishuff-EC1E14.12010512112000@news.povray.org...
> In article <3a0ec1d6@news.povray.org>, "Jan Walzer"
<jan### [at] lzernet>
> wrote:
>
> > I assume it won't work with AA ...
>
> Gamma correction has nothing to do with antialiasing...
>
> --
> Christopher James Huff
> Personal: chr### [at] maccom, http://homepage.mac.com/chrishuff/
> TAG: chr### [at] tagpovrayorg, http://tag.povray.org/
>
> <><

Post a reply to this message

In article <3a0ed1ad@news.povray.org>, "Jan Walzer" <jan### [at] lzernet> 
wrote:

> Looking at the picture I think the effect comes from some
> clipping-effects of the different rgb-values or some
> non-continueities (you know what I mean) of the Gamma-function
> with negative values ...

I think the gamma correction uses a continuous, non periodic function, 
probably just a simple polynomial. The results you see are probably from 
the color values being clipped and otherwise messed up by the function.


> I think, the question is: When is it applied? Before or after AA??

Gamma correction would have to be applied to the final result of 
computing a pixel, correcting each sample wouldn't give the right 
result. Antialiasing has little effect on the image, it mostly makes a 
small difference in edges, so the change would be minor.

-- 
Christopher James Huff
Personal: chr### [at] maccom, http://homepage.mac.com/chrishuff/
TAG: chr### [at] tagpovrayorg, http://tag.povray.org/

<><

Post a reply to this message

Chris Huff <chr### [at] maccom> schrieb in im Newsbeitrag:
chrishuff-7E9319.13261612112000@news.povray.org...
> In article <3a0ed1ad@news.povray.org>, "Jan Walzer"
<jan### [at] lzernet>
> wrote:
>
> > Looking at the picture I think the effect comes from some
> > clipping-effects of the different rgb-values or some
> > non-continueities (you know what I mean) of the
Gamma-function
> > with negative values ...
>
> I think the gamma correction uses a continuous, non periodic
function,
> probably just a simple polynomial. The results you see are
probably from
> the color values being clipped and otherwise messed up by the
function.
>
>
> > I think, the question is: When is it applied? Before or after
AA??
>
> Gamma correction would have to be applied to the final result
of
> computing a pixel, correcting each sample wouldn't give the
right
> result. Antialiasing has little effect on the image, it mostly
makes a
> small difference in edges, so the change would be minor.

So this is what I originaly ment ... If you use the
gamma-correction(with negativ values) on an antialiased image ,
you don't have any longer the effect of antialiasing. The GC
could produce again (unexpected) stairsteps.

But: AFAIK Gamma is only defined for values > 0 ...

So does it make sense to ask if the current model is correct ???
Of course, we often have other functions in POV, working with
unrealistic values (negative pigments or lights ...[New question:
what is with negative IOR ???]), but they are or could be quite
predictable, but what do you expect from a negative gamma value
??? How would define the result ???

BTW: has someone here have the code for the gamma function ???

--

 ,',    Jan Walzer      \V/  http://wa.lzer.net     ,',
',','   student of      >|<  mailto:jan### [at] lzernet ',','
  '   ComputerScience   /A\  +49-177-7403863         '

Post a reply to this message

On Sat, 11 Nov 2000 23:39:00 -0800, Ken wrote:
>This is a multi-part message in MIME format.
>--------------B99E23FC96263D03210CBB0B
>Content-Type: text/plain; charset=us-ascii
>Content-Transfer-Encoding: 7bit
>
>
>Got a wild hair tonight and tried a negative value in the
>assumed_gamma global setting. Interesting effect. The first
>image is how it would normaly appear using assumed gamma 2.2.
>The second is assumed gamma -1.0. There must be an application
>for this but at the moment I can't think of one....

Just tried this on one of my own files, it's like all those 1970s 
pop videos have come back to get me. 

-- 
Cheers
Steve              email mailto:ste### [at] zeroppsuklinuxnet

%HAV-A-NICEDAY Error not enough coffee  0 pps. 

web http://www.zeropps.uklinux.net/

or  http://start.at/zero-pps

  6:40pm  up 32 days, 21:02,  2 users,  load average: 1.06, 1.10, 1.04

Post a reply to this message

In article <3a0efbd8@news.povray.org>, "Jan Walzer" <jan### [at] lzernet> 
wrote:

> So this is what I originaly ment ... If you use the
> gamma-correction(with negativ values) on an antialiased image ,
> you don't have any longer the effect of antialiasing. The GC
> could produce again (unexpected) stairsteps.

You would still have the effect of antialiasing, though there will be 
additional "jaggies" from the gamma function itself.


> Of course, we often have other functions in POV, working with
> unrealistic values (negative pigments or lights ...[New question:
> what is with negative IOR ???]), but they are or could be quite
> predictable, but what do you expect from a negative gamma value
> ??? How would define the result ???

Define it as whatever you get when you feed negative input to the gamma 
function. :-)

Negative ior may not be unrealistic. The ior of a bubble of air in water 
is lower than the ior of water, it just makes the light bend the other 
way. An ior of 1 usually means no effect, because POV sets the ior of 
"space" to 1. It is the change in ior that causes light to bend...


> BTW: has someone here have the code for the gamma function ???

I said it was probably simple. :-)
The variable opts.GammaFactor is set to assumed_gamma/DisplayGamma in 
parse.c.

render.c:
static void gamma_correct(COLOUR Colour)
{
  if (opts.Options & GAMMA_CORRECT)
  {
    Colour[RED]   = pow(Colour[RED],  opts.GammaFactor);
    Colour[GREEN] = pow(Colour[GREEN],opts.GammaFactor);
    Colour[BLUE]  = pow(Colour[BLUE], opts.GammaFactor);
  }
}

And as far as I can tell, Clip_Colour() is always called on the color 
before this, then the gamma function is called, then the result is added 
to the pixel before averaging of values. I'm not sure if the last part 
is correct, I noticed that focal blur was changed so individual samples 
weren't corrected.

-- 
Christopher James Huff
Personal: chr### [at] maccom, http://homepage.mac.com/chrishuff/
TAG: chr### [at] tagpovrayorg, http://tag.povray.org/

<><

Post a reply to this message

On Sun, 12 Nov 2000 21:14:20 +0100, Jan Walzer wrote:

>So this is what I originaly ment ... If you use the
>gamma-correction(with negativ values) on an antialiased image ,
>you don't have any longer the effect of antialiasing. The GC
>could produce again (unexpected) stairsteps.
>
>But: AFAIK Gamma is only defined for values > 0 ...
>
>So does it make sense to ask if the current model is correct ???
>Of course, we often have other functions in POV, working with
>unrealistic values (negative pigments or lights ...[New question:
>what is with negative IOR ???]), but they are or could be quite
>predictable, but what do you expect from a negative gamma value
>??? How would define the result ???
>
>BTW: has someone here have the code for the gamma function ???

Don't think I understand most of your questions but this should
answer some.  The attached image was gamma corrected in a graphics
package, to -1.6, and I get exactly the same results as using
negative gamma in pov.

Post a reply to this message

Attachments:
Download 'neggam.jpg' (29 KB)

Preview of image 'neggam.jpg'