Norbert Kern <nor### [at] t-onlinede> wrote:
> I see application of the method as a special form of image resizing, which
> is certainly allowed.

  I think that's really bending the rules.

  The purpose of the contest is to see what a renderer can do. In order to
achieve this purpose, only the image output by the renderer is allowed.
  Adjusting brightness/contrast is in the limits of this, although still
allowed.
  However, improving the antialiasing of an image by post-processing it
with a paint program clearly breaks this purpose and the rules. The resulting
image is not what the renderer generated, but it has been improved with
a paint program.

  There's a simple way of seeing the difference between adjusting
brightness/contrast and smoothing an image by scaling it smaller (ie.
improving its antialiasing): If the image has some pixel-sized artifacts
(which usually appear due to bad antialiasing), brightness/contrast
adjustments will not do anything about it, but they will still be there.
However, scaling the image smaller can get rid of these artifacts. Thus
what it is doing here is to remove artifacts produced by the renderer with
a paint program. The resulting image is not the one produced by the
renderer, but an improved one.
  If that's not illegal post-processing, then what is?
  How is this different from, for example, applying a blur filter with the
paint program to remove pixelation? Why blurring should be any more illegal?

-- 
#macro N(D)#if(D>99)cylinder{M()#local D=div(D,104);M().5,2pigment{rgb M()}}
N(D)#end#end#macro M()<mod(D,13)-6mod(div(D,13)8)-3,10>#end blob{
N(11117333955)N(4254934330)N(3900569407)N(7382340)N(3358)N(970)}//  - Warp -

Post a reply to this message

Norbert Kern <nor### [at] t-onlinede> wrote:
> resizing is allowed :

> http://www.irtc.org/stills/faq.html#q1.1.14

> [1.1.14] Is it legal to resize my image after it is rendered?
> Yes, it is. Be careful doing it, though--unless done correctly, resizing can
> often add unwanted visual artifacts to an image.

  See my other post.

-- 
#macro N(D)#if(D>99)cylinder{M()#local D=div(D,104);M().5,2pigment{rgb M()}}
N(D)#end#end#macro M()<mod(D,13)-6mod(div(D,13)8)-3,10>#end blob{
N(11117333955)N(4254934330)N(3900569407)N(7382340)N(3358)N(970)}//  - Warp -

Post a reply to this message

Warp wrote:
> 
> Norbert Kern <nor### [at] t-onlinede> wrote:
> > I see application of the method as a special form of image resizing, which
> > is certainly allowed.
> 
>   I think that's really bending the rules.

Resizing is allowed.


_____________
Kari Kivisalo

Post a reply to this message

Kari Kivisalo <pro### [at] luxlabcom> wrote:
>>   I think that's really bending the rules.

> Resizing is allowed.

  My statement still holds.

  I don't think that the people who decided about that rule thought that
it could be used to enhance the quality of the image.
  It's the spirit of the IRTC rules that the quality of the image produced
by the renderer should not be enhanced by post-processing. The image should
be the one produced by the renderer, with no modified quality. If the
image produced by the renderer had artifacts, it's against the rules to
remove these artifacts with a paint program. Using the resizing trick
is essentially doing this.
  We should think about the rules in their whole context and understand
the spirit of the rules, not just their letter.

-- 
#macro N(D)#if(D>99)cylinder{M()#local D=div(D,104);M().5,2pigment{rgb M()}}
N(D)#end#end#macro M()<mod(D,13)-6mod(div(D,13)8)-3,10>#end blob{
N(11117333955)N(4254934330)N(3900569407)N(7382340)N(3358)N(970)}//  - Warp -

Post a reply to this message

On Sat, 30 Mar 2002 17:56:47 -0500, "Chris Becker" <cmb### [at] ritedu>
wrote:

>I have a very thin object that is not showing up correctly because some
>pixels pick it up and others don't hence creating jaggies, even with
>anti-aliasing. My question is, what do I adjust in anti-aliasing to improve
>this? And does increasing or decreasing the threshold provide for a better
>sample?
>
>Thanks in advance!
>
>

I had the same problem just a fw days ago: a rendering with very small
text objects.  The anti-aliasing was not really working because
parts of the letters were falling between adjacent pixels, so that
the anti-aliasing was not triggered.  I ended up rendering the image
at double size, with anti-aliasing ... and sizing it down in a paint
package.  That fixed it.

Post a reply to this message

So what's the big deal of actually modifying POV-Ray with a new AA method?
AA shoots multiple rays for a single pixel which is then averaged. How is
that different than just taking a higher rendering of the image and
averaging the pixels? Why is it such a big deal to modify the renderer?
Megapov does post processing and I don't see such huge debates about that...

"Warp" <war### [at] tagpovrayorg> wrote in message
news:3ca8ed48@news.povray.org...
> Kari Kivisalo <pro### [at] luxlabcom> wrote:
> >>   I think that's really bending the rules.
>
> > Resizing is allowed.
>
>   My statement still holds.
>
>   I don't think that the people who decided about that rule thought that
> it could be used to enhance the quality of the image.
>   It's the spirit of the IRTC rules that the quality of the image produced
> by the renderer should not be enhanced by post-processing. The image
should
> be the one produced by the renderer, with no modified quality. If the
> image produced by the renderer had artifacts, it's against the rules to
> remove these artifacts with a paint program. Using the resizing trick
> is essentially doing this.
>   We should think about the rules in their whole context and understand
> the spirit of the rules, not just their letter.
>
> --
> #macro N(D)#if(D>99)cylinder{M()#local D=div(D,104);M().5,2pigment{rgb
M()}}
> N(D)#end#end#macro M()<mod(D,13)-6mod(div(D,13)8)-3,10>#end blob{
> N(11117333955)N(4254934330)N(3900569407)N(7382340)N(3358)N(970)}//  -
Warp -

Post a reply to this message

On 1 Apr 2002 15:08:40 -0500, Warp <war### [at] tagpovrayorg> wrote:
> > It is much better than my method, which I used in my two last IRTC entries.
> > I will no more use normal POV-Ray antialiasing, as such methods show enhance
> > most pictures considerably.
>
>  I don't think you can do that with an IRTC entry. The rules prohibit
> post-processing.

What about writing the same as function in 3.5 ?

ABX

Post a reply to this message

Chris Becker <cmb### [at] ritedu> wrote:
> So what's the big deal of actually modifying POV-Ray with a new AA method?

  Nothing.

> AA shoots multiple rays for a single pixel which is then averaged. How is
> that different than just taking a higher rendering of the image and
> averaging the pixels?

  It's different because the latter method is post-processing. It's not
something the renderer itself supports, but has to be done with a paint
program.

> Why is it such a big deal to modify the renderer?

  It isn't.

> Megapov does post processing and I don't see such huge debates about that...

  Because it's the renderer itself which does it.
  The idea is to participate with the output image of a renderer, not a
rendered image which has been enhanced with another program afterwards.
The IRTC rules prohibit post-processing for this reason.

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

> What about writing the same as function in 3.5 ?

  How?

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

On 2 Apr 2002 08:15:25 -0500, Warp <war### [at] tagpovrayorg> wrote:
> W?odzimierz ABX Skiba <abx### [at] babilonorg> wrote:
> > What about writing the same as function in 3.5 ?
>
> How?

Knowing algorithm for particular effect in 2d manipulation software writing
macro/pattern applied to plane with ambient 1 and diffuse 0 under orthogonal
camera. Then it use all elements of typical still creation process but has the
same effect. OF course it is cheating but probably works.

ABX

Post a reply to this message