Norbert Kern wrote:
> 
>>I think anyway that this is posprocessing because it is applied on a 2d
>>image and not in a 3d scene
> 
> 
> What, if you make an animation with two images - first the actual render,
> then afterwards the "postprocessing" using the first frame?

The problem is quite simple. What you do this way is circumventing the 
rules of the IRTC.  This is fine IMO but it should lead to the rules 
being modified to be more precise and allow less 'tricks' to circumvent 
them.  Since the IRTC rules have not been changed at all recently post 
processing with external tools is now in fact allowed.  This is not bad 
per se but the rule is a joke then and should be removed.

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 01 May. 2004 _____./\/^>_*_<^\/\.______

Post a reply to this message

news:cbh7i2$50e$1@chho.imagico.de...

> The problem is quite simple. What you do this way is circumventing the
> rules of the IRTC.  This is fine IMO but it should lead to the rules
> being modified to be more precise and allow less 'tricks' to circumvent
> them.  Since the IRTC rules have not been changed at all recently post
> processing with external tools is now in fact allowed.  This is not bad
> per se but the rule is a joke then and should be removed.

I think that the spirit of the original rule is to prevent photoshopping,
because otherwise it would be impossible to prevent obviously unacceptable
(*) modifications of a rendered image, like painting new elements by hand.

Perhaps the rule could just be altered by stating that the image generation
(rendering + ant not-rendered effect) should be done with the same program.
This would prevent the use of paint programs while officially authorising
post-processing when done with built-in features of the renderer, including
SDL-based post-processing.

The phrase:
"Images must not be enhanced or altered ('post-processed') by use of paint
programs such as PhotoShop(tm) etc."

could then be replaced by:

"Images must not be modified by use of a program other than the original
renderer".


G.

(*) in the context of a rendering competition of course


-- 

**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters

Post a reply to this message

Gilles Tran wrote:
> 
> I think that the spirit of the original rule is to prevent photoshopping,
> because otherwise it would be impossible to prevent obviously unacceptable
> (*) modifications of a rendered image, like painting new elements by hand.
> 
> Perhaps the rule could just be altered by stating that the image generation
> (rendering + ant not-rendered effect) should be done with the same program.
> This would prevent the use of paint programs while officially authorising
> post-processing when done with built-in features of the renderer, including
> SDL-based post-processing.

This would surely be much clearer than the current rule.  The difference 
between post processing the image in photoshop and doing the same post 
processing in an awfully slow and memory intensive SDL based system is 
quite academic though.  But unless you want to completely remove the 
post processing rule this probably is the clearest line that can be drawn.

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 01 May. 2004 _____./\/^>_*_<^\/\.______

Post a reply to this message

Samuel Benge wrote:

> Hello everyone. Today I made a scene file which
 > takes an image and applies a 'bloom' effect to
 > the bright parts.

Did you do this by rendering a grayscale map with something like [.5 rgb 
0][.5 rgb 1] blurring the map, and then combining the blur, the original 
map, and the original image? (hope that made sense)

...or, is this a one-step process? For the record, I don't see anything 
procedural/repeatable as unacceptable post-processing.

  -Shay

Post a reply to this message

Samuel Benge wrote:
> Hello everyone. Today I made a scene file which takes an image and 
> applies a 'bloom' effect to the bright parts. The example attached is a 
> bit overdone to show the results. I can control which values in the 
> image are active.
> 
> I think that this technique is probably okay to use in the IRTC, but I 
> wonder if it would be considered a post-processing step or not....
> 
> Happy raytracing~
> 
> -Sam Benge
> 
> ------------------------------------------------------------------------
> 
Congratulations, clever technically and pleasing visually.  Is it my 
imagination or is diffused atmosphere in photos becoming so popular 
lately as to practically define current visual style.  Like in ads, TV, 
movies, photo-art, etc., etc.

Post a reply to this message

Jim Charter wrote:

> Congratulations, clever technically and pleasing visually.  Is
> it my imagination or is diffused atmosphere in photos becoming
> so popular lately as to practically define current visual style.
> Like in ads, TV, movies, photo-art, etc., etc.

The Bruckheimer effect.

 -Shay

Post a reply to this message

Jeremy M. Praay wrote:
<snip>

> 
> Incredible work!  :-)
> 


Thanks!

>>I think that this technique is probably okay to use in the IRTC, but I
>>wonder if it would be considered a post-processing step or not....
> 
> Only if it can be done in one step.  It sounds like it's a 2 part process
> (render scene then take image and add bloom effect), which could disqualify
> it, since it's done through the renderer.


Since POV-Ray has so many options (as a programming language), it might 
be possible to link the post-processing file to execute right after a 
render is complete. I'll have to look into that....

 
> Now having said all of that, I'll say that I actually hate IRTC
> post-processing rule discussions...  bleh
 
Me too.


-Sam

Post a reply to this message

Shay wrote:

> Did you do this by rendering a grayscale map with something like [.5 rgb 
> 0][.5 rgb 1] blurring the map, and then combining the blur, the original 
> map, and the original image?

 >

>(hope that made sense)


It does, and it's almost completely correct. There are two planes. One 
has the original image, and it remains unharmed. The second plane is 
above the first one. It contains the blurred highlights using parts from 
the original image. All darker portions are transparent. The upper and 
lower values are controlled in a pigment_map, and allow for a smooth 
transition from transparent color to image-based color. (if that made 
any sense :)

-Sam

Post a reply to this message

Jim Charter wrote:

>>
> Congratulations, clever technically and pleasing visually. 


Thanks

> Is it my 
> imagination or is diffused atmosphere in photos becoming so popular 
> lately as to practically define current visual style.  Like in ads, TV, 
> movies, photo-art, etc., etc.

It adds an extra dimension of realism to an image at low overhead. Cheap 
to render; worth the extra time for some renders. I'll end up using it 
in too many of my renders for a while, eventually learning when *not* to 
use it :)

-Sam

Post a reply to this message

> Hello everyone. Today I made a scene file which takes an image and
> applies a 'bloom' effect to the bright parts. The example attached is a
> bit overdone to show the results. I can control which values in the
> image are active.

This reminds me of my Glare-Include I've written a while ago. What I did was
to render the scene with Black-Textures and Specular-Highlights only, then
calculate the position for the Glare-Image in-scene. Using average-pattern,
I blurred the image a little and thus got me the glaring effect.
Do you use the colored output of your image as input for the blooming
effect?

Regards,
Tim

-- 
"Tim Nikias v2.0"
Homepage: <http://www.nolights.de>

Post a reply to this message