POV-Ray : Newsgroups : povray.binaries.images : High Quality Prints ;-) : Re: High Quality Prints ;-) Server Time
1 Oct 2024 22:27:58 EDT (-0400)
  Re: High Quality Prints ;-)  
From: Chris Huff
Date: 14 Aug 2000 20:46:18
Message: <chrishuff-C00F72.19472914082000@news.povray.org>
In article <398fe057@news.povray.org>, "Rune" <run### [at] inamecom> 
wrote:

> Ah, but then it would not be the final output image itself that has 
> the effect, only an object inside it. You would still have to do more 
> than one render to get the desired result. Then the question remains 
> - which advantages would a PP effect have over the method I used?

Speed, maybe memory, with later versions the possibility to have 
multiple filtered versions output at the same time...


> Would you add different preprogrammed types of dithering, or would 
> you let the user have complete control over how the dithering should 
> be calculated? The method I use gives me complete control.

Different preprogrammed types, along with user control. Maybe even using 
pigments to control portions, like you use.


> My method is rather fast if it weren't for the AA. It looks *very* bad
> without AA. As far as I know PP and AA doesn't work nicely together...

They work fine, as I recall, the problem was fixed.


> Currently, the PP patch writer has access to some input (image, 
> depth-map, normal-map, and maybe others), and he can then generate 
> the output based on these exactly as he wants.

Color(the image), depth, raw normal, perturbed normal, and intersection 
point. Hopefully ray direction will be added, since that could be used 
in some very interesting filters(and could add a fourth edge-finding 
method to the find_edges filter).


> What I really would like to see is a patch that enables the *user* to 
> do just that. Sort of an "ultimate PP patch". If that were made, many 
> of the current PP effects could be made using #macros and such 
> instead. The user would have *much* greater control, and the whole 
> idea of PP in POV-Ray would work much better. I don't know if this is 
> possible, but if it is, then I certainly hope it will be made.

I have thought about this, only using isosurface type functions instead 
of macros(I still haven't figured out how to call a macro with certain 
parameters from within a patch). I just have to figure out how to add 
accessor functions for rgb color, surface normal, etc.


> Sure, it would be a little slower, but that's a price I would gladly 
> pay in order to get greater control as a user.

Actually, with functions, it might be pretty fast. Certainly faster than 
macros...
And using functions gives the user a way to access pixels other than the 
current one, which could be a very powerful feature. Some things(loops, 
conditionals) are missing, but they might be added to functions in the 
future.

-- 
Christopher James Huff - Personal e-mail: chr### [at] maccom
TAG(Technical Assistance Group) e-mail: chr### [at] tagpovrayorg
Personal Web page: http://homepage.mac.com/chrishuff/
TAG Web page: http://tag.povray.org/


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.