 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
George Pantazopoulos wrote:
>
> It would be a great feature for creators of realistic, high-dynamic range
> scenes ( with 1/r^2 lighting falloff ), to have POV-Ray be able to
> automatically compress the dynamic range of the output image, as seen on
> this web page: http://www.luxlab.com/c/compress.htm It would be especially
> handy for animations.
Automating this is not trivial but logarithmic or some other transfer
curve with adjustable compression factor may be usefull. The curve
presented on that page doesn't fit all cases. It's just an example to
break the 100% mental limit. For low contrast scenes there is dynamic
range expansion. Just flip the curve around the diagonal. I see these
as special cases of tonal range mapping.
Tweaking the curve by hand yields the best results. It's like working
in the dark room controlling the development, exposure and whatever
else photographers do there :)
_____________
Kari Kivisalo
Post a reply to this message
|
|
| |
| |
|
|
From: George Pantazopoulos
Subject: Re: 3.5 feature request - dynamic range compression
Date: 16 Apr 2002 01:14:24
Message: <3cbbb330$1@news.povray.org>
|
|
|
| |
| |
|
|
Thanks for the input Kari!
Does the adc_bailout (in either radiosity or global settings) have to be
changed to a smaller value, in order to allow for true 48-bit resolution
output?
George
"Kari Kivisalo" <pro### [at] luxlab com> wrote in message
news:3CB6EBCE.1A83DA4D@luxlab.com...
> George Pantazopoulos wrote:
> >
> > It would be a great feature for creators of realistic, high-dynamic
range
> > scenes ( with 1/r^2 lighting falloff ), to have POV-Ray be able to
> > automatically compress the dynamic range of the output image, as seen on
> > this web page: http://www.luxlab.com/c/compress.htm It would be
especially
> > handy for animations.
>
> Automating this is not trivial but logarithmic or some other transfer
> curve with adjustable compression factor may be usefull. The curve
> presented on that page doesn't fit all cases. It's just an example to
> break the 100% mental limit. For low contrast scenes there is dynamic
> range expansion. Just flip the curve around the diagonal. I see these
> as special cases of tonal range mapping.
>
> Tweaking the curve by hand yields the best results. It's like working
> in the dark room controlling the development, exposure and whatever
> else photographers do there :)
>
>
> _____________
> Kari Kivisalo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
George Pantazopoulos wrote:
>
> Does the adc_bailout (in either radiosity or global settings) have to be
> changed to a smaller value, in order to allow for true 48-bit resolution
> output?
I have had trouble with global adc and reflections of bright objects
even with normal images. These things were designed for the 0-1 range
but lowering the limit is the obvious choice.
I haven't tested adc_bailout with radiosity. It's yet another uncertain
element in the equation and I'm trying to avoid those.
_____________
Kari Kivisalo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Kari Kivisalo wrote:
>
> I haven't tested adc_bailout with radiosity. It's yet another uncertain
> element in the equation and I'm trying to avoid those.
My tests with MegaPOV back when everyone was trying to copy Arnold
seemed to indicate that increasing adc_bailout had almost no effect on
radiosity, with a slight improvement in quality at higher settings
without affecting rendering speed (and now I use adc_bailout 1 in all my
scenes as a result). This seems counterintuitive to me and I don't know
if it holds true for 3.5 or not.
-Xplo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Xplo Eristotle wrote:
>
> (and now I use adc_bailout 1 in all my
> scenes as a result). This seems counterintuitive to me and I don't know
> if it holds true for 3.5 or not.
There have been some changes so maybe it's time to test again.
Report here if you find out how it works :)
_____________
Kari Kivisalo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Kari Kivisalo wrote:
>
> Xplo Eristotle wrote:
> >
> > (and now I use adc_bailout 1 in all my
> > scenes as a result). This seems counterintuitive to me and I don't know
> > if it holds true for 3.5 or not.
>
> There have been some changes so maybe it's time to test again.
> Report here if you find out how it works :)
I don't like the official interface, and I also slightly prefer the
syntax and feature set of MegaPOV. You're welcome to test it yourself,
though. :)
-Xplo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
