POV-Ray : Newsgroups : povray.binaries.scene-files : Radiosity pretrace Server Time: 23 Jul 2019 11:13:29 GMT
  Radiosity pretrace (Message 1 to 6 of 6)  
From: Cousin Ricky
Subject: Radiosity pretrace
Date: 17 Jan 2014 00:00:01
Message: <web.52d87223da31b3c306548240@news.povray.org>
CLipka's Radiosity Voodoo recommends that pretrace_start be a power of 2
multiple of pretrace_end, and that pretrace_end translate to one pixel.
Calculating the pixel size is easy enough, but deriving a good pretrace_start
from that over various image sizes requires real math, so I incorporated a
formula into one of my include files so I wouldn't have to figure it out each
time.

Last night, I decided to extract this into a stand-alone macro.  Breathe easy;
it allows you to ease up on the pretrace_end recommendation.  I hope you find it
useful.


Post a reply to this message


Attachments:
Download 'radiosity-pretrace.zip' (4 KB)

From: Thomas de Groot
Subject: Re: Radiosity pretrace
Date: 17 Jan 2014 08:45:06
Message: <52d8ed92$1@news.povray.org>
On 17-1-2014 0:58, Cousin Ricky wrote:
> Last night, I decided to extract this into a stand-alone macro.  Breathe easy;
> it allows you to ease up on the pretrace_end recommendation.  I hope you find it
> useful.
>

Thanks a lot. IT certainly is useful. I am going to integrate it in my 
scenes.

Thomas


Post a reply to this message

From: Cousin Ricky
Subject: Re: Radiosity pretrace
Date: 17 Jan 2014 23:05:00
Message: <web.52d9b62fead9079f306548240@news.povray.org>
"Cousin Ricky" <rickysttATyahooDOTcom> wrote:
> CLipka's Radiosity Voodoo recommends that pretrace_start be a power of 2
> multiple of pretrace_end, and that pretrace_end translate to one pixel.
> Calculating the pixel size is easy enough, but deriving a good pretrace_start
> from that over various image sizes requires real math, so I incorporated a
> formula into one of my include files so I wouldn't have to figure it out each
> time.
>
> Last night, I decided to extract this into a stand-alone macro.  Breathe easy;
> it allows you to ease up on the pretrace_end recommendation.  I hope you find it
> useful.

There are errors in the documentation.  I pulled an _Of Pandas and People_[1]
and missed some sentences in my global search and replace.  (Funny thing is I
/remember/ proofreading it.  That's why writers never proofread their own work.)
 The corrected files are attached.

________________________
[1] See Kitzmiller v. Dover.


Post a reply to this message


Attachments:
Download 'radiosity-pretrace.zip' (4 KB)

From: Alain
Subject: Re: Radiosity pretrace
Date: 18 Jan 2014 00:48:01
Message: <52d9cf41$1@news.povray.org>
Le 14-01-16 18:58, Cousin Ricky a écrit :
> CLipka's Radiosity Voodoo recommends that pretrace_start be a power of 2
> multiple of pretrace_end, and that pretrace_end translate to one pixel.
> Calculating the pixel size is easy enough, but deriving a good pretrace_start
> from that over various image sizes requires real math, so I incorporated a
> formula into one of my include files so I wouldn't have to figure it out each
> time.
>
> Last night, I decided to extract this into a stand-alone macro.  Breathe easy;
> it allows you to ease up on the pretrace_end recommendation.  I hope you find it
> useful.
>

The formulas are prety simple:

pretrace_end 1/image_width // stop at 1 pixel block
pretrace_start 64/image_width // start with 64 pixels blocks

Those will have the pretrace start at 64 pixels and end at 1 pixel 
blocks for any resolution for a total of 7 pretrace steps.

image_hight is unimportant and must be ignored. The size of the pretrace 
tiles is strictly a fraction of the horizontal image of your image.


Alain


Post a reply to this message

From: Cousin Ricky
Subject: Re: Radiosity pretrace
Date: 18 Jan 2014 02:40:00
Message: <web.52d9e906ead9079f306548240@news.povray.org>
Alain <kua### [at] videotronca> wrote:
> The formulas are prety simple:
>
> pretrace_end 1/image_width // stop at 1 pixel block
> pretrace_start 64/image_width // start with 64 pixels blocks

Why 64?  Why not 32 or 128?  My formula takes the size of the image into
account.

> image_hight is unimportant and must be ignored. The size of the pretrace
> tiles is strictly a fraction of the horizontal image of your image.

No, it goes by whichever is the larger dimension.  Try it.


Post a reply to this message

From: clipka
Subject: Re: Radiosity pretrace
Date: 18 Jan 2014 03:20:07
Message: <52d9f2e7@news.povray.org>
Am 18.01.2014 00:01, schrieb Cousin Ricky:

> There are errors in the documentation.  I pulled an _Of Pandas and People_[1]
> and missed some sentences in my global search and replace.
> ________________________
> [1] See Kitzmiller v. Dover.

Duh, that was a difficult one... had to watch the Nova documentary to 
get it. (Highly recommended viewing, BTW!)


Post a reply to this message

Copyright 2003-2008 Persistence of Vision Raytracer Pty. Ltd.