POV-Ray : Newsgroups : povray.binaries.images : Playing with bump_map. Doc seam bug. Server Time
5 Feb 2025 15:13:03 EST (-0500)
  Playing with bump_map. Doc seam bug. (Message 1 to 1 of 1)  
From: William F Pokorny
Subject: Playing with bump_map. Doc seam bug.
Date: 3 Feb 2025 18:44:01
Message: <67a154c1$1@news.povray.org>
I recently worked up a bump_map{} scene for the yuqk fork documentation.

Attached is an image with positive and negative bump_size crackle 
pattern in the top row. Lower left is wrinkles. Lower right leopard.

All of the pattern implementations used are somewhat different in the 
yuqk fork, but I've attached a scene file which will work with v3.8 beta 
2 should folks want to play (it 'probably' works back to v3.7).

Of note there is some sort of top / right bug in the way the bump_map{} 
is being handled for use in the normal{} block which has existed for a 
long time. In the attached 'yuqk' fork generated image the seam has been 
rotated into full view. I've not attempted to debug it. There is a note 
in the code about samples falling outside the mapped area needing to be 
handled correctly.

The bump_size as applied to bump_map{}s is a bit funky in that for many 
image types coming in with values in the 0 to 1 range, very large 
bump_size multipliers are needed.

Internally there is a value/255 kind of thing going on. Guess is some 
formats / options (idexed?) come in as integers where the divisions by 
255 makes sense. The bump_size of +-127 equates to an effective 
bump_size of +-0.5 internally where values in the 0-1 range.

Due the sampling, the normals at changes in slope direction are 
sometimes noisy.

Bill P.


Post a reply to this message


Attachments:
Download 'bump_mapstory.jpg' (151 KB) Download 'bump_map_v38.pov.txt' (2 KB)

Preview of image 'bump_mapstory.jpg'
bump_mapstory.jpg

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