I've dug into the normal{} block facets perturbation code a little. I'll 
note a few things and then show the initial set of related keyword 
changes I hope makes facets simpler to use in yuqk.

These notes which apply to official releases and the yuqk fork:

1) Spatial modifiers (ie scale 0.5) as used within the normal{} block 
affect only the flat form.

2) The 'coords' keyword is effectively 'bump_size', thought it only 
affects the flat form.

3) In news to me, there exist already a fully functioning 'metric' 
keyword for both forms. I'm not aware of this being documented anywhere?

4) It was noted somewhere on these news groups that facets replaces the 
normal rather than perturbs it, but this is only true for the 
spherically / curved surface oriented version of facets.

5) The internal code used a crackle implementation but it is an older 
and (by comment at least) slower version than the current crackle 
pattern code implementation.

6) The spherical form is best applied around the origin then transformed 
for use.

7) Both patterns are sensitive to extreme scaling around the default of 
0.1. The spherical form more so than the flat one.

-----------

The attached image reflects the result of the top set of normal{} block 
facets specifications for recent official releases of POV-Ray as well as 
the bottom set of four specifications reflecting the keyword changes 
being adopted for release R15 of yuqk. The core facet implementation has 
not yet been changed.

The aim was to quickly try out the already implemented 'metric' keyword 
code and the keyword changes.

  //--- For official releases
  //facets coords 0.8 size 0.1 metric  0+(frame_number-1)*1.0
  //facets coords 0.8 size 0.1 metric -2+(frame_number-1)*0.5
  //facets coords 0.0 size 0.1 metric  0+(frame_number-1)*1.0
  //facets coords 0.0 size 0.1 metric -2+(frame_number-1)*0.5

  //--- The version with yuqk's keyword changes
    facets type 0 ip_size 0.1
      ip_metric  0+(frame_number-1)*1.0 bump_size 0.8
  //facets type 0 ip_size 0.1
  //  ip_metric -2+(frame_number-1)*0.5 bump_size 0.8
  //facets type 1 ip_size 0.1
  //  ip_metric  0+(frame_number-1)*1.0
  //facets type 1 ip_size 0.1
  //  ip_metric -2+(frame_number-1)*0.5

Bill P.

Post a reply to this message

Attachments:
Download 'facetsstory_yuqkr15.jpg' (218 KB)

Preview of image 'facetsstory_yuqkr15.jpg'

hi,

William F Pokorny <ano### [at] anonymousorg> wrote:
> ...
> 3) In news to me, there exist already a fully functioning 'metric'
> keyword for both forms. I'm not aware of this being documented anywhere?
> ...
> 5) The internal code used a crackle implementation but it is an older
> and (by comment at least) slower version than the current crackle
> pattern code implementation.

guessing not what you meant, but in the wiki reference page 'metric' links to:
<https://wiki.povray.org/content/Reference:Crackle_Pattern>


regards, jr.

Post a reply to this message

On 6/5/24 09:04, jr wrote:
> 5) The internal code used a crackle implementation but it is an older

First, the above bit should have read:

5) The internal code uses a crackle implementation, but it is an older...

And correct, I was referring to the 'metric' keyword not being mentioned 
as a 'facets' pattern / perturbation option when it is supported:

https://wiki.povray.org/content/Reference:Facets_Pattern

At the moment I'm unsure if I'll change any of the existing 'facets' 
core code(a). However, I did want to set up to take a run at adding 
additional types along the lines of the facets pattern I posted about 
last fall to the existing 'facets' perturbation wrapper. See:

https://news.povray.org/povray.pov4.discussion.general/thread/%3C652b157c%40news.povray.org%3E/

Bill P.

(a) - I can somewhat see how the type 0 / flat form of 'facets' could be 
converted to our current 'crackle' implementation (facets is still 
missing the 'form' capability which could be cool). I'm a lot less clear 
on how I might adapt (or replace) the type 2 / curved surface approach. 
In any case, the up front plan is to get one or more of my facet ideas 
running as types 3,4,... so I can play. :-)

Post a reply to this message

On 6/4/24 18:34, William F Pokorny wrote:
>   //--- The version with yuqk's keyword changes
>     facets type 0 ip_size 0.1
>       ip_metric  0+(frame_number-1)*1.0 bump_size 0.8

A quick note. I moved up a planned 'type' to 'ip_type' change to release 
(R15) of the yuqk fork.

This affects yuqk's current 'facets' perturbation syntax.

Bill P.

Post a reply to this message