Le 2020-04-05 à 12:12, William F Pokorny a écrit :
> Previous postings:
> 
>
http://news.povray.org/povray.beta-test.binaries/thread/%3C5da21410%40news.povray.org%3E/

> 
> 
>
http://news.povray.org/povray.beta-test.binaries/thread/%3C5dab3f55%241%40news.povray.org%3E/

> 
> 
>
http://news.povray.org/povray.beta-test.binaries/thread/%3C5dbc6985%241%40news.povray.org%3E/

> 
> 
> and
> 
>
http://news.povray.org/povray.beta-test.binaries/thread/%3C5dd8006c%241%40news.povray.org%3E/

> 
> 
> As discussed previously, I've added a new wave modifier mode called 
> function_interval which works in the -1 to 1 range instead of the usual 
> pattern range of 0 to 1. As part of this working to update all the 
> existing *_wave modifiers to work in both ranges.
> 
> Found with the cubic_wave (does anyone really use it today?) when, eons 
> ago, the inversion of negative values was introduced to get continuity 
> (non-flipping at) while moving from negative to positive values through 
> 0.0 the cubic pattern was apparently not updated and the continuity was 
> mostly broken for it as can be seen in the upper left of the attached 
> image applying it to a gradient x pattern.
> 
> I worked to fix this, but then could see no advantage for cubic_wave 
> over what we already have with sine_wave. I was going to delete it, but 
> have decided for povr, to change the existing cubic wave modifier for 
> both ranges into a bump (selection) function of sorts. The result of 
> this can be seen in the upper right of the attached image again applying 
> it against a gradient x pattern.
> 
> Today if we try a pattern with cubic_wave like:
> 
> granite cubic_wave frequency 3.3 phase -0.5
> 
> and use it as an offset to the y plane in an isosurface, we get the 
> lower right results in the attached image. It's something, but it has 
> awful continuity and is really useful only for noisy results of one kind 
> or another.
> 
> In the lower right is the bump function result with runs in nearly 1/3 
> the time because it maintains the incoming continuity while providing 
> for results much more likely to be useful.
> 
> Bill P.
> 
> 
This looks like an improvement.
As I very rarely encounter that wave type, i don't think that it will 
break anything.

Post a reply to this message