Le 2024-11-22 à 16:07, William F Pokorny a écrit :
> On 11/22/24 08:40, Bald Eagle wrote:
>> William F Pokorny <ano### [at] anonymousorg> wrote:
>>> Some of the tiling pattern's patterns look pretty cool without a color
>>> map - just grey. Attached tiling 21.
>>>
>>> Bill P.
>>
>> Nice!
>> Are these the sort of patterns that are provided with POV-Ray - coded 
>> by Jerome
>> Grimbert?
> 
> Thanks :-) Though, yes, this is Jerome's work so he deserves the credit.
> 
> Aside: In addition to the 'tiling' pattern (implementing 27 sub-tiling- 
> patterns), There is too 'pavement' with a great many possibilities (112 
> base arrangements IIRC - plus exterior, interior and form treatments 
> thereof). I think it a bit harder to understand and use than is tiling, 
> but it's cool too.
> 
>>
>> I'm curious as to what this would look like as a heightfield.
>>
> 
> Attached another grey image on the left that I liked in 'tiling 14'. It 
> required the tiny rotation trick to get away from numerical noise.
> 
> Many of the tiling pattern's patterns are not continuous which make 
> height fields and isosurfaces difficult / jumpy in direct use. Kenneth 
> helps make that point with his recent post. :-)
> 
> On the right isn't a height field, but an isosurface using a new inbuilt 
> function in yuqk called f_npmod() which in wrapping the 'tiling 14' 
> function implements the pattern modifier (frequency, phase, sine_wave 
> etc) code's version of mod() - plus an on the fly scaling. 'tiling 14' 
> has 5 tile types so the f_npmod() function call becomes roughly:
> 
>      f_npmod('tiling 14', 1/5, 'value multiplier')
> 
> The function basically makes 'tiling 14' a continuous pattern where all 
> 5 tile types have the same 0..? value ranges. An unmodified 'tiling 14' 
> function is still used to color / texture the tiles and calk.
> 
> Note. The performance is quite variable where using the 'tiling' pattern 
> in isosurfaces. My up front guess is that certain tiling sub- 
> patterns(1..27) are much slower than others, but there might well be 
> other things in play. I've never dug into what burns the time. The 
> isosurface on the right was relatively slow to render.
> 
> Bill P.
> 

When I use a tiling pattern for an isosurface, I use a simple ramp 
gradient with a frequency parameter matching the number of domains.
So, for your example, I'd use frequency 5.
Then, use a colourful version of that tiling as pigment.

Post a reply to this message