|
 |
I spent some time trying to unravel the calculations in the macro, and indeed
with some time we can probably figure it all out.
After working from the inner parenthetical expressions outwards, it seems to me
that _overall_, most, if not all of the calculations are a method of going from
0 to 1, or 1 to 0 - but in a highly non-linear manner.
{HOW he ever originally conceived to do all of that, and keep it straight, is
something that's a bit mind-bending. I can only guess that must be some form of
probability distribution or some form of specialty nonlinear progression that
might be familiar to mathematicians...}
He then multiplies the result by length or width, or some other factor to cover
the range.
Most notable was the string of sequential _c declares whose whole purpose is to
define the rgb value for that part of the feather.
I'll continue to poke at this - and maybe graph out some of the individual
functions in OpenOffice Calc. Or POV-Ray :D
Post a reply to this message
|
|
|
|
"Bald Eagle" <cre### [at] netscape net> wrote:
> I spent some time trying to unravel the calculations in the macro, and indeed
> with some time we can probably figure it all out.
>
> After working from the inner parenthetical expressions outwards, it seems to me
> that _overall_, most, if not all of the calculations are a method of going from
> 0 to 1, or 1 to 0 - but in a highly non-linear manner.
>
> {HOW he ever originally conceived to do all of that, and keep it straight, is
> something that's a bit mind-bending. I can only guess that must be some form of
> probability distribution or some form of specialty nonlinear progression that
> might be familiar to mathematicians...}
>
> He then multiplies the result by length or width, or some other factor to cover
> the range.
>
> Most notable was the string of sequential _c declares whose whole purpose is to
> define the rgb value for that part of the feather.
>
> I'll continue to poke at this - and maybe graph out some of the individual
> functions in OpenOffice Calc. Or POV-Ray :D
All works...
I wait for your results.
For my part I use an array and a sphere_sweep in _x loop, I have removed all
spheres and cylinder (except for the rachis). The result is good, the barbs
looks better, but the render time was increased (while the parsing time semms to
be the same..).
ACtually I use the same array for the two sides, so the feather is very
symmetric.
Post a reply to this message
|
|
|
|
"Fractracer" <lg.### [at] gmailcom> wrote:
> All works...
> I wait for your results.
Still working through it, though I noticed there are a few calculations that
ought to be throwing a divide-by-zero error.
like:
_dz*(degrees(asin(_y/_length*_depth/_length))/90.)/(degrees(asin(_depth/_length))/90.)*sin(radians(pow(_y/_length,5)*90
.))>,
since _y goes from 0 to _length.
So the first calculation must be (something) / ((something)*(0))
Not sure how POV-Ray works around that.
Post a reply to this message
|
|
