nerdcruft/libcruft-util
1
0
Fork 0
Code Issues Releases Activity

noise-tool: allow single octave output

Browse Source
This commit is contained in:
Danny Robson 2015-06-15 17:48:49 +10:00
parent c885e753d2
commit d9d7cd6f5b
1 changed files with 24 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
tools/noise.cpp
View File

@ -180,6 +180,7 @@ main (int argc, char **argv)
float H; float H;
float scale = 1.f; float scale = 1.f;
float turbulence = 0.f; float turbulence = 0.f;
unsigned single = 0;
// fill variables from arguments // fill variables from arguments
util::cmdopt::parser args; util::cmdopt::parser args;
@ -190,6 +191,7 @@ main (int argc, char **argv)
args.add<util::cmdopt::option::value<fractal_t>> ('f', "fractal", "primary fractal function", fractal); args.add<util::cmdopt::option::value<fractal_t>> ('f', "fractal", "primary fractal function", fractal);
args.add<util::cmdopt::option::value<lerp_t>> ('l', "lerp", "interpolation algorithm", lerp); args.add<util::cmdopt::option::value<lerp_t>> ('l', "lerp", "interpolation algorithm", lerp);
args.add<util::cmdopt::option::value<unsigned>> ('o', "octaves", "total fractal iterations", octaves); args.add<util::cmdopt::option::value<unsigned>> ('o', "octaves", "total fractal iterations", octaves);
args.add<util::cmdopt::option::count<unsigned>> ('1', "single", "single octave", single);
args.add<util::cmdopt::option::value<float>> ('H', "hurst", "Hurst exponent", H); args.add<util::cmdopt::option::value<float>> ('H', "hurst", "Hurst exponent", H);
args.add<util::cmdopt::option::value<float>> ('x', "scale", "frequency multiplier", scale); args.add<util::cmdopt::option::value<float>> ('x', "scale", "frequency multiplier", scale);
args.add<util::cmdopt::option::value<float>> ('t', "turbulence","turbulence scale", turbulence); args.add<util::cmdopt::option::value<float>> ('t', "turbulence","turbulence scale", turbulence);
@ -276,13 +278,32 @@ main (int argc, char **argv)
util::image::buffer<float> img (res); util::image::buffer<float> img (res);
// XXX: offset slightly to avoid origin artefacts in some basis functions
static const auto OFFSET = util::vector2f { -100 };
{ {
// XXX: offset slightly to avoid origin artefacts in some basis functions for (size_t y = 0; y < res.h; ++y)
auto offset = util::vector2f { -100 }; for (size_t x = 0; x < res.w; ++x)
img.data ()[y * img.s + x] = t (util::point2f {float (x), float (y)} + OFFSET);
}
// working on the assumption that all octave images are based on summation,
// subtract the image with one less octave from our current image to leave
// us with the highest octave contribution only. this is hideously
// inefficient, but it's not an operation we care about in general.
if (single && f.octaves () != 1) {
auto oldoctaves = f.octaves ();
auto prev = img.clone ();
for (size_t y = 0; y < res.h; ++y) for (size_t y = 0; y < res.h; ++y)
for (size_t x = 0; x < res.w; ++x) for (size_t x = 0; x < res.w; ++x)
img.data ()[y * img.s + x] = t (util::point2f {float (x), float (y)} + offset); prev.data ()[y * img.s + x] = t (util::point2f {float (x), float (y)} + OFFSET);
CHECK_EQ (img.stride (), prev.stride ());
for (size_t i = 0; i < img.size (); ++i)
img[i] -= prev[i];
f.octaves (oldoctaves);
} }
// rescale into the range [0, 1] // rescale into the range [0, 1]