BuddhabrotĪctually there are two versions of this plot: including points inside the set and excluding them. softhours (and the last one, although, it’s too hard to see)Īnd, again, there are versions of HSV based coloring, grayscale and blue/red palettes.fire and oceanid (gradients based counterparts of Fire Red and Ocean Blue).I decided to just ignore this part (I can generate the bigger image and downscale it any time I want, why bother).Īnd if you were looking for some cool gradients (including “the one from wikipedia”), here are some that I (hardcoded) in the mandel module: Most of the time people just calculate the source image with a twice higher resolution, then downscale the image. There’s another problem you need to keep in mind: aliasing. Saturation value is a bit off, but I like it this way I’ve got the best result with this:Ĭonst auto vc = v > 1 ? 2-v : v auto vl = 0.25+vc*2 auto vl = 0.25+vc*2 auto vs = 0.75+vc*2 auto c = hsv( 360.0*v, vs > 1 ? 1 : vs, vl > 1 ? 1 : vl ) For that you need some additional adjustments, dependant on the library you’re using to work with HSV. But it would be stupid to just feed the v value to these three channels, as with gray gradient. While in RGB colorspace every color is divided into red, green and blue channels, in HSV the first value sets the hue (usually set as an “angle” or a position on a circle), second sets saturation and third sets the value. The most interesting part is using HSV to generate colors. You can look for my multipliers for Fire Red and Ocean Blue generated gradients in the same mandelmodule, function pixelcolor() And if you want to color it, you can just use only one channel (Red/Green/Blue) instead.īut it doesn’t really look as nice with this simple approach, so it’s better to find some nice multipliers to assign to other channels, so the picture won’t look as dark.
0 kommentar(er)
