Title: Buddhabrot Post by: tit_toinou on February 21, 2012, 01:01:09 PM A Buddhabrot : http://titoinou.deviantart.com/art/Buddhabrot-100k-600k-286331491 (http://titoinou.deviantart.com/art/Buddhabrot-100k-600k-286331491)
With the technique http://erleuchtet.org/2010/07/ridiculously-large-buddhabrot.html (http://erleuchtet.org/2010/07/ridiculously-large-buddhabrot.html). Title: Re: Buddhabrot Post by: tit_toinou on February 24, 2012, 04:38:28 PM (http://fc01.deviantart.net/fs71/i/2012/055/0/d/buddhabrot_100k_600k_by_titoinou-d4qh2sj.png)
And a new Buddhabrot i juste made : (http://fc01.deviantart.net/fs71/i/2012/055/8/9/buddhabrot_200k_700k_by_titoinou-d4qsuh2.png) Title: Re: Buddhabrot Post by: ker2x on February 24, 2012, 08:05:42 PM Shiny \o/ ;D
Title: Re: Buddhabrot Post by: Eric Bazan on February 25, 2012, 12:05:27 AM I like the last, most colorful one the most. However they do not appear completely developed as there's an asymmetry about the real axis.
How are you coloring these? Title: Re: Buddhabrot Post by: kram1032 on February 25, 2012, 01:24:20 AM Those are awesome :D
So you ONLY take those orbits close to the border of the set? Also, it LOOKS like you colour each and every orbit you find in a different random colour? Really nice stuff :) Eric: If you look at the links he provided, you'll see that the asymmetry is basically wanted (but a result of the technique sampling only a limited range of escape times and thus dropping out a lot of stuff) Though if you really want symmetry, you could always just mirror it and do an overlay... Title: Re: Buddhabrot Post by: Dinkydau on February 25, 2012, 01:25:24 AM The last one looks really awesome!
Title: Re: Buddhabrot Post by: aluminumstudios on February 27, 2012, 08:19:04 AM That second one is really outstanding. The coloring is great and its so smooth looking. Excellent work!
:thumbsup1: Title: Re: Buddhabrot Post by: tit_toinou on March 02, 2012, 11:45:05 AM Thanks for the comments.
@Eric Bazan: The asymmetry is wanted for aesthetics purposes. And you're right, theses Buddhabrots are not "fully developed". But at theses iterations i don't know if we can compute a complete Buddhabrot (and I don't know if it'll be interesting). @kram1032: "So you ONLY take those orbits close to the border of the set?" Yes, very close. "Also, it LOOKS like you colour each and every orbit you find in a different random colour?" I have "layers" of orbits whose iterations are between two integers. And then I pick a color for this layer (it is not random). The lightness of the color is set according to the density of orbits (not with a linear function of course, we wouldn't be able to see anything..). For example in the last image, the green is from orbits between 560k and 700k. In the next version of my software I will be able to color every orbits separately ;D . Title: Re: Buddhabrot Post by: youhn on April 19, 2015, 10:54:45 AM Just digging up some beautiful images. That symmetrical Buddhabrot, let me find the current topic on that.
Found! http://www.fractalforums.com/new-theories-and-research/buddha%27s-jewel-a-special-subset-of-the-buddhabrot-symmetrical-about-both-axis!/ Is it the same? Quote from the topic started by billtavis: Quote from: billtavis The orbits drawn are for points which are very close to the set and yet take a small number of iterations to escape. In other words, near the tips of branches, equally distributed around the set weighted by the size of the branch (the area around the needle has the most orbits drawn). In this case, all the points I drew were closer than 1e-20 to the set and took less than 200 iterations to reach a very large radius. Most points took less than 100 iterations to escape. There's probably a more precise (i.e. mathematical) way to describe the set of points but I don't know what this would be. Maybe someone here on the forums has an idea? And ... Quote In the next version of my software I will be able to color every orbits separately grin . Wouldn't that give a very meshy greyish results ... ? :hmh: Title: Re: Buddhabrot Post by: tit_toinou on April 19, 2015, 06:25:29 PM Just digging up some beautiful images. That symmetrical Buddhabrot, let me find the current topic on that. Found! http://www.fractalforums.com/new-theories-and-research/buddha%27s-jewel-a-special-subset-of-the-buddhabrot-symmetrical-about-both-axis!/ Is it the same? Quote from the topic started by billtavis: Hi ! No It's not the same. I'm just looking for high iteration orbits because they stay close to points inside mandelbrot for a long time *yet* they escape at some point. I also found images on the internet showing that high order points inside the mandelbrot might be interesting too ! (https://camo.githubusercontent.com/283f2f2d6afd697587e21d6514f64c6f6bad3c1b/68747470733a2f2f7261772e6769746875622e636f6d2f72626172742f62756464686162726f742f6d61737465722f6578616d706c65732f616e746962756464686162726f745f687567655f63726f702e6a7067) (https://github.com/rbart/buddhabrot) It indicates that the Anti-Buddhabrot is not ugly, we just misrepresented it by taking a lot of small iterations points that hide the true high order interesting points and also by plotting too many times the same points that belong to the same cycle. It means that to me, the most beautiful orbits are thoses very close to the border of the mandelbrot set outside (escaping) OR inside (high order). Some research should be done regarding this ! ( By high order I mean very long cycle, z_{n+p}=z_{n}, p is very high. ) Another visualisation of the anti buddhabrot would be to plot only cycles 1 times (and not x times because we run the iterations p^x more times without cycle-checking) OR only plot the orbits before entering in a cycle (that would of course require cycle detection). Buddha's jewel is about to find points close to the set yet escaping very fast. I think (by intuition) that the points concerned are thoses where the derivative of the distance to the set is very high = high values of the Douady-Hubbard Potential = where the thunderstorm would make his way if the inside of the mandelbrot set was electrically charged = near the " needles ". You can look at the images at the end of the first pages posted here : http://www.fractalforums.com/images-showcase-%28rate-my-fractal%29/classic-mandelbrot-with-distance-and-gradient-for-coloring (http://www.fractalforums.com/images-showcase-%28rate-my-fractal%29/classic-mandelbrot-with-distance-and-gradient-for-coloring), you'll see white needles ! Wouldn't that give a very meshy greyish results ... ? :hmh: Hum no because I would have a lot of points, I would take points that go on for 1 millions iterations before escaping, giving weird shapes (but individually colored) like in the images posted here.Title: Re: Buddhabrot Post by: 3dickulus on April 19, 2015, 09:28:47 PM That last image (b/w), to me, is the most interesting with intricate delicacy that leads the eye deeper and deeper. :beer:
Title: Re: Buddhabrot Post by: tit_toinou on May 20, 2015, 07:19:35 PM A friend of mine is also exploring Buddhabrot the way described in this post.
Here's his first image, more is coming ;) (http://nocache-nocookies.digitalgott.com/gallery/17/4891_20_05_15_3_56_18.jpeg) (http://www.fractalforums.com/index.php?action=gallery;sa=view;id=17637) Title: Re: Buddhabrot Post by: kram1032 on June 04, 2015, 04:54:21 PM what kinds of values are you using for these?
If I understood correctly, what you are doing is that you only look at a fairly narrow range of iteration counts, only even beginning to plot them at high values. So what is the minimum value of iterations you consider in these images? I'm currently trying a classical BBrot image with:
And I further restrict the initial selected points such that both the initial point and the first iteration are inside the circle of radius 4 but neither in the main period-1-cardioid nor in the circle of period 2 to ensure that the selected points are at least reasonably close to the border of the set. Is there anything else you are doing besides that? What are your choices for the various values? Title: Re: Buddhabrot Post by: kram1032 on June 05, 2015, 01:34:48 AM Here is the not-nearly-as-dramatic-as-yours result of said settings. (Click on image for larger view)
(http://i.imgur.com/aHQDqsG.png) (http://i.imgur.com/aHQDqsG.png) Title: Re: Buddhabrot Post by: Caleidoscope on June 05, 2015, 10:05:37 AM Wow, it is beautiful, awesome colors and lovely details. :)
Title: Re: Buddhabrot Post by: kram1032 on June 06, 2015, 01:23:57 PM A second one. (Click the image for some extra information/larger view)
(http://orig07.deviantart.net/8e94/f/2015/157/b/a/life_within_life___deep_rendered_buddhabrot_2_by_kram1032-d8w7icb.png) (http://kram1032.deviantart.com/art/Life-within-Life-Deep-Rendered-Buddhabrot-2-537827483) This time, I used the following settings:
(I chose those values such that they are logarithmically evenly distributed) Title: Re: Buddhabrot Post by: Caleidoscope on June 06, 2015, 02:08:00 PM Again a precious gem. Still very different from the other one. Makes me think of artistic Chakra's.
Great picture. Title: Re: Buddhabrot Post by: cKleinhuis on June 06, 2015, 05:10:32 PM can you please try something out ?
what i dont like with the buddhabrots is the randomness of the catched orbits, can you try this? if a long period that bails out is found, e.g. more than 1million, take that location and use a miniminimi radius around that starting value and draw in the budhabrot orbits of those Title: Re: Buddhabrot Post by: kram1032 on June 06, 2015, 05:42:47 PM you mean to make sure that the whole thing is filled more evenly?
I feel like part of this rendering style specifically is the unevenness of it. But it sounds like a nice experiment. I wonder what nice values for a "miniminimi" radius would be. Do you think 1.e-3 would be close enough? EDIT: Ok I launched an experiment. I hope this is not an infinite loop now: As currently implemented, it will do a random walk around anything that's still inside the set and it'll continue to do so until: either it has found a point that DOES diverge or it has gone a previously specified number of steps. (I made it finite since, with a random walk, it might be that it just runs right into the regions that DON'T escape and thus it'd go on practically forever which would... not help.) We'll see what this does. Currently I'm using the adhoc chosen trial depth of 2 steps and a random radius between 0 and 1.e-3 for my steps. Since this actually KEEPS running after whenever it needs the LONGEST and on top of that it will find a point close to that which has a sizeable chance to ALSO be amongst those points that take the longest, I expect this to take FOREVER to render. But we'll see. EDIT: ugh. I think I really need to refactor my code to make it work right. Currently I have to be REALLY careful with what I do. One wrong step and POW, it somehow renders an antibuddhabrot instead of a buddhabrot. The change I did, did just that. I think I need to reorganize my loops a bit for this to work. Such a change is LONG overdue and I'm working on making that happen now. Though it'll take a bunch of time before I get anything out of those changes, so if somebody else would like to try this, please do. Title: Re: Buddhabrot Post by: tit_toinou on June 08, 2015, 12:32:48 AM Nice images kram1032 !
Sorry for not answering I just saw your posts... what kinds of values are you using for these? I think Berdes was using 1k-10k (red) / 10k-100k (yellow) / 100k-1000k (white).If I understood correctly, what you are doing is that you only look at a fairly narrow range of iteration counts, only even beginning to plot them at high values. So what is the minimum value of iterations you consider in these images? I'm currently trying a classical BBrot image with:
And I further restrict the initial selected points such that both the initial point and the first iteration are inside the circle of radius 4 but neither in the main period-1-cardioid nor in the circle of period 2 to ensure that the selected points are at least reasonably close to the border of the set. Is there anything else you are doing besides that? What are your choices for the various values? Checking the cardiod and biggest circle is a good idea. You can also implement a simple periodicity checking algorithm. No nothing else is done ! Another better buddhabrot he made : (http://www.bde.enseeiht.fr/~desmytb/buddha_1000.png) (http://www.fractalforums.com/images-showcase-%28rate-my-fractal%29/berdes-buddhabrot/) you mean to make sure that the whole thing is filled more evenly? I understand cKleinhuis ' idea. I had the same. I feel like points close to each others yield orbits that look alike.I feel like part of this rendering style specifically is the unevenness of it. But it sounds like a nice experiment. I wonder what nice values for a "miniminimi" radius would be. Do you think 1.e-3 would be close enough? I think the radius has to be proportional to the distance to the set. You can crunch it with this formula http://iquilezles.org/www/articles/distancefractals/distancefractals.htm (http://iquilezles.org/www/articles/distancefractals/distancefractals.htm). My friend gave me a file with the list of points of the orbit, I plotted it : (http://nocache-nocookies.digitalgott.com/gallery/17/4891_08_06_15_12_23_10.png) (http://www.fractalforums.com/index.php?action=gallery;sa=view;id=17896) 5 million iterations plotted at the upper left. The others things are enlargement of bulbs rotated and scaled. It appears that they are linked by a simple conformal transformation z->az+b strange.... Red was added for the first third of the orbits (iteration < 5 million / 3), green was added for the second third of the orbits and blue was added for the last third. Discovers :
Title: Re: Buddhabrot Post by: ker2x on June 08, 2015, 02:56:12 PM Cardiod and circle check in this function :
Code: PURE FUNCTION notInMset(c, n_max) Or some other (similar) code in C++ : Code: //Quick rejection check if c is in 2nd order period bulb And some old openCL code : Code: bool isInMSet( Title: Re: Buddhabrot Post by: kram1032 on June 08, 2015, 03:17:56 PM yeah I did that already. Not hard to do :)
Title: Re: Buddhabrot Post by: Alef on June 08, 2015, 04:37:19 PM Quite a good. Nice colours of each image, not too alien as original. |