|
BradC
|
 |
« Reply #15 on: January 05, 2010, 07:21:42 AM » |
|
The problem with multithread :
- With a standard mandelbrot rendering, computing one point write one pixel.
- With buddhabrot, computing one point write on the whole image.
- with millions of point written per seconds, the problem will spend all its time to wait for mutex. Unless i accept some consistancy problem... the effect could be negligible.
Could you have the different threads write to different arrays, and then combine the different arrays down to one final array at the end? |
|
|
|
|
Logged
|
|
|
|
|
ker2x
|
|
« Reply #16 on: January 05, 2010, 08:02:32 AM » |
|
Yes, it should work too.
btw, i added the color version of the Wingedbrot :
|
|
|
|
Logged
|
|
|
|
|
|
|
ker2x
|
|
« Reply #18 on: January 05, 2010, 06:28:38 PM » |
|
i zoomed in the picture above and changed the colors.
|
|
|
|
Logged
|
|
|
|
kram1032
Fractal Iambus
Posts: 922
|
|
« Reply #19 on: January 05, 2010, 06:37:08 PM » |
|
where there any details in that white area?
I don't really like the gradients you chose for killing details...
|
|
|
|
|
Logged
|
|
|
|
|
ker2x
|
|
« Reply #20 on: January 05, 2010, 08:58:41 PM » |
|
i don't know. there is absolutly no post-processing and no color palette. and i can't change it without post processing. I will explain how i color it. but i'm too tired to explain it in english right now. It is the same process as explained here : http://www.superliminal.com/fractals/bbrot/bbrot.htmAfter a long time generating greyscale images I realized that there is a natural way to use color to display more information within the buddhabrot images. Notice that basic buddhabrot images are generated by choosing a "maximum iterations" threshold just as for Mandelbrot images. One main difference between the two techniques is that buddhabrot images have distinctly different appearances depending on the choice of threshold, whereas the effect of different threshold values for mandelbrot images only changes the amount of black (unresolved) pixels. I realized that I should be able to generate meaningful color buddhabrot images by generating three basic images that differ only in the choice of threshold values, and then combining those images as the red, green, and blue channels of a single color image. This is exactly the same technique that astronomers use when generating "false-color" images of astronomical objects. For example, see the famous Eagle Nebula images from the Hubble Space Telescope and read the associated descriptions of color astronomomical mages. For my color buddhabrot images the three different threshold values are analogous to the different frequencies of light which NASA combined into their beautiful false-color images. For the image below I used threshold values of 500, 5000, and 50000, and assigned them to the blue, green, and red channels respectively in order to generate images that most resemble the NASA nebula images. I'm wondering if i could rewrite a buddhabrot renderer for Ultrafractal5 ... |
|
|
|
|
Logged
|
|
|
|
kram1032
Fractal Iambus
Posts: 922
|
|
« Reply #21 on: January 05, 2010, 10:24:02 PM » |
|
hmmm...
interesting: colours for that don't really look right...
It especially looks like it lacks the yellow part a lot.
Which bailout values are you using?
|
|
|
|
|
Logged
|
|
|
|
|
ker2x
|
|
« Reply #22 on: January 06, 2010, 06:31:30 PM » |
|
i don't remember for this one.
i should print the parameters on the screen, so i can remember.
I usually use the green color for high iteration count, because i don't like green and with this kind of fractal the high iteration pixels are rare.
in this picture, all the green pixel are in the "white area". (because you need green pixel make white pixel, heh !)
i tried with differents values, iteration. the white area don't have interetings details.
|
|
|
|
|
Logged
|
|
|
|
kram1032
Fractal Iambus
Posts: 922
|
|
« Reply #23 on: January 06, 2010, 07:12:49 PM » |
|
ok  It looks surprisingly dull in the red region too... |
|
|
|
|
Logged
|
|
|
|
|
stigomaster
|
|
« Reply #24 on: February 06, 2010, 09:20:18 PM » |
|
My winged buddhabrot/nebulabrot looks quite different from yours, I wonder why. Parameters:
Top left: -2, 1.5
Bottom right: 2, -1.5
Iterations: R = 10000, G = 1000, B = 100
Bailout: 20*20
|
|
|
|
Logged
|
Temporary placeholder for signature
|
|
|
|
stigomaster
|
|
« Reply #25 on: February 06, 2010, 09:38:10 PM » |
|
I tried another variant as well, it's a little nice.
This is the mandelbar orbit plot of points escaping the mandelbrot set.
For some reason I named it James.
|
|
|
|
Logged
|
Temporary placeholder for signature
|
|
|
kram1032
Fractal Iambus
Posts: 922
|
|
« Reply #26 on: February 06, 2010, 11:34:28 PM » |
|
That's beautiful  I really need to implement the three-colour method in a reasonable way... (What I do right now, if I use the tribailout colouring method, is to render three images seperately... It would be better to reuse values from lower bailout in higher ones and colour the results directly...) |
|
|
|
|
Logged
|
|
|
|
|
Nahee_Enterprises
|
|
« Reply #27 on: February 07, 2010, 05:33:48 AM » |
|
I tried another variant as well, it's a little nice.
This is the mandelbar orbit plot of points escaping the mandelbrot set.
For some reason I named it James. I guess your monitor must be adjusted quite a bit different from mine, for this image was so dark when I viewed it that hardly anything was noticeable within the graphic. |
|
|
|
|
Logged
|
|
|
|
|
Timeroot
|
|
« Reply #28 on: February 07, 2010, 08:59:30 AM » |
|
Awesome pictures. That modification to the formula you made is actually identical to the Chaosbrot formulas... there's a thread about them here somewhere... Has anyone written a general class-based buddhabrot algorithim for UF? Loading one to determine whether its orbit is included or not, choosing inside/outside, then choosing a second whose orbits to plot. It would be quite useful. Has anyone made a 3D buddhaghost based on triplex numbers? Or a Julibrot-equivalent buddhabrot? I think the z^2 buddhaghost would look quite frightening... Buddhabrots would lend themselves well to distributed computing, would they not? |
|
|
|
|
Logged
|
Someday, man will understand primary theory; how every aspect of our universe has come about. Then we will describe all of physics, build a complete understanding of genetic engineering, catalog all planets, and find intelligent life. And then we'll just puzzle over fractals for eternity.
|
|
|
|
ker2x
|
|
« Reply #29 on: March 07, 2010, 10:10:29 PM » |
|
My winged buddhabrot/nebulabrot looks quite different from yours, I wonder why. Parameters:
Top left: -2, 1.5
Bottom right: 2, -1.5
Iterations: R = 10000, G = 1000, B = 100
Bailout: 20*20
rotate by -90° and set iteration to a very low value (i don't remember exactly but something below 20) |
|
|
|
|
Logged
|
|
|
|
|
Dilber MC Theme by HarzeM