Welcome to Fractal Forums

I fired up this pic for an avatar a year or two ago, then went and deleted the fractal I used... the problem is, that part of the Mandelbrot set has a very particular density and layout of colour, and I haven't been able to find another part of the set that still looks much like a fractal at very low resolution...

So if anyone recognises whereabouts this is from the combination of features apparent here, and can give me a rough idea where to look, that would be awesome.

(http://i4.photobucket.com/albums/y111/Kimmo666/largemeblkfract.jpg)

If someone can tell me where to find this spot, they are truly steeped in the ways of the Thumbprint...


...Oh yeah, it was done in Fractint 20 with all default values.

Well, it was easy enough to see that it's from somewhere in the center/far-right spike but I couldn't find that combination where you have the two focal points and the central feature.  The area I was looking around in was 0.28, 0.01 & 500 mag.  I'm using Ultra Fractal now but gods was Fractint a ball for so many years.  It's the real reason I got a 387 chip, not Acad!  :D

Good luck in your search, lot o landscape!

'tomix

Hmm, just had a squiz at that location, and can't seem to find anywhere all that promising there... oh well... I'll keep poking around.
Thanks for your help, though : )

Ater having a browse around on that spike with UF, I found this interesting area here...
(http://i20.tinypic.com/11gj7ev_th) (http://i20.tinypic.com/11gj7ev.jpg)
center=0.2506812972612106 / 0.000028362307663746385
magn=3000000
Rendered @ 2048x1536, then resized 150% so I could soften it very slightly before being resized down to 1600x1200.
Anti-aliased, basically.

I got there by zooming in on this promising bit.
(http://i20.tinypic.com/2v1ok5j_th) (http://i20.tinypic.com/2v1ok5j.jpg)
center=as above
magn=100000
Rendered @ 1600x1200

You can see how that antialiasing technique reduces noise and shows more detail than this plain 1600x1200 copy of the first image.
(http://i20.tinypic.com/f4ef6x_th) (http://i20.tinypic.com/f4ef6x.jpg)

...Damn Tinypic, reduced the quality of those jpgs... they still look pretty sweet though.

At least the antialiased one does... hm, I wonder where the diminishing returns stop? My 21" CRT displays every pixel @ 1600x1200, but not every one gets its chance to shine @ 2048x1536. However, I guess there's a kind of antialiasing going on within the monitor with the scan pattern having higher resolution than the shadow mask... I know stuff looks nicest on my monitor @ 2048x1536, except for the 75Hz flicker...
Up until just now, I've only thought of rendering fractals at up to 2048x1536, thinking that softening effect was making it look as good as possible, but comparing these two versions of the same pic here, I just realised -

1:1 pixels calculated to pixels displayed gives you garbage for detail. Softened garbage is just softened garbage.

So what's the best way to do it? Is 150% larger than final size enough? Or is 200% the go? Is there any possible value to starting even larger?
The existence of 4xAA modes on graphics cards suggests so, but I gather it really is a matter of diminishing returns...

I hear ya, I can't wait to upgrade from my 1.6 Athlon, for the same reason among others... but WTF's an Acad? Never heard of it.

Wouldn't be these guys (http://www.acadcomp.com/) by any chance, you reckon? ; )

Cheers... even the upper area we can get to without our PCs bogging down is pretty freakin huge...

antialiasing is a complex and subtle topic, mixing numerical analysis with aesthetic judgement. i have a bunch of great online references if you're interested to learn more about it - it really is a fascinating subject.

the essence of what it tries to do is produce "a perfect signal"; i use quotes because we can never hit perfection in practice, and this is where the engineering comes in. the idea is to keep exactly as much information as the resolution can hold, and no more. to make an analogy with audio recording, if you're producing a cd-quality (44.1khz) track, you want to eliminate all frequencies above 22.05khz (the nyquist limit is always 1/2 sampling rate) because:

1. the sampling rate is incapable of representing higher frequencies
2. in trying to do so it negatively affects (through aliasing) the quality of the signal

so antialiasing is nothing more than jamming as much information into a signal as is possible, and no more!

people have differing views about antialiasing, some think it's just a blur filter applied in postprocess which somehow reduces the crispness of the resulting signal (?!), but it's well established that antialiasing of the central concern when producing high quality images/audio.


the take-away point is this: if your image is NxM resolution, let it be NxM perfect pixels ;) put as much cpu power into antialiasing your works as you can, and have faith that the software you're using (if you didn't write it yourself) will do a good job!

Kimmo,

I think I'm pretty close. I looked around in "Elephant Valley (the cleft of the main cardioid) and found this:

(http://www.pbase.com/image/88297566/original.jpg)

It's centered at about 0.32136, -0.03512i, at a width of about 4E-5 (or magnification of about 75,000)

I used monochrome colors to show the detail most clearly.

It looks to me like the image in your avatar is rotated some, and you erased the center tendril between the two clusters, so it didn't obscure your eyes.

I created this image with my app, FractalWorks. If you have a bigger version of the avatar image I can try to duplicate the colors.



Duncan C

Wow... I'm blown away, Duncan : D

But I didn't alter the fractal at all... I prolly rotated it, though.

You're obviously very close, with those virtually identical elements there...

Thanks a lot; I'm off to have a browse...