Home » forums » Programmer Forum

New Flame Paper

Submitted by spot on Fri, 2008-02-15 03:02.

It's been a long time coming but Erik and I are prereleasing a new and updated version of the fractal flame paper. You can download it here. This is the definitive explanation of all aspects of the visual language that's the basis for Apophysis and Electric Sheep.

‹ FLAM3 2.7.14 released Time frame on full quality for the beta? ›
» login or register to post comments

Can't download new paper...

Submitted by apalomba on Tue, 2008-07-08 15:30.

I have been trying to download the new paper but the download stalls.
I am using Firefox, WinXP. Is there some kind of limiter in this file?

» login or register to post comments

i just tried the link to the

Submitted by spot on Tue, 2008-07-08 17:12.

i just tried and it downloaded fine in about a minute.

» login or register to post comments

I still can not download

Submitted by apalomba on Thu, 2008-07-10 12:22.

I still can not download this file. Anyone else have trouble with it?

» login or register to post comments

Is this how it works?

Submitted by Zippo251 on Mon, 2008-03-03 22:58.

Ok so I am really trying to understand how the numbers in IFS work for fractals, but I'm only a senior in high school so there is still a lot I need to learn in the way of math. The notation is half my problem.

this is how I understand how it works:

to make the Sierpinski triangle first a point is chosen (I still don't understand how the initial point is chosen)
Then a random function is chosen F1, F2, or F3

F1 = input is the previous point and output is halfway to (0,0)
F2 = input is the previous point and output is halfway to (1,0)
F3 = input is the previous point and output is halfway to (0,1)

for each pixel there is a counter, when the point falls into the area of the pixel it's counter goes up one.
the new point is the input of the next random function.
this repeats thousands of times (or more).
the pixels with its counter greater than zero are colored in. the ones that aren't are transparent. the ones with the largest count are completely opaque and the ones with less count are slightly transparent.
all of these things things come together to make an image (not including the other enhancements Flam3 has.)

Is this how it works or am I completely off?

» login or register to post comments

Initial points

Submitted by Scotoma on Wed, 2008-04-09 16:03.

There is more than one initial point. The initial points are randomly disbursed throughout the biunit square (the square that covers x from -1 to 1 and y from -1 to 1). I am confused as to whether all these points are transformed in parallel (all at once) by the same randomly (well with a certain probability) chosen function, or if a different function is used for each point. My guess is that probably the later happens.

» login or register to post comments

I believe so

Submitted by denaje on Tue, 2008-03-04 12:01.

It's been a while since doing Sierpinski's triangle, but that sounds right. Each point is chosen randomly from inside the triangle and then moved halfway to one of the vertices. This is repeated until you have the image. It would make sense that the pixels with the greatest counts are opaque and the ones that don't have any are transparent.

--
Jesse Denardo

» login or register to post comments

Nice work! That Xform list

Submitted by keldor314 on Mon, 2008-02-18 07:29.

Nice work! That Xform list was in bad need of updating. :-)

» login or register to post comments

Hole in NGon (variation 38) documentation

Submitted by keldor314 on Wed, 2008-03-12 13:36.

The documentation for NGon seems to have a couple of lines that accidently got deleted. t2 (and presumably t1) are not defined anywhere.

» login or register to post comments

Thx for the catch

Submitted by Erik Reckase on Sun, 2008-03-16 08:35.

t2 should actually be p2 in that calculation. I've fixed that and sent the correction to spot.

Erik Reckase
flam3 developer
AIM: flam3dev

» login or register to post comments

NGon docs

Submitted by keldor314 on Fri, 2008-03-21 01:24.

In that case, why is there only t3 and t4? What happened to t1 and t2? Given that all the other xforms have inside functions listed as t1, t2, t3... it might be a good idea to change this to avoid confusion.

» login or register to post comments