|
rickz65
Guest
|
 |
« on: November 21, 2009, 01:45:41 AM » |
|
Hey Now ! How's everybody doing ? I found this forum from a link in the skytopia page by Daniel White about the Mandelbulb. Very neat, new stuff, check it out if you haven't yet: http://www.skytopia.com/project/fractal/mandelbulb.html . Anyway, A little about myself: I've been fascinated by Fractals ever since reading a Scientific American article about the Mandelbrot set way back in July or August 1985. I've written a few generators over the years for my own uses, and currently am working on one that is rather fast, for now, limited to IFS in 2D, but has many nice features such as high precision mathematics(beyond 16 digits) which is also quite fast, user entered equations(also quite fast), and, well, a bunch of nice features. The App. is written in C++, and some Assembler(x86 SIMD), for now just runs on Windows, and uses a multi-threaded calculation engine that can take advantage of multi-core CPUs. Although I've worked with various programming languages for 30 years now, the web stuff is not at all my strong suit, so I don't have a site yet for my new creation, but It is almost ready for initial publishing. It's Freeware, of course, as I don't really see why anyone would spend money on a generator, other than a donation, and besides Freeware Rocks! Never-the-less looking at the multitude of types of Fractal images and 3D images out there(and here), I feel like I've been living under a rock. Hope to meet other fractal Authors and Artist at this site and learn more (I always give credit in my software (in the About Box) to those who have helped me in anyway). Here's to talking to you soon !  |
|
|
|
|
Logged
|
|
|
|
|
David Makin
|
 |
« Reply #1 on: November 21, 2009, 02:06:42 AM » |
|
|
|
|
|
|
Logged
|
|
|
|
|
rickz65
Guest
|
|
« Reply #2 on: November 21, 2009, 03:36:56 AM » |
|
Thanks for the Welcome, David. Glad to be here. I downloaded some commercial software awhile back that I think was called Ultra-fractals, it had a 30 day trial period and cost somewhere around $100(US), is this the same program? Unfortunately, I uninstalled it after it crashed on me just trying to produce a simple Mandelbot set at a particular set of parameters so I could judge it's speed against my software. For the Curious, I'm able to generate a Mandelbrot image for -2-2i to 2+2i (the whole set) to 200 iterations in 500x500 pixel image in 7.4 milliseconds on a 2.0 GHz dual core AMD Turion X2(RM-70) on Vista ( 6.8 ms on win 7) and drops to 6.2 ms (on Vista) if I use SSE which limits the resolution to 8 digits. if you've any ideas as to how to impliment a fast ray-tracing renderer for doing non-affine 3D IFS using the escape-time method then I'd love to hear them I'ld love to help you, but you'll have to excuse my ignorance, what is 'affine' ? I might be familiar with the concept without knowing it's formal name, which is true for me with most aspects of Fractal geometry, as I have no formal education in the subject. Do you know a good link for an explanation of Hart's method ? Strangely, I can usually understand something better if it's in code, rather than pure mathematical jargon. I'm much more of a programmer than mathematician. Thanks for the tip about deviantart. I'll put togeather a movie posting in a little while of some interesting example of what I like to do with my software. It has the Fractal image respond to music being played on the computer. Makes for nice eye candy, although, with certain settings can cause seizures |
|
|
|
|
Logged
|
|
|
|
|
David Makin
|
|
« Reply #3 on: November 21, 2009, 04:17:56 AM » |
|
Thanks for the Welcome, David. Glad to be here. I downloaded some commercial software awhile back that I think was called Ultra-fractals, it had a 30 day trial period and cost somewhere around $100(US), is this the same program? Unfortunately, I uninstalled it after it crashed on me just trying to produce a simple Mandelbot set at a particular set of parameters so I could judge it's speed against my software. For the Curious, I'm able to generate a Mandelbrot image for -2-2i to 2+2i (the whole set) to 200 iterations in 500x500 pixel image in 7.4 milliseconds on a 2.0 GHz dual core AMD Turion X2(RM-70) on Vista ( 6.8 ms on win 7) and drops to 6.2 ms (on Vista) if I use SSE which limits the resolution to 8 digits. if you've any ideas as to how to impliment a fast ray-tracing renderer for doing non-affine 3D IFS using the escape-time method then I'd love to hear them I'ld love to help you, but you'll have to excuse my ignorance, what is 'affine' ? I might be familiar with the concept without knowing it's formal name, which is true for me with most aspects of Fractal geometry, as I have no formal education in the subject. Do you know a good link for an explanation of Hart's method ? Strangely, I can usually understand something better if it's in code, rather than pure mathematical jargon. I'm much more of a programmer than mathematician. Thanks for the tip about deviantart. I'll put togeather a movie posting in a little while of some interesting example of what I like to do with my software. It has the Fractal image respond to music being played on the computer. Makes for nice eye candy, although, with certain settings can cause seizures How long ago did you try Ultra Fractal ? I'm amazed it crashed unless you're running under WINE or something - it's been stable since version 2 - around 2000  It's impossible to time anything using UF's timer more accurately than to around 0.1 of a second, even a blank "pixel" formula registers some time in the counter displayed in UF (though the real time is actually virtually nil in that case). In UF my formula for 3D IFS renders a Menger Sponge in just over 4 seconds @640*480 on a 2GHz core2duo (to well beyond the actual display resolution), considering that the escape-time method is inherently slower than the convergent method when rendering a "whole" IFS this compares well with render times in Xenodream. My test quaternion renderer renders the z^2+c quat in just over 4 seconds @640*480 on this decrepit old 3GHz P4HT which rarely manages better than 2GHz due to throttleback because of overheating. The real reason I was recommending UF was because of the speed/ease of coding up new algorithms you want to try out, the only free software I know that can compare in that sense is ChaosPro. Also it makes it easier to browse through the formulas from the UF database to get ideas I'll have to dig around to find the link I used for Hart's method - I think I got it off lycium so you could try messaging him through fractalforums. |
|
|
|
|
Logged
|
|
|
|
|
lycium
|
|
« Reply #4 on: November 21, 2009, 05:04:22 AM » |
|
being "hart's method" one should be able to find it on john c hart's publications page: http://graphics.cs.uiuc.edu/~jch/papers/ (near the bottom) welcome to the forums rickz! |
|
|
|
|
Logged
|
|
|
|
|
rickz65
Guest
|
|
« Reply #5 on: November 21, 2009, 05:31:22 AM » |
|
I tried UF back in mid 2008, I think. The Animation version, which was a new thing for it. Was running Win XP sp2 or maybe 3 if it was out, the hardware and OS were stable and not overclocked, but were somewhat old: 2.66 Celeron on Asus P5P800. I just looked up UF's web site and Yes, it was the same App., just an earlier version, glad to see the price is down to $69. I will say the 'crash' was without a doubt the most graceful crash I've ever seen, and was happy to send the log report in. I'm rather familiar with ChaosPro, and gave up trying to use it because it crashed way too many times on me(dozens, about every 3rd or 4th rendering). My software can make use of a fast math parser if selected to do so, the original parser is called the muParser(found @ CodeProject.com) written by Ingo Berg. I've made many changes to it. The compiled 'byte-code' now runs about 8 to 10 times faster and it can handle 16, 32, 64 digit and Arbitrary Precision math. of Complex Numbers with the 'stage' set (template) for quad. and octo. math libraries. This Allows me to test formulas/equations very quickly and easily. Still, UF I found was about the same speed as my 'User entered' renderings. I feel your pain with the decrepit old P4, I got a p4 640 (3.2 ghz w/ HT) for an upgrade to my old desktop(the 2.66 celeron) but due to job loss, I'm keeping it in my new rig(hopefully for not too long). I got a Xigmatek sd964 cpu cooler for it, and OCed to 4.5 ghz and under full load, it still only gets about 12 degrees C above Ambient. (true, really good case cooling helps too  ... summers get really hot here) Really great cooler for $24, except it's 775 attachment really sucks, so it's hard for me to recommend it. It pops loose on thick boards, until you play with it just right. I'll have to try UF again, to check out your formulas, and to get some ideas from it's database. ... Thank You for the link Lycium !! I'll check that out very soon ... and thank for the welcome. |
|
|
|
|
Logged
|
|
|
|
|
cKleinhuis
|
|
« Reply #6 on: November 21, 2009, 01:06:11 PM » |
|
you can still use uf5, you just can not save the images without watermarking !  |
|
|
|
|
Logged
|
---
divide and conquer - iterate and rule - chaos is No random!
|
|
|
|
David Makin
|
|
« Reply #7 on: November 21, 2009, 01:21:43 PM » |
|
Last night just after I posted my last entry in this thread I realised I hadn't explained "non-affine", I went to add to my message but fractalforums went down and it was time to go to bed.
Anyway "affine" transforms basically means linear transforms i.e. rotation/scaling/skewing and translation only with no non-linear functions in other words basically matrix transforms only, "non-affine" basically means that some non-linear maths is used.
|
|
|
|
|
Logged
|
|
|
|
|
David Makin
|
|
« Reply #8 on: November 21, 2009, 01:25:40 PM » |
|
I tried UF back in mid 2008, I think. The Animation version, which was a new thing for it.
I perhaps should have said "generally stable", there have been maybe a couple of times when something was missed in a release version that caused a crash but it's always got fixed within a week or so - you must just have had bad timing |
|
|
|
|
Logged
|
|
|
|
|
rickz65
Guest
|
|
« Reply #9 on: November 24, 2009, 04:57:24 PM » |
|
Hello Everybody ! I'm back again, sorry to have disappeared for a few days. My Dad is here for Thanksgiving, and I had a lot to do. The Forums here inspired me to finally get off my 'duff' and get my web site togeather so my generator can be downloaded by you folks (and everybody else out there) ... so I invite everyone here at FractalForums to goto http://manyfractals.org and download a copy of ManyFractals.exe for themselves. I don't have much content posted yet, but everything at the site does work. David, Thanks for the explanation about affine/non-affine ... I'll keep it in mind when trying things out in UF5, and when reading Hart's work. Maybe in a week or two, I'll have an Idea that will help your quest. Trifox, Thanks for the tip on UF5, but do you know if the App still works after 30 days ? Right now I don't have the $$ to purchase it. |
|
|
|
|
Logged
|
|
|
|
|
Nahee_Enterprises
|
|
« Reply #10 on: January 05, 2010, 12:06:48 PM » |
|
Hey Now ! How's everybody doing ? I found this forum from a link in the skytopia page by Daniel White about the Mandelbulb.
I've written a few generators over the years for my own uses Hope to meet other fractal Authors and Artist at this site.... Greetings, and a belated Welcome to this particular Forum !!!  Just downloaded ManyFractals awhile ago, will give it a try when I have a spare moment. Thank you. |
|
|
|
« Last Edit: September 18, 2013, 01:03:01 PM by Nahee_Enterprises »
|
Logged
|
|
|
|
|
Nahee_Enterprises
|
|
« Reply #11 on: January 05, 2010, 12:30:26 PM » |
|
How long ago did you try Ultra Fractal ? I'm amazed it crashed unless you're running under WINE or something
- it's been stable since version 2 I tried UF back in mid 2008, I think. The Animation version, which was a new thing for it. I perhaps should have said "generally stable", there have been maybe a couple of times when something was missed in a release version that caused a crash but it's always got fixed within a week or so - you must just have had bad timing Your first response was definitely a falsehood!!! There have been many problems with UF over the years, and I have the quotes and comments from many people to prove it. As to your trying to recover from such a blatant error in truthfulness, there have been way more than "a couple of times". And the fixes did not come out as quickly as you are implying, I can prove that as well. Version 2 (with its many releases) spanned a few years (approximate dates): - UF_2-00-b1 early 1998
- UF_2-00-b2
- UF_2-00-b3
- UF_2-00-b4
- UF_2-00-b5 07/21/1998
- UF_2-00-b6 08/14/1998
- UF_2-00-b7 10/04/1998
- UF_2-00-b8 12/02/1998
- UF_2-04 05/13/1999
- UF_2-05 04/19/2001
Version 3.00 08/20/2002 - UF_3-01a 01/04/2003
- UF_3-02
- UF_3-03 08/21/2003
- UF_3-04 01/02/2004
- UF_3-05 08/19/2004
Version 4.00 - UF_4-01
- UF_4-02
- UF_4-03
- UF_4-04
Version 5.00 - UF_5-00
- UF_5-01
- UF_5-02
- UF_5-03
And each version/release had many bug issues, not to mention all of the crashes that were reported. Far from being "stable" !!!! |
|
|
|
« Last Edit: September 18, 2013, 01:11:17 PM by Nahee_Enterprises »
|
Logged
|
|
|
|
|
David Makin
|
|
« Reply #12 on: January 05, 2010, 01:27:20 PM » |
|
I said "stable" not "perfect" - it's been more stable than any Windows operating system since the very first version Also any problems get fixed (unlike with most freeware where you are basically stuck being unable to do certain things). |
|
|
|
|
Logged
|
|
|
|
|
Dilber MC Theme by HarzeM