From: fractint-owner@xmission.com (fractint Digest) To: fractint-digest@xmission.com Subject: fractint Digest V1 #28 Reply-To: fractint@xmission.com Sender: fractint-owner@xmission.com Errors-To: fractint-owner@xmission.com Precedence: fractint Digest Saturday, October 11 1997 Volume 01 : Number 028 In this issue: Re: (fractint) Integer v. float (fractint) Conformal formulas Re: (fractint) Integer v. float (fractint) And now the PAR you've all been WAITING FOR! See the end of the digest for information on subscribing to the fractint or fractint-digest mailing lists and on how to retrieve back issues. ---------------------------------------------------------------------- Date: Fri, 10 Oct 1997 15:09:02 -0400 From: "Damien M. Jones" Subject: Re: (fractint) Integer v. float [NOTE: This is somewhat technical material. Non-programmers should skip this message.] Evin, - Well, I say keep the integer math. Despite what some people have said, - integer math is, in general, faster than floating point, with a few - exceptions. OK, to clarify the point: on an Intel Pentium, the FPU multiply of 64x64 (mantissa only, exponents are just added) occurs with one-cycle throughput rather than 10-cycle throughput for a 32x32 integer multiply. This is the most common case where the FPU is faster. Other cases are square roots (70 vs. a few hundred, depending on your integer algorithm) and trig or log functions (huge differences). For adding and subtracting, the FPU is *slower*. These instructions have one-cycle throughput on the FPU, but integer instructions on the CPU have throughput of half a cycle (two are executed in one cycle). Thus if your process involves a lot of addition and subtraction, and very little else, you are unlikely to benefit from moving things into the FPU. Division is also about the same speed (39 vs. 41 cycles). It is the disparity between the multiplies and the trig/log functions that makes this such a compelling option, because they play such a large role in fractals. I had a Pentium Pro machine here up until last week, and I spent quite a bit of time generating fractals on it. As far as I could tell, floating-point performance still outran integer performance by a considerable margin. I would dispute the claim that integer multiplies are again faster until I have seen comparative benchmarks that indicate otherwise. MMX will probably not help fractals much. It is true that the MMX multiply is fast--one cycle to do four 16x16 multiplies--but the catch is, it is ONLY a 16-bit multiply. Sixteen bits of precision just doesn't get you very far in fractals; and while it is possible to use this for extended precision, it is highly unlikely to be faster than the FPU's 64x64 multiply, especially when you factor in the rest of the work that needs to be done. - On everything else (P-pro, PII, K6, each better than the regular pentium - and pentium MMX), integer multiplication is faster than floating point. Unfortunately Intel hasn't (to my knowledge) published basic cycle counts for Pentium Pro and Pentium-II instructions. The last published counts I have are for Pentiums. However, the trend from 386 through Pentium is for FPU speed to increase drastically, and integer multiplies to increase only moderately. (13-18 cycles on a 486, 10-11 for Pentium, for integer; 16 cycles on a 486, 1 on a Pentium, for FPU.) I'd be very surprised if the PPro and P-II had radically redesigned integer multiplies, given that overall their performance is not leaps and bounds above Pentiums of the same clock speed. Various indications I have received are that AMD's K6 processor does not *quite* have the FPU performance of their Intel counterparts, but they are still much faster with FPU stuff than Cyrix processors. And even on a Cyrix 6x86, where FPU operations (including multiply) take 4-9 cycles, and multiplies take 4-10, the FPU routines in FractInt are marginally faster even for plain Mandelbrot. Cyrix's 6x86MX processors (the "M2") have the only slightly better FPU performance than the original 6x86. (And I also test in 32-bit protected mode to verify this.) - The problem Fractint has now with integer math is that it's running in - 16-bit mode, so every 32 bit operation requires prefixes which can slow - things down. Actually, the problem FractInt has is that multiply instructions (IMUL and MUL) are *not* pairable instructions on the Pentium, and so not only do they consume 10-11 cycles each, but they *also* prevent other instructions from running at the same time. On the PPro and P-II, the assignment of instructions to available execution pipes is better, preventing one complex instruction from blocking simple instructions that are not dependent on it, but MUL and IMUL are still complex instructions, and there is only one complex instruction pipe. Thus you are still blocked, waiting for those three multiplies for Mandelbrot iteration to go through. Even on a Pentium, the cycle penalty for executing an instruction with a size prefix is only one cycle; compared to the 30 it takes to do your multiplies, this is not enough to account for the discrepancy between integer and FPU math times. - Just porting fractint to 32 bits and keeping the integer math would - speed it up. DJGPP and NASM, please... Yes, it would speed it up. But it would also speed up the FPU routines slightly (which also have to deal with a few integer things, while the FPU is cranking) and the speed-up on the integer side would still not make up for the difference. Would you like to see who can write the fastest 32-bit protected mode M-set generation routine? You write integer code, I'll write FPU code. :) I think the bigger issue is whether it is *worth* porting all that integer assembly code to a 32-bit environment. Me, I don't think it's worth it. Yes, there are some folks who have images based on artifacts in the integer code. I'm not convinced it's worth preserving the artifacting, especially if it hinders the growth of FractInt in other areas. If I had to choose between having synchronous orbits working fast, or integer math support in a 32-bit environment, guess which one I'm gonna choose? (Good thing I don't decide stuff, huh? :) In an ideal world, FractInt would keep support for the oldest PAR files known to man. :) But that isn't always possible. And there's enough work porting FractInt to a 32-bit environment just with FPU support, without needing to move all the optimized integer assembly as well. IMO. Damien M. Jones / temporary sanity designs / http://www.emi.net/~dmj/ dmj@emi.net / my gallery: http://www.geocities.com/SoHo/Lofts/2605/ - ------------------------------------------------------------ Thanks for using Fractint, The Fractals and Fractint Discussion List Post Message: fractint@xmission.com Get Commands: majordomo@xmission.com "help" Administrator: twegner@phoenix.net Unsubscribe: majordomo@xmission.com "unsubscribe fractint" ------------------------------ Date: Fri, 10 Oct 1997 23:39:03 -0400 (EDT) From: ao950@freenet.carleton.ca (Paul Derbyshire) Subject: (fractint) Conformal formulas After some experimentation I determined that conformal mappings aren't enough. In fact to get the sort of equant form of typical Mandelbrots and Julias (esp. the spirals) requires the functions f(x,y) and g(x,y) that determine the next x and the next y respectively, have a certain structure. Specifically the partial derivativwe matrix _ _ |f_x f_y| |g_x g_y| ~ ~ may need to have the form _ _ |a -b| |b a| ~ ~ (I took apart f(x,y)=x^2-y^2 and g(x,y)=2xy and noted that the resulting matrix had that form. That is just complex z^2 interpreted as being from R^2 to R^2 instead of C to C.) This will always generate orthogonal families too. I tried coming up with pairs of functions f, g with this pattern; most tended to come out as some complex function or other. But I came up with this: Conform8J { c=p1, z=pixel: a=sinh(real(z))*cos(imag(z)) b=cosh(real(z))*sin(imag(z)) z=a+b*(0,1)+c, |z|<=1000 } Conform8M { c=pixel, z=(0,1)*3.141592654/2: a=sinh(real(z))*cos(imag(z)) b=cosh(real(z))*sin(imag(z)) z=a+b*(0,1)+c, |z|<=1000 } It generates some interesting Julias and Mandelbrots in R^2. Much weirder is: Conform7J { c=p1, z=pixel: a=exp(real(z))*sin(imag(z)) b=-exp(real(z))*cos(imag(z)) z=a+b*(0,1)+c, real(z)<=1000000 } Conform7M { c=pixel, z=0: a=exp(real(z))*sin(imag(z)) b=-exp(real(z))*cos(imag(z)) z=a+b*(0,1)+c, real(z)<=1000000 } This behaves strangely and is very slow. Both require float=y, should use periodicity=no for the latter as well. - -- .*. Where feelings are concerned, answers are rarely simple [GeneDeWeese] -() < When I go to the theater, I always go straight to the "bag and mix" `*' bulk candy section...because variety is the spice of life... [me] Paul Derbyshire ao950@freenet.carleton.ca, http://chat.carleton.ca/~pderbysh - ------------------------------------------------------------ Thanks for using Fractint, The Fractals and Fractint Discussion List Post Message: fractint@xmission.com Get Commands: majordomo@xmission.com "help" Administrator: twegner@phoenix.net Unsubscribe: majordomo@xmission.com "unsubscribe fractint" ------------------------------ Date: Sat, 11 Oct 1997 02:08:35 -0400 (EDT) From: Evin C Robertson Subject: Re: (fractint) Integer v. float **Warning**: Again, perhaps too technical for the casual Fractint user > Unfortunately Intel hasn't (to my knowledge) published basic cycle counts > for Pentium Pro and Pentium-II instructions. The last published counts I > have are for Pentiums. However, the trend from 386 through Pentium is for > FPU speed to increase drastically, and integer multiplies to increase only > moderately. (13-18 cycles on a 486, 10-11 for Pentium, for integer; 16 > cycles on a 486, 1 on a Pentium, for FPU.) I'd be very surprised if the > PPro and P-II had radically redesigned integer multiplies, given that > overall their performance is not leaps and bounds above Pentiums of the > same clock speed. Go to www.x86.org/intel.doc for links to lots of intel documentation. Great site. PPro, P-II and K6 are very different from their predecessors as they rewrite the assembly into RISC code (microinstructions), which they execute very quickly. > Actually, the problem FractInt has is that multiply instructions (IMUL and > MUL) are *not* pairable instructions on the Pentium, and so not only do > they consume 10-11 cycles each, but they *also* prevent other instructions > from running at the same time. On the PPro and P-II, the assignment of On the PPro, P-II, and K6, I believe an integer multiply takes 1-2 cycles when scheduled properly. > Various indications I have received are that AMD's K6 processor does not > *quite* have the FPU performance of their Intel counterparts, but they are > still much faster with FPU stuff than Cyrix processors. The K6 doesn't have a floating point pipeline, so if you have mulitple floating point instructions they can't all be running at the same time. I've heard that the K6 actually does single floating point instructions faster than the P-II. >Would you like to see who can write the fastest 32-bit >protected mode M-set generation routine? You write integer code, I'll >write FPU code. :) You're on. I'll be optimizing for my K6. I may well put some floating point stuff in there with my integers as the K6 does floating point and integer math simultaneously quite well. I guess I'll probably need to reoptimize for the P-II and P-pro to widen my target of computers a bit... If somebody would like to donate a fast P-II to me which I could optimize on... Quoting myself from above: < I may well put some floating point stuff in there with my integers... Perhaps this would be a good solution for fractint. Unfortunately, this doesn't work to well on some machines... >if it hinders the growth of FractInt in other areas. If I had to choose >between having synchronous orbits working fast, or integer math support in >a 32-bit environment, guess which one I'm gonna choose? (Good thing I >don't decide stuff, huh? :) I go for whatever goes fast. If integer math can go faster, let it go faster. I just have to ask: What's all this talk about synchronous orbits with fractals? Did an altavista search and came up with nothing related to the topic. - ------------------------------------------------------------ Thanks for using Fractint, The Fractals and Fractint Discussion List Post Message: fractint@xmission.com Get Commands: majordomo@xmission.com "help" Administrator: twegner@phoenix.net Unsubscribe: majordomo@xmission.com "unsubscribe fractint" ------------------------------ Date: Sat, 11 Oct 1997 03:32:23 -0400 (EDT) From: ao950@freenet.carleton.ca (Paul Derbyshire) Subject: (fractint) And now the PAR you've all been WAITING FOR! Sorry this is late. First the FRM then the PAR... ; Precogna.frm, formulas PrecognaJ, PrecognaM ; Copyright (C) 1997 PGD. ; Distribute it as you see fit, as long as these comments remain attached and ; the two formulas unaltered. Formulas you derive from these by changing ; anything, and images and par files generated with this, are yours to do ; with as you please. PrecognaJ { ; p1 is Julia parameter. z=pixel,p=0,c=p1: t=sqr(p)*p+sqr(z)+c p=z z=t, lastsqr<=2048 } PrecognaM (XAXIS) { ; p1 is Julia parameter. z=0,p=0,c=pixel: t=sqr(p)*p+sqr(z)+c p=z z=t, lastsqr<=2048 } ; Precogna.par, image parameters for beautiful images generated with the ; Precogna formulas. These formulas are found in precogna.frm ; which must be in your /frm directory. You should be able to obtain ; precogna.frm from where you obtained this file. ; If you obtained this file from usenet sci.fractals, the .frm was probably ; embedded in the same posting with this .par file. ; ; Copyright (C) 1997 PGD. ; Distribute it as you see fit, as long as these comments remain attached ; and the original entries unaltered. Images and PAR files you derive from ; these by changing anything, zooming in or out, or changing the formula, ; are yours to do with as you please. Black_Carrot { reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaJ center-mag=5.78e-006/-7.005e-006/0.6666636 params=-0.4/0 float=y maxiter=256 inside=0 colors=000000<28>w00z00z00<61>zy0zz0zz0zz0<157>zzz } Carrot_Detail { reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaJ center-mag=0.500791/0.313147/2.380941 params=-0.4/0 float=y maxiter=256 inside=0 colors=000000<28>w00z00z00<61>zy0zz0zz0zz0<157>zzz } Tractor_Beam { ; An alien starship towing a runabout in a tractor beam. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaJ center-mag=-0.751174/1.11022e-016/3.205128 params=-0.4/0.03 float=y maxiter=256 inside=0 logmap=5 colors=000<15>000000001112<43>TbxUczWdz<12>wxzzzzzzz<174>000 } Plasma_Reactor { ; Plasma storage tanks and reactor nexi. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaJ center-mag=0.346992/7.77156e-016/2.541735 params=-0.44/0 float=y maxiter=256 inside=0 logmap=5 colors=000000<45>xTbzUczWc<4>zjfzmgzni<5>zxwzzzzzz<30>d2zc0zc0z<29>LTzKU\ zKUzKUz<124>zzz } Pod_Detail { ; Detail of a plasma storage pod reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaJ center-mag=+0.34994753302034220/-0.06833724926931038/110.5102 params=-0.44/0 float=y maxiter=256 inside=0 logmap=5 colors=000000<45>bTxcUzcWz<4>fjzgmzinz<5>wxzzzzzzz<30>z2dz0cz0c<29>zTLzU\ KzUKzUK<124>zzz } Pod_Detail_2 { ; Detail of a plasma storage pod fuel siphon clamp reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaJ center-mag=+0.34904122331142200/-0.07381572826722271/394.6794 params=-0.44/0 float=y maxiter=256 inside=0 logmap=5 colors=000000<45>TbxUczWcz<4>jfzmgzniz<5>xwzzzzzzz<30>2zd0zc0zc<29>TzLUz\ KUzKUzK<124>zzz } Mandibles { ; A Precogna Mandelbrot image. Looks like bug jaws. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.424664/0/21.5839 float=y maxiter=256 inside=0 colors=000A0A<44>x0Ez0Fz0F<110>zzzzzzyyy<93>00A } Whirlpools { ; Whirlpools around a promontory. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.45579244166987560/+0.03772238872010915/88.4586 float=y maxiter=500 inside=0 logmap=21 colors=00015A<125>hlzhlzilzimzjmz<26>yyzzzzzzyzzx<12>zzezzczzc<78>A0A } Tongue { ; The bug's tongue reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.424664/1.30104e-017/68.30348 float=y maxiter=2000 inside=0 logmap=28 colors=000zzz<44>z2zz0zy0z<29>20z00z00y<29>00C00A00A<28>901901A00B00<29>\ x00z00z10<45>zx0zz0zz1<30>zzz } Tongue_Tip { ; The bug's tongue's tip reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.43731076613586700/+0.00000000000000002/216.1503 float=y maxiter=2000 inside=0 logmap=28 colors=000zzz<44>2zz0zz0yz<29>02z00z00y<29>00C00A00A<28>0910910A00B0<29>\ 0x00z01z0<45>xz0zz0zz1<30>zzz } Mandelbud { ; A minibrot tastebud. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM passes=b center-mag=-0.43873947536942330/-0.00135218707365365/4323.005 float=y maxiter=4096 logmap=66 colors=000000K0A<76>z00<16>zzV<14>zzz<30>2z20z00y0<29>0C00A00A0<31>00A00\ A10A10A20A20A<41>K0A } Intersection { ; Strands intersect in a superstring reactor. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.43379926315796940/+0.00000000000000002/1541.604 float=y maxiter=2000 inside=0 logmap=65 colors=000ezz<76>B1VA0UA0UA0U<46>U0AV0AW1B<42>xJTyJTzKUzLV<29>zxxzzzzzz<\ 46>ezz } Tantalon_Beam { ; Detail of the reactor core. reset=1930 type=formula formulafile=precogna.frm formulaname=precognam center-mag=-0.43379384907332540/+0.00000000000000002/11335.32 float=y maxiter=2000 inside=0 logmap=106 colors=000zez<76>1BV0AU0AU0AU<46>0UA0VA1WB<42>JxTJyTKzULzV<29>xzxzzzzzz<\ 46>zez } Precogna_Mandelbrt { ; The M-set of the Precogna formula. reset=1930 type=formula formulafile=precogna.frm formulaname=precognam center-mag=-9.99201e-016/7.77156e-016/1.449275 float=y maxiter=2000 inside=0 logmap=yes colors=000ezz<76>B1VA0UA0UA0U<46>U0AV0AW1B<42>xJTyJTzKUzLV<29>zxxzzzzzz<\ 46>ezz } Gullet { ; In the mandibles' grip. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.40822902177024260/+0.00000000000000000/431.678 float=y maxiter=2000 inside=0 logmap=75 colors=000zmz<60>B1zA0zA0yA0x<44>A02A00B10<45>xk0zm0zm0<94>zzz } Symbiont { ; A symbiotic crustacean lives in the mandibles. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.40969846075901170/+0.00000000000000000/3174.103 float=y maxiter=2000 inside=0 logmap=75 colors=000zmz<60>z1Bz0Ay0Ax0A<44>20A00A01B<45>0kx0mz0mz<94>zzz } Symbiont_Left { ; Left of the symbiont. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.41006067746974330/+0.00000000000000000/3174.103 float=y maxiter=5000 inside=0 logmap=75 colors=000zmz<60>z1Bz0Ay0Ax0A<44>20A00A01B<45>0kx0mz0mz<94>zzz } Spiral_Hook { ; On the bug's tongue. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.41006593651454530/+0.00005261789606087/36908.17 float=y maxiter=32767 inside=0 logmap=131 colors=000cmz<60>B1zA0zA0zA0y<44>A0LA0KA0KA0K<29>K0AK0AL1B<60>yyyzzzzzzz\ zz<45>zzz } Unicorn_Horn { ; Neat spirals and stuff. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.41004671470579420/+0.00006766398343948/271383.6 float=y maxiter=32767 inside=0 logmap=138 colors=000mzm<140>B1LA0KB1L<29>xxxzzzzzz<78>A05 } Convergence { ; Many converging lines reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.41054918859461490/+0.01354137030978280/431.678 float=y maxiter=2000 inside=0 logmap=112 colors=000zzz<108>dLBcKAcKAbKAbKA<76>00A00A00A00A109<25>901901A00B10<30>\ zz0 } Two_Strands { ; Two strands of mandelmatter from Convergence. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.41088754625650250/+0.01426680086209259/1541.707 float=y maxiter=4096 inside=0 logmap=127 colors=000zzz<92>znzzmzzmzzlyzkx<27>zLdzKczKcyKb<45>K00K00L10<77>zz0 } One_Strand { ; One of the Two_Strands. IMO the more interesting one. :) reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.41035970830395770/+0.01463241786045672/5506.097 float=y maxiter=4096 inside=0 logmap=133 colors=00000A<97>0BV0BV0BV0CW0CW0CW<53>0Tl0Um1Um2Vm<28>alycmzcmz<62>zz0 } Mandible_Edge { ; A serration on the mandible. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.41000324267340650/-0.00103920344720220/9321.416 float=y maxiter=4096 inside=0 logmap=84 colors=000czz<77>K0UK0UK1U<93>zzzzzzzyy<77>z00 } Verteron_Emitter { ; A strange alien device. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.41004129707625650/-0.00106025627565845/54194.28 float=y maxiter=4096 inside=0 logmap=89 colors=000zcU<77>U0KU0KU1K<93>nnznnzmmz<77>00z } GoldenFluff { ; reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.41003944897537460/-0.00107111953514188/444215.4 float=y maxiter=4096 inside=0 logmap=107 colors=000zzm<93>K00K00L10<76>yy0zz0zz0zz1<77>zzm } MandelApocalypse { ; Mandelbrot Set after the Apocalypse? reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.0250391/0.507307/3.436426 float=y maxiter=4096 inside=0 logmap=yes colors=000<2>000100100200200<87>c00c00c10c20<75>yy0zz0zz0zz1<77>zzm } Bud-oid { ; Mangled bud. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.171981/0.441695/14.08371 float=y maxiter=4096 inside=0 logmap=yes colors=000U00<142>00A01A11B<108>mzz } Spiroid { ; Spiral fusion reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.20368671048513370/+0.39277803757828890/72.59646 float=y maxiter=4096 inside=0 logmap=yes colors=000mzm<108>11B00A00A00A<31>A0AB1BB1BC2CC2C<105>zmz } Twisted_Matter { ; Twisted Mandelmatter. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM passes=t center-mag=-0.15721668519561880/+0.38557609741127420/17827.49 float=y maxiter=4096 inside=0 logmap=107 colors=000mmz<174>448448448447447447<26>000000100<44>K00 } Twisted_Matter_2 { ; A zoom of the twisted matter. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM passes=t center-mag=-0.15724514084310800/+0.38552973237442150/849575.2 float=y maxiter=4096 inside=0 logmap=112 colors=000KK0<13>hsfjvjjvj<157>448448448448447<28>000000100<44>K00 } Burning_Brot { ; Bright purple plasma fire burns a Mandelbrot bud. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.206057/0.491205/67.71016 float=y maxiter=4096 inside=0 logmap=19 colors=000zzU<108>881770770770<29>00A00A10A<109>zmz } Complexity { ; A complex region where chaos collides with chaos. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM passes=2 center-mag=-0.20975520050078320/+0.48269548719833070/159.6938 float=y maxiter=4096 inside=0 logmap=22 colors=00000A<141>z00z00z11<109>zzz } BrambleSpiral { ; A spiral of thorny protrusions. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.20243811866441390/+0.48865679211317000/3193.876 float=y maxiter=4096 inside=0 logmap=74 colors=00000A<76>b01c00c00c10<60>yy9zzAzzAzzA<109>zzz } Fracture_Zone { ; A fault line in Mandelbrot Land. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM passes=t center-mag=-0.20952000858461420/+0.47848596925408020/654.4827 float=y maxiter=4096 inside=0 logmap=32 colors=00000A<141>z00z00z11<109>zzz } Membranes { ; Complexity: a thin membrane between Brown and Purple. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM passes=t center-mag=-0.20855080883051680/+0.47805215359115470/2845.577 float=y maxiter=4096 inside=0 logmap=94 colors=000z0z<126>AKUAKUAKUAKUBKT<55>UKAUKATJA<21>B11A00B11<29>kxxmzznzz\ <8>zzz } Membrane_Segment { ; Part of the Membrane. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM passes=t center-mag=-0.20869013129516830/+0.47775869005446980/33088.1 float=y maxiter=32767 inside=0 logmap=106 colors=000z0z<127>AKUAKUBLU<123>zzz } Frost_Fronds { ; A beautiful, complex, and slow image. reset=1930 type=formula formulafile=precogna.frm formulaname=PrecognaM center-mag=-0.20866679844871980/+0.47777408515160420/661762.1 float=y maxiter=1048576 inside=0 logmap=1599 colors=000z0z<127>KAUKAULBU<123>zzz } Twistron { ; A very peculiar Julia set. reset=1960 type=formula formulafile=precogna.frm formulaname=PrecognaJ center-mag=0/0/0.6666667 params=-0.427/0.018 float=y maxiter=4096 inside=0 logmap=yes colors=000<11>733844944A55A55<88>0Uz0Uz0Uy<147>000 } Energon { ; Another very peculiar Julia set. reset=1960 type=formula formulafile=precogna.frm formulaname=PrecognaJ passes=t center-mag=0/0/0.6666667 params=-0.427/0.008 float=y maxiter=4096 inside=0 logmap=yes colors=000<11>33744844955A55A<88>Uz0Uz0Uy0<147>000 } Terduron { ; Another very peculiar Julia set. reset=1960 type=formula formulafile=precogna.frm formulaname=PrecognaJ passes=t center-mag=0/0/0.6666667 params=-0.427/0.002 float=y maxiter=4096 inside=0 logmap=yes colors=000<11>733844944A55A55<88>0zU0zU0yU<147>000 } Generon { ; Another very peculiar Julia set. reset=1960 type=formula formulafile=precogna.frm formulaname=PrecognaJ passes=t center-mag=0/0/0.6666667 params=-0.427/0.0005999999999999999 float=y maxiter=4096 inside=0 logmap=yes colors=000<11>33744844955A55A<88>zU0zU0yU0<147>000 } Vivid_Explosion { ; Very vivid. reset=1960 type=formula formulafile=precogna.frm formulaname=PrecognaJ passes=t center-mag=-8.88178e-016/6.66134e-016/1.103753 params=-0.45414/0.03511 float=y maxiter=4096 inside=0 logmap=yes colors=000<14>40950A50A60A<92>yJ1zK0zK0zL0<28>zb0zc0zc0zc1<108>zzz } Twistor-Drive { ; This looks very weird. Periodicity checking screws it up reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-8.88178e-016/6.10623e-016/1.041667 params=-0.4079/0.0026 float=y maxiter=4096 inside=0 logmap=yes periodicity=0 colors=000<12>830840940A50A50<91>1xJ1xJ1yJ0zK0zK0zL<28>0zb0zc0zc1zc<60>y\ zyzzzyyy<45>00A } Pinched_Carrot { ; Strange. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-8.88178e-016/6.10623e-016/1.041667 params=-0.40518/0 float=y maxiter=4096 inside=0 logmap=yes periodicity=0 colors=000yzz<11>rwzrvzqvzpuzpuz<91>y2gy2gy1gz0fz0fz0e<28>z0Oz0Nz0Ny0N<6\ 0>101000111<45>zzp } Glowing_Asteroids { ; Aftermath of a space collision. From a minibrot. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-8.88178e-016/6.10623e-016/1.041667 params=-0.45255/0.00338 float=y maxiter=4096 inside=0 logmap=yes periodicity=0 colors=000<12>30840840950A50A<91>xJ1xJ1yJ1zK0zK0zL0<28>zb0zc0zc0zc1<57>z\ xvzxwzyxyzyzzz<46>A00 } Deimos { ; Amid the asteroids. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=0/0/20.83333 params=-0.45255/0.00338 float=y maxiter=4096 inside=0 logmap=yes colors=000<12>30840840950A50A<91>xJ1xJ1yJ1zK0zK0zL0<28>zb0zc0zc0zc1<57>z\ xvzxwzyxyzyzzz<46>A00 } Phobos { ; Amid the asteroids, slight param tweak. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-9.57567e-016/6.10623e-016/10.41667 params=-0.446477/0.002736 float=y maxiter=4096 inside=0 logmap=28 colors=000<12>30840840950A50A<94>z0Kz1Kz2J<28>za1zc0zc0<58>zxvzxwzyxyzyz\ zz<46>A00 } Spiral_Night { ; A hole in the space time continuum. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=0/0/0.6666667 params=-0.40746/0.00143 float=y maxiter=16384 inside=0 logmap=20 periodicity=0 colors=000005<11>30840940950A50A<78>z0Kz1Kz2J<28>za1zc0zc1<45>zyxzzzzyy<\ 45>IAAH88H88<29>000 } Spiral_Spiral_S { ; Very interesting and unusual. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=-8.88178e-016/5.55112e-016/1.041667 params=-0.411579/0.013829 float=y maxiter=4096 inside=0 logmap=yes periodicity=0 colors=000<10>702703803804904<96>K0zK1zJ2z<28>1az0cz0cz<61>yyzzzzyyy<45>\ 00A } Intricacy { ; Very complex patterns of spirals! ; Very slow... reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=2 center-mag=-8.88178e-016/5.55112e-016/1.041667 params=-0.4115331/0.0138375 float=y maxiter=32767 inside=0 logmap=yes periodicity=0 colors=000<10>702703803804904<96>K0zK1zJ2z<28>1az0cz0cz<61>yyzzzzyyy<45>\ 00A } Triplicacy { ; Three-fold forms reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-8.88178e-016/5.55112e-016/1.041667 params=-0.139238/0.40686 float=y maxiter=4096 inside=0 logmap=yes colors=000<8>026026027037038048<97>0zK1zK2zJ<28>az1cz0cz0<61>yzyzzzyyy<4\ 5>0A0 } Alien_Starship { ; Very odd. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-8.88178e-016/5.55112e-016/1.041667 params=-0.438768/0 float=y maxiter=4096 inside=0 logmap=yes colors=000<16>A00B10B10C20C31<88>xxJyyJzzKzzL<29>zzxzzzyyz<45>C2LA0KA0K<\ 61>A00 } Starship_Mark_II { ; A revised model. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-8.88178e-016/5.55112e-016/1.041667 params=-0.436084/0 float=y maxiter=4096 inside=0 logmap=yes colors=000<16>00A01B01B02C13C<88>JxxJyyKzzLzz<29>xzzzzzzyy<45>L2CK0AK0A<\ 61>00A } Mark_II_Reactor { ; The Mark II craft has a powerful reactor core. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-9.57567e-016/6.10623e-016/7.659316 params=-0.436084/0 float=y maxiter=4096 inside=0 logmap=yes colors=000<16>00A01B01B02C13C<88>JxxJyyKzzLzz<29>xzzzzzzyy<45>L2CK0AK0A<\ 61>00A } Fluxion_Chamber { ; In the Mark II's reactor core. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-9.81853e-016/6.17562e-016/44.53091 params=-0.436084/0 float=y maxiter=4096 inside=0 logmap=92 colors=000<16>00A10B10B20C31C<88>xJxyJyzKzzLz<29>zxzzzzzyy<45>LC2KA0KA0<\ 61>0A0 } Starship_Mark_III { ; Another revised model. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-8.88178e-016/5.55112e-016/1.041667 params=-0.433787/0 float=y maxiter=4096 inside=0 logmap=yes colors=000<16>0A01B01B02C03C1<87>wwJxxJyyJzzKzzL<29>zzxzzzzyy<43>N5EM3DL\ 2CK0AK0A<61>00A } Mark_III_Reactor { ; Different... reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-9.71445e-016/6.10623e-016/10.41667 params=-0.433787/0 float=y maxiter=4096 inside=0 logmap=yes colors=000<16>0A01B01B02C03C1<87>wwJxxJyyJzzKzzL<29>zzxzzzzyy<43>N5EM3DL\ 2CK0AK0A<61>00A } Luminon_Chamber { ; The center of the Mark III reactor uses fractal geometry reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-0.00000000000000098/+0.00000000000000062/60.56203 params=-0.433787/0 float=y maxiter=4096 inside=0 logmap=103 colors=000<16>A00B01B01C02C13<87>wJwxJxyJyzKzzLz<29>zxzzzzyyz<43>5EN3DM2\ CL0AK0AK<61>0A0 } Crystal_Lumps { ; A blue corona surrounds crystals that formed in the core reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=-1.00614e-015/6.38378e-016/32.96415 params=-0.4339/0 float=y maxiter=4096 inside=0 logmap=69 colors=000<15>A00A00B11B22<75>yyyzzzyyy<29>22C00A00B<37>JJlKKmKMn<6>Kcz<\ 78>zzz } Crystal_Resonance { ; One crystal in emerald green, looks like a Tholian ship. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=2 center-mag=-0.01954057947918072/+0.00000000000000064/135.099 params=-0.4339/0 float=y maxiter=16384 inside=0 logmap=105 colors=000<15>A0AA0AB1BB2B<75>yyyzzzyyy<29>2C20A00B0<37>JlJKmKKnM<5>Kx`K\ zcKzc<77>zzz } Golden_Spiral { ; A beautiful delicate spiral of gold wrapped in mystery. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=+0.01123836285323686/-0.00283608552263435/659.283 params=-0.43382/0 float=y maxiter=4096 inside=0 logmap=299 periodicity=0 colors=0000A0<110>000001012<44>9TxAUzATw<14>A0A<46>xx1zz0zz2<29>zzz } Deep_Core { ; Deep in the core of the Darkness is the Light reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-8.88178e-016/6.10623e-016/1.041667 params=-0.136275/0.397373 float=y maxiter=4096 inside=0 logmap=yes colors=000A00<44>18909A09A09B09B<91>8Sw9Tx9Su<17>124000110<57>vv0xx1ww1<\ 30>A00 } Complex_Stuff { ; reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=2 center-mag=-8.88178e-016/6.10623e-016/1.041667 params=-0.411435/0.013704 float=y maxiter=4096 inside=0 logmap=yes periodicity=0 colors=000A00<44>18909A09A09B09B<91>8Sw9Tx9Su<17>124000110<57>vv0xx1ww1<\ 30>A00 } Warm_Cool_Spirals { ; Very odd reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=-8.88178e-016/6.10623e-016/1.041667 params=-0.410899/0.014262 float=y maxiter=4096 inside=0 logmap=yes periodicity=0 colors=000zzz<29>z33z00z00<108>L09L09K0AJ0A<13>201000000<59>0av0bw1cx2cx\ <30>zzz } Swirl_Tunnels { ; Strange reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-8.88178e-016/6.10623e-016/1.041667 params=-0.157247/0.385551 float=y maxiter=4096 inside=0 logmap=yes periodicity=0 colors=000zzz<29>33z00z00z<108>90L90LA0KA0J<13>102000000<59>v0aw0bx1cx2c\ <30>zzz } Hypertronic_Pulse { ; This formula produces a lot of funky alien power cores.. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=t center-mag=-9.57567e-016/6.10623e-016/16.27604 params=-0.157247/0.385551 float=y maxiter=4096 inside=0 logmap=104 periodicity=0 colors=000<14>90IA0KA0K<140>TlyTlyUmzVmz<15>xynzzmzkp<2>z0z<72>A00 } Galaxies { ; Galaxies in the night reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=0/0/0.6666667 params=-0.210613/0.391426 float=y maxiter=4096 inside=0 logmap=3 periodicity=0 colors=000<14>90IA0KA0K<140>TlyTlyUmzVmz<15>xynzzmzkp<2>z0z<72>A00 } Tango { ; Spiraling things do an intimate tango. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=-6.66134e-016/4.44089e-016/0.9861933 params=-0.192851/0.394992 float=y maxiter=4096 inside=0 logmap=3 periodicity=0 colors=000<14>90IA0KA0K<140>TlyTlyUmzVmz<15>xynzzmzkp<2>z0z<72>A00 } Chitoid { ; This complex chitinous creature inhabits Alpha Centauri. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=2 center-mag=-8.88178e-016/0.0677411/1.152095 params=-0.189172/0.390449 float=y maxiter=4096 inside=0 logmap=3 periodicity=0 colors=000<14>90IA0KA0K<140>TlyTlyUmzVmz<15>xynzzmzkp<2>z0z<72>A00 } Chitoid_Zoom { ; This complex chitinous creature inhabits Alpha Centauri. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=2 center-mag=-1.05471e-015/0.0278755/4.114625 params=-0.189172/0.390449 float=y maxiter=4096 inside=0 logmap=3 periodicity=0 colors=000<14>90IA0KA0K<140>TlyTlyUmzVmz<15>xynzzmzkp<2>z0z<72>A00 } Green_Goo { ; Strange stuff reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=2 center-mag=-8.88178e-016/0.0677411/1.152095 params=-0.157234/0.385567 float=y maxiter=4096 inside=0 logmap=3 colors=000zyy<11>zrjzrizqhzpfzpfzpf<139>EY1EY1DX0DW0<15>12C00D00D<75>000 } Purple_Haze { ; Strange stuff reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=2 center-mag=7.77156e-016/0.0083288/0.8895923 params=0.209132/0.011115 float=y maxiter=4096 inside=0 logmap=3 colors=000yyz<11>rjzrizqhzpfzpfzpfz<139>Y1EY1EX0DW0D<15>2C10D00D0<75>000 } Fuzzy_Blue { ; Strange stuff reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj passes=2 center-mag=7.77156e-016/0.0083288/0.8895923 params=0.209735/0.010512 float=y maxiter=4096 inside=0 logmap=3 colors=000yyz<13>hhzffzffz<139>Z2FZ2FY1EX1E<29>21B00A00A<62>000 } Frondy_Feathers { ; Attractive fluffy spirally stuff reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=7.77156e-016/0.0083288/0.8895923 params=0.210294/0.009349 float=y maxiter=4096 inside=0 logmap=3 colors=000yyz<13>hhzffzffz<139>Z2FZ2FY1EX1E<29>21B00A00A<62>000 } Blue_Green_Algae { ; Interesting critters in pond water. reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=7.77156e-016/0.0083288/0.8895923 params=-0.203006/0.488152 float=y maxiter=4096 inside=0 logmap=3 colors=000yyz<11>jykiyihyhfzffzf<140>2FZ1EY1EY1EX1DW<27>11B00A00A00A<61>\ 000 } Pastel_Spirals { ; reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=-1.11022e-015/6.66134e-016/1.131455 params=-0.202296/0.488633 float=y maxiter=4096 inside=0 logmap=3 colors=000yyz<11>yykyyiyyhzzfzzf<123>L7`L7`K6_K6Z<13>F2ZE1YE1YE1XD1W<12>\ 71M71M82N<77>zzz } Pastel_Double { ; Double spiral reset=1960 type=formula formulafile=precogna.frm formulaname=precognaj center-mag=-1.16573e-015/-0.00369026/17.67899/1/97.5 params=-0.202296/0.488633 float=y maxiter=4096 inside=0 logmap=3 colors=000yyz<11>yykyyiyyhzzfzzf<123>L7`L7`K6_K6Z<13>F2ZE1YE1YE1XD1W<12>\ 71M71M82N<77>zzz } - -- .*. Where feelings are concerned, answers are rarely simple [GeneDeWeese] -() < When I go to the theater, I always go straight to the "bag and mix" `*' bulk candy section...because variety is the spice of life... [me] Paul Derbyshire ao950@freenet.carleton.ca, http://chat.carleton.ca/~pderbysh - ------------------------------------------------------------ Thanks for using Fractint, The Fractals and Fractint Discussion List Post Message: fractint@xmission.com Get Commands: majordomo@xmission.com "help" Administrator: twegner@phoenix.net Unsubscribe: majordomo@xmission.com "unsubscribe fractint" ------------------------------ End of fractint Digest V1 #28 ***************************** To subscribe to fractint Digest, send the command: subscribe fractint-digest in the body of a message to "majordomo@xmission.com". If you want to subscribe something other than the account the mail is coming from, such as a local redistribution list, then append that address to the "subscribe" command; for example, to subscribe "local-fractint": subscribe fractint-digest local-fractint@your.domain.net A non-digest (direct mail) version of this list is also available; to subscribe to that instead, replace all instances of "fractint-digest" in the commands above with "fractint". Back issues are available for anonymous FTP from ftp.xmission.com, in pub/lists/fractint/archive. These are organized by date.