ftp.xmission.com/2014.06.ftp.xmission.com.tar/ftp.xmission.com/pub/lists/fractint/archive/v01.n126

Internet Message Format | 1998-03-08 | 42KB

From: [e-mail] (fractint-digest)
To: [e-mail]
Subject: fractint-digest V1 #126
Reply-To: fractint-digest
Sender: [e-mail]
Errors-To: [e-mail]
Precedence: bulk


fractint-digest         Monday, March 9 1998         Volume 01 : Number 126




----------------------------------------------------------------------

Date: 08 Mar 1998 19:31:08 +0100
From: Brederlow <[e-mail]>
Subject: Re: (fractint) fractal hardware engine

[e-mail] (Paul Derbyshire) writes:

> Won't the serial port I/O speed become a serious bottleneck? It must be
> orders of magnitude slower than a PC bus. I wouldn't think all the
> megahertz in the world would make much difference if your FPU communicated to
> the CPU/memory at only 57600 baud...

Considering that one pixel in 19 decimals accuracy at 32000 iterations
takes about 1 second on my maschine and that you can communicate at up
to 496000 BAUD over a normal Amiga Serial, thats 62000 Pixel a second
or roughly one 320x200 Picture in one second.

No matter how slow the serial ist, if the conected hardware is fast
enough and complex enough, its a speedup. Of cause its worthless if it
can't do at least n iterations with m dezimals accuracy by itself.

May the Source be with you.
            Mrvn


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: 08 Mar 1998 19:48:47 +0100
From: Brederlow <[e-mail]>
Subject: Re: (fractint) Arbitrary Precision pisses me off!

[e-mail] writes:

> Sorry bout the subject, but it does. I was generating =9Fberzoomed M-Se=
ts at
> 320x200x256, c. 2,000,000,000 iterations, and suddenly it stopped and n=
o matter
> how long I waited it wouldn't draw another point. I tried again later, =
and at
> 17-decimal arbitrary precision mode, it stopped at the same place. I wa=
ited 24
> hours, and it was still at the same point, with the time around 25:32:0=
0. Is
> there something wrong with my computer, (P133), with Windows 95. It doe=
s the
> same thing in DOS mode too.

Maybe you just hit the M-Set. At 17-decimals and 2,000,000,000 a Point =

in the M-Set will take rather long.

Considering that 2**16 (65536) iterations take about 1 second on your
comp and you are trying to do 2**31 iterations, which will take 32768
seconds a point in the M-Set. 32768 s =3D 546 min =3D 9 hours.

After those 9 hours you should get ONE pixel further.
With 3 Pixels a day (well, lets round up) you would need 21333
days. Lets say half your image is in the M-Set, which gives 10666 days =

=3D 30 years.

I hope you see that you eigther have to decrease the itterations or
get a far bigger maschine. With 17 decimals theres no need for that
many itterations anyway.

May the Source be with you.
            Mrvn

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: 08 Mar 1998 19:48:47 +0100
From: Brederlow <[e-mail]>
Subject: Re: (fractint) Arbitrary Precision pisses me off!

[e-mail] writes:

> Sorry bout the subject, but it does. I was generating =9Fberzoomed M-Se=
ts at
> 320x200x256, c. 2,000,000,000 iterations, and suddenly it stopped and n=
o matter
> how long I waited it wouldn't draw another point. I tried again later, =
and at
> 17-decimal arbitrary precision mode, it stopped at the same place. I wa=
ited 24
> hours, and it was still at the same point, with the time around 25:32:0=
0. Is
> there something wrong with my computer, (P133), with Windows 95. It doe=
s the
> same thing in DOS mode too.

Maybe you just hit the M-Set. At 17-decimals and 2,000,000,000 a Point =

in the M-Set will take rather long.

Considering that 2**16 (65536) iterations take about 1 second on your
comp and you are trying to do 2**31 iterations, which will take 32768
seconds a point in the M-Set. 32768 s =3D 546 min =3D 9 hours.

After those 9 hours you should get ONE pixel further.
With 3 Pixels a day (well, lets round up) you would need 21333
days. Lets say half your image is in the M-Set, which gives 10666 days =

=3D 30 years.

I hope you see that you eigther have to decrease the itterations or
get a far bigger maschine. With 17 decimals theres no need for that
many itterations anyway.

May the Source be with you.
            Mrvn

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: 08 Mar 1998 19:56:33 +0100
From: Brederlow <[e-mail]>
Subject: Re: (fractint) Arbitrary Precision pisses me off!

[e-mail] (Paul Derbyshire) writes:

> >Sorry bout the subject, but it does. I was generating =9Fberzoomed M-S=
ets at
> >320x200x256, c. 2,000,000,000 iterations, and suddenly it stopped and =
no matter
> >how long I waited it wouldn't draw another point. I tried again later,=
 and at
> >17-decimal arbitrary precision mode, it stopped at the same place. I w=
aited 24
> >hours, and it was still at the same point, with the time around 25:32:=
00. Is
> >there something wrong with my computer, (P133), with Windows 95. It do=
es the
> >same thing in DOS mode too.
> =

> I had this happen also. I reported it, and was told it might be a bug i=
n
> the GIF encoder's fault that can cause a restored half-done image to ha=
ng.
> However, after I pointed out that it happened to a fresh image (not loa=
ded
> as a GIF but started fresh), there has been no response. I suspect a
> rare(?) arbitrary precision bug causing the occasional hang, unrelated =
to
> GIF encoding/decoding.

When a gif is loaded back in xfractint only the lower 16 Bit of the
itteration is restored. When the Image had 65536 itterations set, it
will have 0 itterations after loading and then crash. Its awullfy
buggy.

May the Source be with you.
            Mrvn

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 08 Mar 1998 11:12:03 -0800
From: Wizzle <[e-mail]>
Subject: (fractint) documentation

I've been reading the documentation for fractint (!!!!!!!!!)
Goodness.....it's wonderful!!

I must say I still don't understand some of it....but it's much more
meaningful than when I started using Fractint a year ago....documentation
for a graphics program is about as hard a problem as there is, I think.
Thank you Stone Soup Group.

I noticed in my reading a new-ish standard formula....voltera_lotta or
something like that. Really neat!! First try outta the box yielded this image

wizvoltera_lotta   { ; wizzle 3/7/98 w. creams4.map
                     ; something different
  reset=1960 type=volterra-lotka
  center-mag=+1.04755944931163900/+1.38606010016694400/0.9661836/1/-90
  params=0.739/0.5 float=y colors=@creams4.map
  }

The color map is a really good one to re-do in other colors.  It is what I
think of as a "blacks" map.....yup....lots of black here and there.  It was
part of my "cream and gold" period....I'm now in my "grey" period. It
cycles nicely too....so do try that....there is an amber-ish point, even.

Speaking of amber....Pete

Did you get anything close to what you were visualizing for bugs in amber?

Angela

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: 08 Mar 1998 20:07:39 +0100
From: Brederlow <[e-mail]>
Subject: Re: (fractint) Re: A Geom Construction of the M-set

Jason Hine <[e-mail]> writes:

> To all interested,
>
> Ok, I've got a few pictures to help with my description of how to find the
> tangent points for all the bulbs off the main cardioid.  I'm not sure where to
> take it from here... I can see how the buds off the period 2 bulb might be
> described, but what about period 3?  How does one describe the line which
> represents the 'x-axis' for the period 3 bulb?  I'd check out relative areas of
> bulbs, but I think this has already been done - anyone have two cents to add?
>
> Anyway, For a mediocre improvement of my last couple emails on this subject,
> check out:
>
> http://boralf.agsci.colostate.edu/~jason/fracpages/geo_mset.htm

And how can one determine the radius of the bulbs? Witout that it
would mean nothing to now the tangent points.

May the Source be with you.
            Mrvn


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 08 Mar 1998 11:10:59 -0800
From: Mark Christenson <[e-mail]>
Subject: Re: (fractint) Viewwindows / correction

At 09:45 AM 3/8/98 -0800, I wrote:
>Oddly, though, askvideo behaves
>differently in 19.3 and 19.6.  In 19.3, it does what one would like,
>resets the viewwindow without complications, but in 19.6 an image
>restored from gif is displayed in the current video mode (as opposed
>to the mode it was saved in).

This applies to restores from GIF, in both cases.  Of course, askvideo
has no effect (good or bad) on restores from PAR.

>For those having trouble with losing slices of images when returning
>from option screens or editing the palette...

I think I meant "loading a new palette", which is just another option
screen.


Bud


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 8 Mar 1998 13:18:47 -0600
From: "Tim Wegner" <[e-mail]>
Subject: (fractint) How to use simplgif (long)

Wizzle asked:

> 1. what is simplegif's purpose?

The main purpose of simplgif is to convert a multiple image GIF file
to a single image GIF. A multiple image GIF is a single file, but
within that file are separate images. The reason you might want to
do this is if you wanted to make a super resolution image.

Suppose you have just created the world's most dazzling fractal, and
you want to create a wall-sized version of it for your exhibition at
the Metropolitan Museum.

You might do the following steps. (In real life you can use a higher
resolution - this is just a low resolution example for you to
follow).

a. Generate a relatively low resolution version of your fractal, say
320x200 (F3)

b. press <b> to take you to the screen that generates PAR files (The
"Save Current Parameters" screen.

For safety, enter the name of a brand new PAR file. At the bottom,
enter integral values for "X Multiples" and "Y Multiples". If you
used 2 and 2, this would result in a final image of 640x400.

After you press <enter>, exit Fractint, and in addition to a PAR file
you will see the file makemig.bat in the current directory. The
contents of this file is:

Fractint batch=yes overwrite=yes @fractint.par/test_00
If Errorlevel 2 goto oops
Fractint batch=yes overwrite=yes @fractint.par/test_01
If Errorlevel 2 goto oops
Fractint batch=yes overwrite=yes @fractint.par/test_10
If Errorlevel 2 goto oops
Fractint batch=yes overwrite=yes @fractint.par/test_11
If Errorlevel 2 goto oops
Fractint makemig=2/2
Rem Simplgif fractmig.gif simplgif.gif  in case you need it
:oops

The first lines of this batch file generate four 320x200 images that
each contain 1/4 of your total image.

The line

Fractint makemig=2/2

combines these images into a  multiple-image GIF file called
fractmig.gif. Unfortunately most service bureaus who print images
have never heard of a multiple image GIF. (You can view it in
Fractint though - you will see the images added the the total
image one piece at a time - try it.) This is where simplgif comes
in. The last line of the batch file, which is commented out, invokes
simplgif. If you uncommented it, it would look like:

Simplgif fractmig.gif simplgif.gif

This simply converts the multiple image GIF file fractmig.gif (the
file with fractint creates using that name) to a single image GIF
called simplgif.gif (or any other name you used as the second
parameter).

This example is silly because it only results in a 640x400 image. The
real excitement happens when you use a video mode of
something like 1600x1200 and, say, multiples of 4x4. This would
result in a final image of 6400x4800, which is a seriously high
resolution. You might want to go higher.

The earlier simplgif released with version 19.6 had various problems
with some images. The new simplgif is rock solid and capable of up to
32000x32000 pixels IF you have enough disk space for the temporary
file. For that huge size the temporary file would be about a
gigabyte, which is no longer out of the question with today's disk
drives. A more practical limitation is the maximum image size a
service bureau who prints your image can take. Up to 16K pixels isn't
too hard to find, and an image that big looks good blown up to wall
sized. Jon Noring and others have made images for exhibitions that
big. Maybe Lee Skinner can tell us his experience.

I hope this helps. Try the simplfied example, then when you have lots
of free disk space and time. try a larger resolution.

Tim Wegner




 (A  , which
> 2. what are the steps for use?
> 3. any warnings?
>
> Angela
>
> At 09:37 PM 3/4/98 -0600, you wrote:
> >A while ago I mentioned that a new version of simplgif is available
> >at:
> >
> >ftp://ftp.phoenix.net/pub/USERS/twegner/simplgif.zip
> >
> >This utility combines the images in multiple-image GIFs created with
> >Fractint's "divide-and-conquer" mode.
> >
> >Just curious, is it bug free or have folks just not gotten around to
> >trying it? I'll tell you a secret if you promise not to tell. If
> >you add the -i option to the simplgif command line before the
> >filename arguments, it will produce an interlaced file. But no,
> >Fractint doesn't read interlaced files correctly yet <g!>
> >
> >Tim
>
>
> -
> ------------------------------------------------------------
> Thanks for using Fractint, The Fractals and Fractint Discussion List
> Post Message:   [e-mail]
> Get Commands:   [e-mail] "help"
> Administrator:  [e-mail]
> Unsubscribe:    [e-mail] "unsubscribe fractint"
>
>

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 8 Mar 1998 13:31:03
From: "Lavondyss" <[e-mail]>
Subject: RE: (fractint) documentation

Wizzle

Yeah, it does help if one reads the manuals. I haven't read all of the
fractal stuff. Especially about writing your own formula files. Hehehe. I
don't know how to program. But I assume it's fairly difficult. I'm a chemist
mind you. Computer related subjects are just my hobby as well as graphics.

Speaking of which, I might be getting a newer, faster computer soon. I'm
still running at P100Mhz, 2.1G, 64MB RAM. What is everyone else running at?
I've gotta admit, my computer does run smoother than some new ones. It
doesn't crash as often and doesn't freeze as often either. I brought the
Fractint program to work and loaded it onto a p233 or higher and those
perform slightly better than the one I have now. I thought they would
perform astoundingly. Is that due to the program or the computer?


- -Lavondyss




- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 08 Mar 1998 15:44:33 -0500
From: Theran Cochran <[e-mail]>
Subject: (fractint) DJGPP

I'm pretty sure that someone has asked this before, but I'll try
anyways.

Will Fractint compile with DJGPP? I want to try my hand at adding PNG
support, but Fractint is a real mode program and DJGPP makes 32 bit
protected mode exe's. The C part would probably compile OK, but I don't
know about the assembler. It would be nice to know before going through
the trouble of downloading, unpacking, and rewriting the makefiles.

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 08 Mar 1998 17:34:51 -0800
From: Peter Jakubowicz <[e-mail]>
Subject: Re: (fractint) How to use simplgif (long)

Tim,

>The main purpose of simplgif is to convert a multiple image GIF file
>to a single image GIF.

I have a bunch of new Fractint images I want to generate at 1600 x 1200 and
then reduce to 640 x 480 (for anti-aliasing) for my web page. So if I make
a mulipart GIF with simplegif will I then be able to load that into Picture
Publisher whence I can reduce their size?

Also, if this can be done, I am wondering how long this tends to take to do
per image (say, on a Pentium/166). If they do not take too long, I wld be
interested in checking whether either the DJGPP-compiled or the
MSVC++-compiled executable is faster.


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 08 Mar 1998 16:50:49 -0600
From: "Damien M. Jones" <[e-mail]>
Subject: Re: (fractint) How to use simplgif (long)

Peter,

 - I have a bunch of new Fractint images I want to generate at 1600 x 1200 and
 - then reduce to 640 x 480 (for anti-aliasing) for my web page.

You don't need SIMPLGIF for that.  I do this sort of thing all the time
(although I don't use Picture Publisher to resize).

 - So if I make a mulipart GIF with simplegif will I then be able to load that
 - into Picture Publisher whence I can reduce their size?

SIMPLGIF doesn't make multi-part GIFs, it takes multi-part GIFs built with
FractInt's divide & conquer process and creates a *simple* GIF, one that
any program with GIF support will read.  I have used this process before,
and most programs do not support the multi-part GIF properly.  SIMPLGIF is
essential.

 - Also, if this can be done, I am wondering how long this tends to take to do
 - per image (say, on a Pentium/166).

It depends very much on the image.  I have images which render at 1600x1200
in a matter of minutes on my P166; I have others which take hours.

Damien M. Jones   \\
[e-mail]  \\  http://www.icd.com/tsd/ (temporary sanity designs)
                    \\  http://www.fractalus.com/ (fractals are my hobby)

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 08 Mar 1998 19:17:11 -0800
From: Peter Jakubowicz <[e-mail]>
Subject: (fractint) banana bug par

Opus #44           { ; The Bug Series (Banana Bug in Amber)
                     ; Not to be confused w/ Fractal Excursions, or w/ Bach's
                     ; sixth suite for unaccompanied cello
  reset=1960 type=formula formulafile=*.frm
  formulaname=tri-fn+fn function=cosxx/log
  center-mag=-0.168051/0.00386238/0.6915625/1/90 params=16/4/0.5/0
  float=y maxiter=1024 outside=summ decomp=256 periodicity=0
  colors=000T90<199>zz0000<37>000R50<13>T80
  }


frm:tri-fn+fn    { ; Sylvie Gallet <[e-mail]>, May 1997
               ; Thanks to Kerry Mitchell for the idea!
               ; real(p1) = bailout
               ; imag(p1) must be different from 0 (>=1 recommended)
               ; z=fn1(z) + p2*fn2(z)
   ; This formula must be used with decomp = 256 and periodicity = 0
   ;
   z1 = pixel , mz1 = cabs(z1) , summ = iter = 0
   b1 = abs(real(p1)) , f = b1^0.1
   b2 = b1*f , b3 = b2*f , b4 = b3*f , b5 = b4*f , b6 = b5*f
   b7 = b6*f , b8 = b7*f , b9 = b8*f , b10 = b9*f , b11 = b10*f
   iter1 = iter2 = iter3 = iter4 = iter5 = iter6 = iter7       \
         = iter8 = iter9 = iter10 = summ_tot = iter_tot = 0
   k = imag(p1) * (0.0,6.28319530718)
   :
   iter = iter + 1 , za = fn1(z1) , a = cabs(za)
   zb = p2*fn2(z1) , b = cabs(zb)
   z1 = za + zb , mz1 = cabs(z1) , a_b = abs(a - b)
   summ = summ + (mz1 - a_b) / (a+b - a_b)
   IF (mz1 > b1)
    IF (iter1==0)
     iter1 = iter
     iter_tot = iter_tot + iter1 , summ_tot = summ_tot + summ
    ENDIF
    IF (mz1 > b2)
     IF (iter2==0)
      iter2 = iter
      iter_tot = iter_tot + iter2 , summ_tot = summ_tot + summ
     ENDIF
     IF (mz1 > b3)
      IF (iter3==0)
       iter3 = iter
       iter_tot = iter_tot + iter3 , summ_tot = summ_tot + summ
      ENDIF
      IF (mz1 > b4)
       IF (iter4==0)
        iter4 = iter
        iter_tot = iter_tot + iter4 , summ_tot = summ_tot + summ
       ENDIF
       IF (mz1 > b5)
        IF (iter5==0)
         iter5 = iter
         iter_tot = iter_tot + iter5 , summ_tot = summ_tot + summ
        ENDIF
        IF (mz1 > b6)
         IF (iter6==0)
          iter6 = iter
          iter_tot = iter_tot + iter6 , summ_tot = summ_tot + summ
         ENDIF
         IF (mz1 > b7)
          IF (iter7==0)
           iter7 = iter
           iter_tot = iter_tot + iter7 , summ_tot = summ_tot + summ
          ENDIF
          IF (mz1 > b8)
           IF (iter8==0)
            iter8 = iter
            iter_tot = iter_tot + iter8 , summ_tot = summ_tot + summ
           ENDIF
           IF (mz1 > b9)
            IF (iter9==0)
             iter9 = iter
             iter_tot = iter_tot + iter9 , summ_tot = summ_tot + summ
            ENDIF
            IF (mz1 > b10)
             IF (iter10==0)
              iter10 = iter
              iter_tot = iter_tot + iter10 , summ_tot = summ_tot + summ
             ENDIF
             IF (mz1 > b11)
              iter_tot = iter_tot + iter , summ_tot = summ_tot + summ
              z = exp (summ_tot * k / iter_tot)
             ENDIF
            ENDIF
           ENDIF
          ENDIF
         ENDIF
        ENDIF
       ENDIF
      ENDIF
     ENDIF
    ENDIF
   ENDIF
   mz1 <= b11
   }



- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 8 Mar 1998 19:26:33 -0500
From: "Peter Gavin" <[e-mail]>
Subject: (fractint) Bug report, I think?...

I got this wierd as heck looking picture using the params given in this
par...  It must be a bug.  If not, it's the wierdest looking julia set I've
EVER seen!!!  Really, you MUST see this.  Plug it in, and tell me what
happens.  I won't tell you now, but I will once you email me back.  I want
to find out if it's just my computer, or something.

I insist you try it out!!!
Pete

WhatTheHell Julia  {
  reset=1960 type=julia passes=1
  corners=-1.0128777921781637e-008/-1.0115340421781637e-008/2.379957685677\
  4754e-008/2.3809654981774755e-008
  params=-0.5059243436387786/0.6932148034001859 float=y maxiter=65536
  inside=0 logmap=1114 symmetry=none periodicity=4
  colors=000U0U<12>303000000<14>zzzzzz<14>GGG<13>333222000000<13>0ET0GW0GW\
  <13>023000000<14>kW0kW0<14>000000<13>T00W00W00<13>w0uz0zz0z<14>zzzzzz<14\
  >000<60>U0UU0UV0VW0W
  }




- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 8 Mar 1998 16:55:20 -0800
From: "Jay Hill" <[e-mail]>
Subject: Re: (fractint) Bug report, I think?...

Hi Pete,

The problem is that

"Floating-point flag is activated"

and it should be arbitrary at this point. Even zooming in more
does not activate it though.

I went to the MSet (hit space bar) and tried another Julia.
It did switch to arbitrary precision after zooming to about
the depth of yours.

Jay

- ----------
> From: Peter Gavin <[e-mail]>
> To: FractInt Mailing List <[e-mail]>
> Subject: (fractint) Bug report, I think?...
> Date: Sunday, March 08, 1998 4:26 PM
>
> I got this wierd as heck looking picture using the params given in this
> par...  It must be a bug.  If not, it's the wierdest looking julia set I've
> EVER seen!!!  Really, you MUST see this.  Plug it in, and tell me what
> happens.  I won't tell you now, but I will once you email me back.  I want
> to find out if it's just my computer, or something.
>
> I insist you try it out!!!
> Pete
>
> WhatTheHell Julia  {
>   reset=1960 type=julia passes=1
>   corners=-1.0128777921781637e-008/-1.0115340421781637e-008/2.379957685677\
>   4754e-008/2.3809654981774755e-008
>   params=-0.5059243436387786/0.6932148034001859 float=y maxiter=65536
>   inside=0 logmap=1114 symmetry=none periodicity=4
>   colors=000U0U<12>303000000<14>zzzzzz<14>GGG<13>333222000000<13>0ET0GW0GW\
>   <13>023000000<14>kW0kW0<14>000000<13>T00W00W00<13>w0uz0zz0z<14>zzzzzz<14\
>   >000<60>U0UU0UV0VW0W
>   }


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 8 Mar 1998 20:02:47 -0500
From: Lee Skinner <[e-mail]>
Subject: (fractint) documentation

HI Wiz,

>>wizvoltera_lotta   { ; wizzle 3/7/98 w. creams4.map
                     ; something different
  reset=3D1960 type=3Dvolterra-lotka
  center-mag=3D+1.04755944931163900/+1.38606010016694400/0.9661836/1/-90
  params=3D0.739/0.5 float=3Dy colors=[e-mail]
  }

The color map is a really good one to re-do in other colors.  It is what =
I
think of as a "blacks" map.....yup....lots of black here and there.  It w=
as
part of my "cream and gold" period....I'm now in my "grey" period. It
cycles nicely too....so do try that....there is an amber-ish point, even.=
<<

For those of us that don't have creams4.map, could you repost this par wi=
th
all the colors, please?  Thanks,

Lee

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 8 Mar 1998 20:16:30 -0500
From: "Peter Gavin" <[e-mail]>
Subject: Re: (fractint) Bug report, I think?...

Ignore this post, unless you want a good laff... : )

Pete

- -----Original Message-----
From: Jay Hill <[e-mail]>
To: [e-mail] <[e-mail]>
Date: Sunday, March 08, 1998 7:53 PM
Subject: Re: (fractint) Bug report, I think?...


>Hi Pete,
>
>The problem is that
>
>"Floating-point flag is activated"
>
>and it should be arbitrary at this point. Even zooming in more
>does not activate it though.
>
>I went to the MSet (hit space bar) and tried another Julia.
>It did switch to arbitrary precision after zooming to about
>the depth of yours.
>
>Jay
>
>----------
>> From: Peter Gavin <[e-mail]>
>> To: FractInt Mailing List <[e-mail]>
>> Subject: (fractint) Bug report, I think?...
>> Date: Sunday, March 08, 1998 4:26 PM
>>
>> I got this wierd as heck looking picture using the params given in this
>> par...  It must be a bug.  If not, it's the wierdest looking julia set
I've
>> EVER seen!!!  Really, you MUST see this.  Plug it in, and tell me what
>> happens.  I won't tell you now, but I will once you email me back.  I
want
>> to find out if it's just my computer, or something.
>>
>> I insist you try it out!!!
>> Pete
>>
>> WhatTheHell Julia  {
>>   reset=1960 type=julia passes=1
>>
corners=-1.0128777921781637e-008/-1.0115340421781637e-008/2.379957685677\
>>   4754e-008/2.3809654981774755e-008
>>   params=-0.5059243436387786/0.6932148034001859 float=y maxiter=65536
>>   inside=0 logmap=1114 symmetry=none periodicity=4
>>
colors=000U0U<12>303000000<14>zzzzzz<14>GGG<13>333222000000<13>0ET0GW0GW\
>>
<13>023000000<14>kW0kW0<14>000000<13>T00W00W00<13>w0uz0zz0z<14>zzzzzz<14\
>>   >000<60>U0UU0UV0VW0W
>>   }
>
>
>-
>------------------------------------------------------------
>Thanks for using Fractint, The Fractals and Fractint Discussion List
>Post Message:   [e-mail]
>Get Commands:   [e-mail] "help"
>Administrator:  [e-mail]
>Unsubscribe:    [e-mail] "unsubscribe fractint"


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 8 Mar 1998 20:46:55 EST
From: Genealogy1 <[e-mail]>
Subject: (fractint) A Julia variant ?

A Julia variant? Takes about 20 minutes on a Pentium 166 at a resolution of
1600 X 1200. Looks best at higher resolutions.

- --Bob Carr--

Carr3371(YAXIS){;Modified Sylvie Gallet frm.1996
;passes=1 needs to be used with this PHC formula
pixel=-abs(real(pixel))+flip(imag(pixel))
b6=conj(0.1/pixel),b7=flip(0.01/pixel),b8=cabs(0.002/pixel+cabs(0.002/pixel))
b4=abs(conj(conj(pixel^2)))*conj(flip(pixel^3))-conj(0.1/pixel-
flip(0.01/pixel))
b5=conj(conj(pixel^3))*cabs(pixel+0.125)^4-b7-b6+1.75*b8-0.2
c=whitesq*b4-(whitesq==0)*b4
z=whitesq*b5^0.7456-(whitesq==0)*b5
c1=1.5*z^1.2,c2=2.25*z,c3=3.375*z,c4=5.0625*z
l1=real(p1),l2=imag(p1),l3=real(p2),l4=imag(p2)
bailout=16,iter=0:
t1=(iter==l1),t2=(iter==l2),t3=(iter==l3),t4=(iter==l4)
t=1-(t1||t2||t3||t4),z=z*t,c=c*t+c1*t1+c2*t2+c3*t3+c4*t4
z=z*z+(-0.7456,-0.13214)+conj(c/10)
iter=iter+1
(conj(|z|))<=bailout
}

Carr3371          { ; Julia variant ?
                         ; Copyright (c) 1998 Robert W. Carr, M.D.
  reset=1960 type=formula formulafile=43gCarr.frm formulaname=Carr3371
  passes=1 center-mag=0.0010622/-0.00591256/0.8687799
  params=200/300/400/550 float=y maxiter=647 periodicity=0
  colors=000<3>000C00<9>Z00`00`45<8>Xcr<9>`00<11>0000FM<12>zz0<10>`00<7>zy\
  n<8>yUF<7>x4R<7>S6TO7TK7TG7UC5M83F427<14>VEtXFwZIv<14>zyn000<13>Xcr<15>0\
  00232<10>5Pw<6>zo`<5>5Pw<6>000CC9<14>zyn<8>XFw<4>XVsXYsX`rXcrV`o<13>16F
  }





- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 8 Mar 1998 19:27:40 -0800
From: "Jay Hill" <[e-mail]>
Subject: Re: (fractint) Arbitrary Precision

> From: Brederlow <[e-mail]>
> [e-mail] writes:
>=20
> > Sorry bout the subject, but it does. I was generating =9Fberzoomed M-=
Sets at
> > 320x200x256, c. 2,000,000,000 iterations, and suddenly it stopped and=
 no
matter
> > how long I waited it wouldn't draw another point. I tried again later=
, and
at
> > 17-decimal arbitrary precision mode, it stopped at the same place. I =
waited
24
> > hours, and it was still at the same point, with the time around 25:32=
:00.
Is

Where is this point. You should post the par.  To be hung up this long wi=
thout=20
being a bug would require the point being VERY accurately on the edge of =
the
MSet. These points are quasichaotic, never converge nor diverge.=20

> > there something wrong with my computer, (P133), with Windows 95. It d=
oes
the
> > same thing in DOS mode too.
>=20
> Maybe you just hit the M-Set. At 17-decimals and 2,000,000,000 a Point=20
> in the M-Set will take rather long.
>=20

We should be getting more than a million iterations per second.  Over on =
the=20
development lists they are bragging about 10 cycle iteration times. So wi=
th=20
a P100 that would be 10,000,000 iterations/sec or 200 sec per point at=20
2,000,000,000 iterations using floating point.  If you failed to back off
enough,
you might still be in arbitrary precision math.=20

Has anyone benchmarked iterations per second for arbitrary precision math
just after you leave double?

>=20
> I hope you see that you either have to decrease the iterations or
> get a far bigger machine. With 17 decimals there is no need for that
> many iterations anyway.

If you are using period detection, the limit of 2,000,000,000 would occur
very rarely. The amount of time spent above what looks good and what
hits the limit is normally small.

Jay



- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Sun, 8 Mar 1998 22:43:06 -0500
From: "Brian E. Jones" <[e-mail]>
Subject: Re: (fractint) A Julia variant ?

Hi Doc,

>>A Julia variant?<<

Cool image, Doc!

Brian
- -
http://ourworld.compuserve.com/homepages/Brian_E_Jones
- -
[e-mail]
- -
[e-mail]



- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: Mon, 09 Mar 1998 01:39:15 -0600
From: "Justin A. Kolodziej" <[e-mail]>
Subject: Re: (fractint) Bug report, I think?...

Peter Gavin wrote:
>
> I got this wierd as heck looking picture using the params given in this
> par...  It must be a bug.  If not, it's the wierdest looking julia set I've
> EVER seen!!!  Really, you MUST see this.  Plug it in, and tell me what
> happens.  I won't tell you now, but I will once you email me back.  I want
> to find out if it's just my computer, or something.
>
> I insist you try it out!!!
> Pete

Yep, looks like we got a major problem here.  Even on Linux and
XFractint.  Something's definitely the wrong precision.

Regarding the arbitrary precision, it doesn't make the switch until a
"mag" on the order of 10^21, where it suddenly jumps to 26 decimals!!!
Then the strange thing is that when you zoom back out, it stays at
arbitrary until you get down to below 17 decimals, the pops back into
floating point.

There is good news, however.  There is a way to control where the
switchovers occur between floating-point and arbitrary precision.
Unfortunately, I couldn't find it in the XFractint help files.  Perhaps
someone else would like to help out???

Justin K.

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

Date: 09 Mar 1998 13:37:22 +0100
From: Brederlow <[e-mail]>
Subject: Re: (fractint) Arbitrary Precision

Jay Hill <[e-mail]> writes:

> > From: Brederlow <[e-mail]>
> > [e-mail] writes:
> > =

> > > Sorry bout the subject, but it does. I was generating =9Fberzoomed =
M-Sets at
> > > 320x200x256, c. 2,000,000,000 iterations, and suddenly it stopped a=
nd no
> matter
> > > how long I waited it wouldn't draw another point. I tried again lat=
er, and
> at
> > > 17-decimal arbitrary precision mode, it stopped at the same place. =
I waited
> 24
> > > hours, and it was still at the same point, with the time around 25:=
32:00.
> Is
> =

> Where is this point. You should post the par.  To be hung up this long =
without =

> being a bug would require the point being VERY accurately on the edge o=
f the
> MSet. These points are quasichaotic, never converge nor diverge. =

> =

> > > there something wrong with my computer, (P133), with Windows 95. It=
 does
> the
> > > same thing in DOS mode too.
> > =

> > Maybe you just hit the M-Set. At 17-decimals and 2,000,000,000 a Poin=
t =

> > in the M-Set will take rather long.
> > =

> =

> We should be getting more than a million iterations per second.  Over o=
n the =

> development lists they are bragging about 10 cycle iteration times. So =
with =

> a P100 that would be 10,000,000 iterations/sec or 200 sec per point at =

> 2,000,000,000 iterations using floating point.  If you failed to back o=
ff
> enough,
> you might still be in arbitrary precision math. =

> =

> Has anyone benchmarked iterations per second for arbitrary precision ma=
th
> just after you leave double?

I think 2**16 iterations per second is a good guess for arbitrary
precision with 17 digits, which would make it 9 hours a pixel and
about 3 pixels a day at 2**31 iterations.

> > I hope you see that you either have to decrease the iterations or
> > get a far bigger machine. With 17 decimals there is no need for that
> > many iterations anyway.
> =

> If you are using period detection, the limit of 2,000,000,000 would occ=
ur
> very rarely. The amount of time spent above what looks good and what
> hits the limit is normally small.

Murphys Law: It will occur!

And then you have to wait 9 hours.

How good is the period detection with 17 decimals precision? Points on =

the edge of the M-Set will converge slowly to the fix points and they
will most likely have a high period. Say you have a period of 50, then =

the period seen by the periodicy check could be >1000 because of
rounding errors. Also it could take millions of iterations until the
periodicy becomes clear.

Remember, some points converge real slowly towards their periodic
points and that will be the points you are zooming into, because there =

lie the intresting parts.


What would realy help here, would be SOI (Simultanious Orbital
Itteration). It does a bunch of iterations for 9 control points of a
rectangle and then divides it into 4 smaler rectangles doing another
bunch of iterations. The speedup of this method is poor for the
complete M-Set and the first zooms, but the deep one goes, the better
it gets. Its speedup grows with zoom and not decreases like the
methods fractint knows currently.

May the Source be with you.
            Mrvn

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

------------------------------

End of fractint-digest V1 #126
******************************