Tglad Dihedral Pseudo-Klien Newton Triplet cardiod bulb-ish thing

  • 6 Replies
  • 342 Views

0 Members and 1 Guest are viewing this topic.

Offline gforce

  • *
  • Fractal Friend
  • **
  • Posts: 11
« on: July 21, 2018, 05:27:38 PM »
Seems to show spinal features everywhere, seems to have M-Set like properties(also has what looks like higher power M-Sets at the origin!), (if you want to turn on julia you need to swap _abox for the other one) Changing the fake bailout on the newton triplet formula gives different shapes. Changing the log2map power gives a more filled surface but looks worse to me.  I really wouldn't mess with the DE stop at all.

To really see the detail in this fractal you need to zoom in while roataing.(preferably in helices) This is because the detail in this fractal happens as the surface "expands", Still renderings while nice become a little dull once you see it in motion.   

I will make an image thread here to post some more renders when i find nice spots to be in, for now you have to make do with some admittedly terrible renderings(they are all on outside only) of a VERY cool fractal.(if you look on the right side at the beginning of the purple and orange zoom you can see the Mandelbrot valley shapes, its in the other two gifs but its more subtle)  :yes: Also if somebody with more mathematical and programming experience could sum this up into one formula that would be awesome.  :peace:   

ps: in the future I will post my gifs on an external site :)


Offline gforce

  • *
  • Fractal Friend
  • **
  • Posts: 11
« Reply #1 on: July 21, 2018, 05:29:13 PM »
Update: these params do show some kind of pow2-3-4ish mset!(just change logmap2power on the tetrahedra), if anybody does open this up you can zoom around arbitrarily on the outside and see nice spirals and stuff. Looks the the cubic flying squirrel or buffalo brot. You can also rotate the brot by inverting the pi3 value on the first set of numbers in the mobius triplex, looks awesome animated.

Mandelbulb3Dv18{
g....M6...UV....61....2..sNJ4c1XyB/zzm/Mr6hAT15Es0/4wxxyjz96qxFcD66Ezw1eR3/laE/E
................................dx4w3aAJZz1........Y./x3................y.2...wD
...Uz6....U/.....M.0/......A/...wM....E3.....sHSDL6BsVtD/.........sC1/......1LDm
z.UaNaVD16../.kw9GzJumLwzog1IZo1Y9zj5k6Ca6urEz9...........U0.....y1...sD...../..
.zXaNadDqtTk1G1SMy1..........iXA4akc9EgjRjWrzMxOkrfRy5wUoU5azgGiLTkc9EgjfYvp5Au0
2vnCmMM0Xi.lysuUo46BsVtDUoAnAt1B2.............sD.E....sD..G.0...................
.............oAnAt1...sD....zw1.........................................V0...k1.
.....CnAnz1.......kz.wzzz1.U..6......o0...EB....O....U2...EE....0/...I1.....SJrH
SU1U.0aZczDM6/nzMg2czX6dE1660c..yfwe...........U6.....................U.8.UQgk0.
.wUmc2beYz1.dA8E5Exwz0........../EU0.wzzz1...........s/...................E.2c..
zzzz.............0...................2./8.kzzzD............8....................
/EU0.wzzz1...................................UTVR..y3q/yzz/k.1A..1Ak.vzDsnwF.UDn
5VTzThdht0kaqaPiQs5y3q/.sLM5s1bTsLM5.UTVRUDlyVTVR..y3q/yMw5y3q/.sLM5snqTsLM5.UTV
RUDkzVTVR..y3q/y..kzfcDbzzlzfcDkyznzfczmxzpzfcjf................................
E....ME/F3E.....I....U....kLI7LMiBbNj7LP..................................E/3IU.
...........4hE2Jv5G0./..........................................................
..............zD................................................................
.....................o5....3....A....wJHj7KOpB5JkV5.............................
.....................................................6R6nw5TN90EG5mAzlLqW.A.....
...wz...........vAIpNlccn/2..........6R6nw5TN90kG5mAzlLqW.2........wz...........
................................P....E/...U.....T3IPVdLOiRaEjV5.dtqN............
............5...........................kz9.....................................
................................................................................
..........................................E.....I.........kLalKOkBI9sZbS........
...................................................6./..........................
................................................................................
.....................................................E.....3....C....cIGIxpNidqL
GJKMg/pPrtINrFrPix3Ao.........................................zDMo0FIhT67.2.....
...wzMYtI1JeWaxj..........UFaHBId8OqzgPjLbxrwhRDSVHJdcbOj.2.....................
...................../..........................................n1...E/...U.....
IR4PVF4FdVKNY/............................U.0...................................
...../..........................................................................
............................................................................}
{Titel: MSET}

« Last Edit: July 21, 2018, 11:31:36 PM by gforce, Reason: cool stuff »

Offline Fraktalist

  • *
  • Administrator
  • *******
  • Strange Attractor
  • Posts: 916
« Reply #2 on: July 22, 2018, 02:34:44 PM »
very interesting!!
Looking forward to see some more details.

when I try to render these parameters it takes ages, even in the preview navigator, so I'll be patient and wait for more from you ;)

Offline gforce

  • *
  • Fractal Friend
  • **
  • Posts: 11
« Reply #3 on: July 27, 2018, 04:04:59 AM »
yeah it does take forever  :(, i'm just finishing up a multi-day render.
  Also i found that hybrid formulas of JIT_gnj_RealPowNewton_04 and tglad tetra or dihedral rendering make m-sets close to other ones i've seen before.(they all seem to be perpendicular, no nice mset at z=0) They still have that flat or stretched look that i don't want but they are much faster I'd recommend trying it out your self.  (in my formula i guess i'm kind of trying to expand space with the mobius triplex abox combo((psuedo-klein)) , probably not a good way to word that sorry) ??? also the best renders i have made have the values of tetrafoldingpower3 default fold value(divided by some value 1/2 or 1/4) as the inversion values of x and z in the newton formulas. (one being positive and one negative)1 iteration of realpowernewton to 3 of tglads formula seems to work the best.

  I've also had mild success in doing that with different m-sets 1 iteration of hopf4d or tetrafoldingpow3 then 3 of another then 9 ect... Works best leaving bhuddibrot formula as the last one. Ends up has a result similar to burning bridge but with the boundary of that first fractal persevered, still looks wrong though. (and you have to have a lot of iterations)

Am i incorrect in saying that mandelbulb 3-d will not make nice DE's for hybrid formulas? I think this can and should be sped up even if just for the pretty pictures.

« Last Edit: July 27, 2018, 04:30:10 AM by gforce »

Offline mclarekin

  • *
  • Fractal Furball
  • ***
  • Posts: 299
« Reply #4 on: July 28, 2018, 03:19:32 AM »
Quote
Am i incorrect in saying that mandelbulb 3-d will not make nice DE's for hybrid formulas?
Generally the DE problem is when hybridizing  log and linear type formulas

I have not used M3D for a while, but seem to remember that there is not many options for control over DE calculations when making hybrids. 
With M3D being written in machine language I cannot check it out.With Fragmentrarium and Mandelbulberv2, you can see the DE Analytic calcs, which are generally of two types, linear or logarithmic

These are the basic Final Calculations that are performed on the DE value at termination of the iteration loop.

basic log (bulbs)         return   0.5 * r * log(r) / DE;

basic linear                   return    r / DE;

(but they can be much more complicated.)  So when making a hybrid between linear and log type formulas what Final Calculation is used ? well that depends on a lot of things, but often log.

In the formula the DE is updated every iteration. If in a linear formula the DE is a simple updated DE = DE * (a number),  then it  can often "fit"  with a log formula in a hybrid.

It gets more difficult if DE = DE * (a number e.g scale )  + 1.0.

Sometimes you want to "tune" the DE in a hybrid,     so to provide some  flexibility  we can use  DE = DE * (a number e.g scale ) * DEtweakScale + DEtweakOffset;


Offline gforce

  • *
  • Fractal Friend
  • **
  • Posts: 11
« Reply #5 on: July 29, 2018, 09:17:31 AM »
Thank you, :embarrass: I'm pretty sure that is becoming a problem.(not too bad though only the really detailed parts are being over/under sampled). I think it is just a relatively computational heavy formula. 
« Last Edit: August 01, 2018, 03:16:46 AM by gforce »

Offline gannjondal

  • *
  • Fractal Friend
  • **
  • Posts: 14
« Reply #6 on: August 26, 2018, 11:23:11 PM »
Nice pictures indeed, gforce  8)
Just two comments:
.... when I try to render these parameters it takes ages, even in the preview navigator ....
There are some glitches around (well, one could call it bugs?) that can make some formulas (often JIT, but not exclusively) slower as needed. 

Your params (i have taken the MSET one as example) show one typical example: 
It seams that if the Z-end plane is too large MB3D appears to search something in the background (again, only for some formulas).In below params I have just lowered the 'Z end' param from about 270 to 7.
Well, it's not really fast, but it is about twice as fast as the original:
Code: [Select]
Mandelbulb3Dv18{
g....UZ....K0...61....2..stt0eIwfR6zz0.......k/EJqtYfJyyjzvDGcigFd4EzgFWNU9m7E/E
................................dx4w3aAJZz1........Y./x3................y.2...wD
...Uz6....k/.....M.0/.....kX....wM....E3.....MZdILT/QLrD/.......IBSI2/......6GDs
./UaNaVD16../.kw9GzJumLwzog1IZo1Y9zj5k6Ca6urEz9...........U0.....y1...sD...../..
.zXaNadDaWnznL.rpx1..........mCcEDmEc5ejvNHonfhJBrfdswzw3kRRzEe2E7mEc5ejZ8/IWA2u
VuHvU0x61VScyQjmd6T/QLrDUoAnAtHl/.............sD.E....sD..G.....................
.............oAnAt1...sD....zw1.........................................Y0...k1.
.....83iyz1.......kz.wzzz1.U..6......o0...EB....O....U2...EE....0/...I1.....SJrH
SU1U.0aZczDM6/nzMg2czX6dE1660c..yfwe...........U6.....................U.8.UQgk0.
.wUmc2beYz1.dA8E5Exwz0........../EU0.wzzz1...........s/...................E.2c..
zzzz.............0...................2./8.kzzzD............8....................
/EU0.wzzz1...................................UTVR..y3q/yzz/k.1A..1Ak.vzDsnwF.UDn
5VTzThdht0kaqaPiQs5y3q/.sLM5s1bTsLM5.UTVRUDlyVTVR..y3q/yMw5y3q/.sLM5snqTsLM5.UTV
RUDkzVTVR..y3q/y..kzfcDbzzlzfcDkyznzfczmxzpzfcjf................................
E....ME/F3E.....I....U....kLI7LMiBbNj7LP..................................E/3IU.
...........4hE2Jv5G0./..........................................................
..............zD................................................................
.....................o5....3....A....wJHj7KOpB5JkV5.............................
.....................................................6R6nw5TN90EG5mAzlLqW.A.....
...wz...........vAIpNlccn/2..........6R6nw5TN90kG5mAzlLqW.2........wz...........
................................P....E/...U.....T3IPVdLOiRaEjV5.dtqN............
............5...........................kz9.....................................
................................................................................
..........................................E.....I.........kLalKOkBI9sZbS........
...................................................6./..........................
................................................................................
.....................................................E.....3....C....cIGIxpNidqL
GJKMg/pPrtINrFrPix3Ao.........................................zDMo0FIhT67.2.....
...wzMYtI1JeWaxj..........UFaHBId8OqzgPjLbxrwhRDSVHJdcbOj.2.....................
...................../..........................................n1...E/...U.....
IR4PVF4FdVKNY/............................U.0...................................
...../..........................................................................
............................................................................}
{Titel: MSET-faster}
I don'treally believe that it goes much faster, mainly due to the usage of this quite edgy combination of -mostly not too fast- formulas with often many iterations for each.

... Am i incorrect in saying that mandelbulb 3-d will not make nice DE's for hybrid formulas? ...
Good question. 
What I know is that the only way to control the used DE mechanism in MB3D for each single formula is to place it on a ceartain formula tab using the .DEoption parameter. 
I also don't know which one is taken into account if you mix for instance formulas from the 3D tab with formulas from the 3Da tab. Probably you will need to do research in the old Fractalforums.com pages  ;)
The only experiment you could do would be to create a copy of a formula m3f , and to change the value of .DEoption parameter (you can find the correct values on the existing formulas).



clip
Pseudo-Klien-Tglad-Newton hyrbrids

Started by gforce on Fractal Image Gallery

0 Replies
44 Views
Last post August 01, 2018, 03:15:16 AM
by gforce
wink
Gforce's M-set like Newton-Tglad(Diheadral)-PsuedoKlein hybrid

Started by gforce on Image Threads

1 Replies
90 Views
Last post July 29, 2018, 06:46:17 AM
by gforce
xx
The Newton Bulb - Some fruits need to ripe long

Started by gannjondal on Fractal Image Gallery

0 Replies
64 Views
Last post March 13, 2018, 12:00:58 AM
by gannjondal
xx
pseudo kleinian

Started by Adam Majewski on Programming

3 Replies
426 Views
Last post September 05, 2017, 06:36:35 PM
by knighty
clip
Pearls - Pseudo Kleinian 3D Fractal

Started by Alon Nahary on Fractal movie gallery

3 Replies
61 Views
Last post August 20, 2018, 08:54:44 PM
by Alon Nahary