I'd like to report a bug in the algorithm. I opted to double the pattern (3/4 to the minibrot) from the original formation located at 1219 zoom levels. First I zoomed about 20 or so zoom levels into the point where I wanted to zoom into, to insure the correct path was chosen. After 5 passes and some hours later, it honed in on a position at 1519 zoom levels. That's only a quarter of the path to the minibrot, not half! It should have placed me at the target formation around 1800 or so.
So I am redoing it. Lesson learned, this time I will tell it to locate the minibrot, then manually back out to a distance halfway between the minibrot and the original location. Still it is useful even if unoptomized and non-multitreaded, because I don't have to baby it all the way to the target formation. Even Mandel Machine is taking over a minute just to calculate the reference and Series approximation.
Edit: Kalles fractaler found the minibrot. I awoke to find my PC burning with KF at 3% on a black image some e600+ deep, 225 million iteration depth. I saved the file and copied the coordinates into notepad for use with Mandel Machine. The minibrot is not always located exactly twice the zoom depth of the forked path, and can occasionally be at a shallower depth if the Julia formation zoomed into is sufficiently complex. I can take the depth of coordinates using log function on calculator to determine the exact formation at equidistant zoom depth to the fork point and minibrot, expressed as log base 2 or "zooms". Load coordinates in Mandel Machine to frame and render the object. This is certainly more convenient to manual advance as I am not tethered to the PC for hours on end.EDIT2: Well working out the exact location was easier than I expected. I saved the KF file and loaded it in Mandel Machine. The image was solid black with bailout of 225 million. I took the minibrot location (2027 zooms) and the fork location (1219 zooms) and calculated the point exactly halfway between those two (1623 zooms). This placed me smack dab in the center of the target formation!
As it turns out, the 1519 figure was "wrong" because I forked into a highly detailed Julia formation. The Minibrot (2027) was much closer than the expected 2438, because of this fact. So 3/4 to the minibrot is not always accurate. The period duplicity is always equidistant to the fork formation and the final minibrot.I should come up with something to show a bit later. Peace...