as far as creating a basic synch,say with COSM

you need noise..either with words or specific sounds.

then you need pictora moments that connect to same moment or moments of either a bunch of words or a sound.

WHAT YOU DONT WANT is anything you dont want. playback that say nonsense or untruth or ngative crap should be rejected (unless)

so what is relly required is a series of one or more correlating things.

the goal being to get as many as possible or at least consistantly, without distasteful errors. the counterintuitive secret of prgramming is that its not the positive matching data thats importan, even thats what we want. the real issue is negative connatations and combinations. so instead of making MASSIVE EXTINSIVE DTABASES of ALL SOTS OF POSITIVE DATA (whic has already been done in the movie for you(

all one really needs to do is scan NEGATIVE DATA and the remainder is ALWAYS an articulate synch.

thats jsust ONE REASON im the master. ive never met another person who has understood this withour my personal instruction and demonstration

and yet i myself have proven the (law) therum (methodolgy) with the simplest programs.

$ = (A+V)2 - SCB(PCB)

(this applize only to str8 line thru the field) The Total amount of synchable startponts for any 2 sorces = the entire volume of possibilities, but reject thos that violate either a standard or personal confirmation bias, or RULEBASE. Complete rulebase makes non-idiosynchratic results as all interactions can be wexplained by a rule or pressident

FOR InSTANCE say you name is MIKE. and so your watching a movie with a character named MIKE so you want to get all the good playback out of the combo without hurting the name MIKE either in picture or in caption.

so there is a list of words that when combined with MIKE = error. Lets say the lyrics are i dunno

"mental wounds not healing who and whats to blame? im going off the rails on the crazy train"

ANd YOU are MIKe, right, so maybe you say OK crazy + mike is ok, and Mike + healing, and 'who' is good, but not NOT so

basically there would be a mike filter aand some + and - words and id feed that into the synchomatic and then a list spits out taht show you WOW you can play RIGHT HERE nd everythings good but 2 seconds earlier or later and thers dozens of errors.

this system is nice because also tells you the exact AMOUNT OF TIME that a line stays in an error state.

The fun shti that is hard and time consuming is

programming the computer to combine error free segmentws of errored line using keywaords to move around the errors.

Also i would do these statistics to individual tracks rather than str8 lists in order in order to combine error free list reorders but thats not how i confirm them because all our lists are reusable. but that wold be fun just for picking stuff at random but i havent evolved my decoder to that lever because i can already see everything with the naked eye.