Thanks for reply, yeah i knew about all those threads you posted, dont worry im not that dumb or lazy and wasting your time 
. But i read them only partially they are huge and maybe something changed since they were written... I will focus mainly on that multicam thread with 35 pages tommorow.
I understand, it's a lot of material and things have changed over time... but it's what's available. So far no one has publicly documented one of these projects in more digestible format. I would certainly welcome anyone posting more detailed write-ups!
Sometimes skimming backwards through threads is more productive than starting at the beginning.
Whats currently the sync delay while running the chdkptp with maximum of 20 cameras? How much delay can i expect in ms? Does someone know...? I think that for my purposes with "only" 20 cameras the delay could be manageable... what is it?
Do you mean the delay between when you command shot and when the shot starts, or spread between the times each cameras shutter opens?
The way the chdkptp multicam script work is:
1) the init_sync function attempts to estimate offsets between each cameras 10ms precision tick counter and "real time"
2) When you shoot, a command is sent to each camera telling it to shoot at a particular time in the future, as a tick count calculated from the offsets in #1
So the delay between sending the command and the cameras shooting scales linearly with the number of cameras, due to the time required to actually send the commands. For reasonably modern cameras like yoursm 5ms per camera should be a decent ballpark, plus a few tens of ms in fixed overhead. (This is ignoring the half-press you need to do beforehand to get all the cameras ready to shoot.)
The spread between each cameras shots depends on the quality of the sync estimates, the precision limitations of the counter, and uncertainties introduced by the cameras task scheduling and shooting code. It shouldn't really depend on the number of cameras, except that with more cameras you will cover more of the overall distribution in any given shot.
I haven't measured this rigorously, I only have a few cameras (all of different models) and I haven't set up a precision test rig. From the limited testing that I have done and what I know about the cameras, I'd expect it to be ~50ms or less most of the time, probably closer to 30. However, this is little more than a WAG and there could easily be camera specific variations.
