Hmm that's an interesting thought, I'll experiment with it.
The following would require developer input/comment but could be of interest to you.
In recent testing of precision sync (on SX150IS) it transpires that it does not work well with either cameraside script or ptp connection still active - the camera crashes in up to 50% of cases. This is probably not surprising given the tight loops involved with that method.
In any case, it seems that (for a single camera), assuming ptp comms has already been locked and the switch wire then set to 0V (say by a usbhub per port 5V switch), by subsequently:
- setting the remote / twopush / sync enable (via multicam.lua)
- stopping the cameraside script and then closing the ptp connection (via multicam.lua)
- pulsing the 5V per port usb hub switch twice
a shot is reliably taken without camera crashes (famous last words).
So, if the inbuilt chdk twopush switching action were re-coded to operate from the opposite polarity to that which currently exists then triggering of all the cameras from a single mains power switch supplying all the hubs might look like this:
- lock ptp comms (once only via multicam.lua)
- set the remote / twopush / sync enable (once only via multicam.lua)
- stop the camera side script and then close the ptp connection (via multicam.lua)
- unplug the usb hub from the laptop/PC (or at least break 5V connection from PC)
- turn the usb hub mains power off (to prefocus) then on
- turn the usb hub mains power off (for lcd black screen) then on for sync shot (but see further below).
- reconnect the usb hub to PC
- start ptp connection (via multicam.lua)
- start cameraside script (via multicam.lua)
- download image / set zoom etc
- repeat.
(Edit: at the end of the session disable remote via multicam.lua)
Issues:
using two cameras and two equivalent but unbranded externally powered 10 port usbhubs to act as a simple usb switch using mains power as the trigger (with no ptp connection & normal usb remote voltage level operation), decimated precision sync in a 10 stereo pair test, that crt tester result was:
Minimum mis-sync = 8.86ms
Maximum mis-sync = 29.65ms
Average mis-sync = 17.47
If that was no worse over a large number of cameras & usb hubs it could still be significantly better than using existing multicam ptp based sync (I don't currently plan on testing it though).
Again , a dev would need to build or advise how much effort could be involved in swapping current usb remote voltage levels for twopush (and probably onepush) operation.
PS: There would be no need to recycle power to cameras using this proposed method (assuming none crash).
fmi:
e/ "Powered" USB-2 Hubs also have "Additional Capacitive" Power By-pass Filters. !!
f/ "Powered" USB-2 Hubs also have a "Current Limiting" Resistor for short circuit protection. !!
So, All this will have some "Sync" effect when the triggering your a550 Cameras via "Powered Hubs" ....