Measuring Multi-Camera Shutter Sync #twinrig #homebrew #3d-cameras


 

This topic comes up often enough to deserve its own thread.

I see that Werner still prefers the CRT synch test method.  Very
accurate, but of course it is cumbersome to find/keep a CRT for this
purpose.
I've got a large CRT and old Windows tower salted away just for this
one and only reason. But I will need to get rid of them at some
point. In fact the longer I keep them, the more they will cost to
dispose of.

We *really* need a good way to measure multiple camera sync that is
affordable and simple enough for the average 3D photo nerd to put
together.

There has been recent discussion about the laser/pendulum approach.
But I had forgotten about the flash shoe idea.

Perhaps this is a forgotten method, but the old SLR Shepherds took
readings from the flash shoe ports and compared the shutter release
signals through there.
At first blush this seems like an obvious idea, but there may be
limitations that are not occurring to me. Are there reasons why using
the timing of the flash signal would not work?

As for how to measure this, perhaps using a USB device and any laptop.
A few ideas that come quickly to mind:

- There are devices that will turn any laptop into a dandy two channel
Oscilloscope via USB.

- An inexpensive USB data logger?

- With a cheap USB external sound card, using the dry contact of the
hot shoe to drive a simple circuit that turns the event into a clean
'tick' that can be measured using a sound recording app?

So, I'll ask the question again, is the output of a hot shoe
sufficient to provide enough precision for useful measurement of
shutter sync?

...BC


Werner Bloos
 

is the output of a hot shoe
sufficient to provide enough precision for useful measurement of
shutter sync?
Yes.this should be fine since both cameras should act in same way so shooting differece is identical with the differece of the hot shoe signal -> you can measure missynch

The problem is that a hot shoe is not present on all camera models...

Werner


Depthcam
 

Werner made a most pertinent post on another thread which I feel needs to be reproduced here:


I see that Werner still prefers the CRT synch test method.  Very accurate, but of course it is cumbersome to find/keep a CRT for this purpose.
Well, the CRT synch test right now is the best test method... but i am about to develop two new test methods - both even more precise and without all disadvantages each of the other test methods have (including the CRT synchtest since even here we have some disadvantages)

An electronic-savvy person could conceivably construct a similar device that is entirely passive, and just reads the flash shoe outputs on the Sony models that have flash shoes
There might be also a way to do this via the MULTI terminal port.

Werner


Antonio F.G.
 

On Thu, Sep 10, 2020 at 10:06 AM, Bill Costa wrote:
I've got a large CRT and old Windows tower salted away just for this
one and only reason. But I will need to get rid of them at some
point. In fact the longer I keep them, the more they will cost to
dispose of.
If you are a bit familiar with the oscilloscope and solder iron, I would suggest
you trash the CRT and Windoze tower and replace them with an oscilloscope:-)

For timing measurements (as stereo synch is) the oscilloscope is THE TOOL (full stop)

Moreover: with an oscilloscope you can asses the synch error of an stereo rig using
only ONE CAMERA. Most of my measurements I show below are done with one
camera (because I only have one:-)


There has been recent discussion about the laser/pendulum approach.
But I had forgotten about the flash shoe idea.
Yes, I have measured the synch of all cameras I have at hand that have a hot-shoe.
The hot-shoe pulse can be taken as a proxy of the actual photo shot. It would be
better to measure the synch error in the actual photo but this can not be done with
a digital scope, which is the one I used for the tests.
Hot-shoe measurement also limits the maximum shutter speed, because most cameras
do not trigger the flash above some speed.

I have recently bought a very old analog scope and hope I could make synch tests
from the photos proper instead of the hot-shoe. I hope the analog scope can be used
in a similar way as the CRT test, only with way better precision and capability of control.
I have two test projects for this analog scope: testing my A720 SDM rig (that has no
hot-shoe), of which I only made a few CRT-like tests using an old TV. But the photos I got
on the (interlaced) TV screen are pretty blurred and confusing.


Following are reports of the synch tests I did using a digital scope. All of them looking at the hot-shoe:

Canon 350D: http://afgalaz.es/stereo/synch_canon_350d/

Fuji X-M1, X-E2: http://afgalaz.es/stereo/synch_fuji_x/

Samsung NX1000: http://afgalaz.es/stereo/synch_nx1000/



So, I'll ask the question again, is the output of a hot shoe
sufficient to provide enough precision for useful measurement of
shutter sync?
I hope so because the flash trigger should be carefully timed respect the shutter opening,
to ensure the lighting always happens with the curtains open.

Regards
   Antonio