Samsung NX-1000 Synchronism Performance

The following are measurements of the synchronism of the shutter of the Samsung camera NX1000.
One set of measures are done with a single camera, that is, I measure how well one camera synchronizes with itself. What I mean by that is the difference between the delays of several shots.
This way it is possible to predict the sync of a stereo rig if there were a second camera identical to this.

Other set of measures is done with a rig of two cameras. In this test I measure the timing difference between the shutter of one camera respect to the other.

In both cases an oscilloscope is used to watch the timing of the Hot-Shoe flash trigger of the cameras, which can be taken as a proxy for the actual photo shot.

SINGLE CAMERA:
The following GIF animations shows several successive camera shots, which are commanded using the remote release SR2NX02 which is connected to the micro-USB port of the camera.
The scope is triggered with the top (CH1) which is connected to the Full-Press signal at the remote release. The timing of the Half-press (Focus) signal is ignored.
The signal at the bottom (CH2) shows the flash trigger taken from the Hot-Shoe. Hot shoe needs to be powered externally, for which I used a 6V battery through a 39K resistor.
Setup


The following animation shows a sequence of shots with the camera in "A" setting (Aperture Priority). The test shows an average trigger delay ~100ms (from full press to shot), which is shown by the "CurA" indicator. And most relevant for synch testing: a dispersion of the delay of ~4 display units at 500us/Div, that is ~2 ms.
This means that if we had another camera identical to this one, the synch error would be a random between -2ms and +2ms.

Scope Animation


The following is a similar test with the camera in "S" setting (Shutter Priority). The results are essentially the same as "A" mode:

Scope Animation



RIG OF TWO CAMERAS:
The testing procedure is similar to the one followed with the single camera, except that this time the two channels of the oscilloscope are connected to the Hot-Shoe of each camera, and the cameras are shot using a wireless release shutter that is connected to a 2.5mm jack built in the rig.
In some tests the scope is synchronized with the full-press signal (which is not visible because my scope only has two channels).
In another tests the scope is synchronized with one of the Hot-Shoes, which is most meaningful to assess the synch performance of the stereo rig.
Note that in spite both cameras of the stereo rig are shot simultaneously, this test is not based in a comparison between the two images. In fact we only show one of the images and the other is ignored because it is identical.

The test results that are shown here are done with the both cameras updated to Firmware V01.14. This improved the synch performance respect the first tests I did with one camera at V01.01 and the other at V01.14.
Setup


The following animation shows the Hot-Shoe pulses of both cameras respect their common Full-Press signal. It is noticed that both cameras show about the same dispersion as in the single tests above, but the camera at CH2 has more delay than can the one at CH1.
Camera 1 & 2 respect shot



This is the same test as above, only synchronizing the scope with the Hot-Shoe at CH1. The time difference shown here is the synch error of the stereo camera rig. Notice that further to the random dispersion of each camera there is a permanent error because one of the cameras systematically has more delay than the other:
Camera 2 respect camera 1


Following is an assortment of test conditions, trying to find the sweet setting that gets the best synch. Apparently the best is achieved when using the external shutter, but it may be perhaps because of the limited amount of tests because the synch error of the paired cameras can be worse than a single camera but never better.

Auto-Focus pushing on camera 1 Auto-Focus - Camera 1 shutter button
Auto-Focus pushing on camera 2 Auto-Focus - Camera 2 shutter button
Auto-Focus pushing on wireless release Auto-Focus - Remote release shutter
through 2.5mm jack in rig
Manual-Focus pushing on camera 1 Manual-Focus - Camera 1 shutter button
Manual-Focus pushing on camera 2 Manual-Focus - Camera 2 shutter button



Conclusion:
These tests show that a single NX1000 camera has a 0-2ms uncertainty of the shutter delay. This should be understood as an intrinsic error, in the sense that if we make a rig with another identical camera the synch error of the stereo pair would be somewhere between -2ms and +2ms.
Note that the extreme errors -2ms and +2ms are very unlikely because they require one camera exactly at 0ms and the other at 2ms. The sync error has a symmetrical distribution of probability between -2ms and +2ms centered at 0 ms

All this with two identical cameras, but this test highlights the cameras may not be fully identical. The cameras used for this test have the same intrinsic error due to delay dispersion, but additionally they show a permanent difference in delay about ~1ms. This increases the probability of the synch error, whose probability becomes non symmetrical

The good news is the permanent difference in delay can be compensated with a bit of electronics. Only the intrinsic dispersion can not be compensated.


Antonio Fernández Galaz