Home About Me Content Gallery 3D Events Links Search Results

Type in a keyword to locate the information you seek.



Dual Camera Synchronisation Test

When photographing in 3D with Dual camera rigs, it is useful to have a means of comparing the synchronisation accuracy of the pair when triggered with with various shutter release devices. With this in mind I built a very simple test rig, one that didn’t require specialist knowledge or equipment.

In this article,  I do not propose to offer a diagnosis of my test results, rather to describe the the procedure that I have adopted to conduct the tests.  Hopefully, this will help others to carry out their own investigation.

There have been several offerings in the public domain on this subject, but the measurement tools involved often require photographing a moving target having a defined speed, such as a motorised calibrated disk or wheel and all that this entails, known rotational speed, lighting etc.  Digital counter displays are sometimes used, but these have their own latency issues.

My own measurement solution is to photograph the trace provided by a pendulum mounted laser pointer.  To do this we simply need a metric scale or rule having millimetre divisions and a readily available laser pointer.  The laser pointer is mounted to a ‘home made’ pendulum of some sort.

The pendulum is positioned above the scale, so that the laser projects a trace on the scale as the pendulum swings.

Laser based test rig

Pndulus laser pointer and tablet used to trigger the cameras

The cameras, fitted with appropriate close-up lenses, are positioned at close range and are manually triggered by the chosen remote shutter release.  In practice, the speed of the swinging pendulum need not be known, as the photographically captured laser trace length will correspond to the cameras Tv setting.  Consequently, any spatial difference between the recorded traces will provide a synchronisation disparity value.

In practice, a pendulum length of around 1 meter is about right, but some practice is required to repeatedly capture the laser trace at the right moment, midway through the swing and directly in front of the cameras being ideal.  It is important that the images are all captured with the pendulum travelling in the same direction, to avoid confusion about which camera may be lagging the other.

Each cameras are set to manual mode, with an exposure time value (Tv) of 1/200th  second (5 milliseconds).  At this setting, the laser trace will appear on the captured image as a line corresponding in length to the Tv.  A  slower Tv of say 1/50th second for example, will therefore present a trace length of four times the latter (20 milliseconds) and so on.

Knowing that the length of each trace corresponds directly with the Tv setting of the two cameras; then the distance between the leading edge of each trace will represent the synchronisation disparity between the cameras.  

In the following example the Tv was set to 1/125th second (8 milliseconds).  The laser traces produced were 12 mm. in length (12 mm. = 8 milliseconds) and the spatial difference between the two traces is 15 mm., a disparity of 10 milliseconds between the triggering of the two cameras.

To visualise a left to right swing for comparison purposes,  images from each camera were opened in StereoPhoto Maker (SPM) and mounted in side-by-side format, aligned, cropped and saved as a side-by-side pair for record purposes.  

Further cropping was then carried in preparation for an above/below printout. For calculation purposes

In this example, the cameras were set with a Tv of 1/200th second, the resulting trace length is about 13mm therefore 15/13 X 5 = Error ( 5.2 milliseconds approx.)

This may all sound very simple, but in fact it is quite an accurate way of measuring the synchronisation discrepancy of dual camera rigs, requiring little skill and effort.  My own test rig is pieced together as and when required, some of the components being used for other purposes.

Final image from SPM (left above right) indicating 15mm triggering lag of right hand camera.