Shutter lag measurements - Canon EOS 400D / Digital Rebel XTi
After triggering a camera (mechanically or electrically) it takes a certain time until the shutter opens and the exposure can begin. This delay is called the shutter lag. The values vary strongly depending on the camera model. The focusing time is not considered in this test. In all test images the focus was adjusted manually for fastest response times.
There are several ways to determine the shutter lag of a camera such as photographing falling objects, rotating discs or an electronic counter. I chose the digital method, due to a simple implementation and good readability of the results. To implement the timer I used my FPGA board on which I implemented a binary 8-bit counter with a time step of 1 ms. This counter was connected to 8 LEDs showing the time elapsed since the triggering. The camera was triggered by an optocoupler to reduce the risk of damage. To record the time stamp the camera was set on manual focus, 1/1250s, f5.6, ISO1600 and pointed on the counter. Having everything working properly I shot some series of test pictures. An example is shown below.
Results for Canon EOS 400D / Digital Rebel XTi
The shutter lag was measured with and without mirror lock-up. To make the results more reliable both test series consist of 16 images. Pictures with ambiguous time steps due to a jumping counter were sorted out before the analysis. The results are shown in the histogram below.
The bars correspond to the frequency of occurrence for each shutter lag value. The shutter lag varies around 115 ms for the case without mirror lock-up. Enabling the lock-up mode reduces the lag almost by a factor of two to about 66 ms. An interesting fact revealed by the test is that the shutter lag is not constant. A lag of 65 ms is an average value, but good enough to use the camera with a lightning trigger. A microcontroller based DIY lightning trigger circuit using mirror lock-up is described here. I also measured the minimum trigger impulse duration for reliable trigger operation (see plot below). For pulses shorter than approx. 15 ms the trigger failure rate increased rapidly. The results with the corresponding timing diagram are summarised below.
|minimum trigger time||tt||15 ms|
|shutter lag w/o mirror lock-up||td||116 ms|
|shutter lag with mirror lock-up||tdml||66 ms|
I repeated the measurements later to find out the influence of the aperture setting on the shutter lag. The aperture can potentially increase the shutter lag, because it must be closed each time before the exposure begins. The results were very similar to those described above. The aperture had no influence on the shutter lag.
Comparison of shutter lag and startup times for different cameras
Also, it should be noted that the measured times can vary by something like 4 ms depending on the vertical position of the counter in the frame.
fantastic web site... You gave me new inspirations.
Looks like can be ignored
Not even close to a second. IMO can be ignored.
Depends on the purpose.. If you're doing highspeed photography 100ms is huge!
How about the same data in teethered or triggered by software SDK?
Anyone did this test?
Any idea what is the expected lag time?
Using an old flash on a modern DSLR is usually no problem, unless the trigger voltage is too high. Trigger voltages of many old flash units can be found on this great site: http://www.botzilla.com/photo/strobeVolts.html
Do you have any cirquit to eliminate this problem so I could use my exelent old flash as an external? Thanks, Yoram