Inspired by this recent posting,
Photographing strobes. - PentaxForums.com, wondering (as I understand it) among other things how one might measure the burst rate and/or burst interval when shooting, I have slightly re-purposed my shutter speed timing array (
Shutter Speed Measurements: Pentax (K-1, K-3, Q7, *istD) and Canon (G15, Rebel XS) - PentaxForums.com) so that I can measure the interval between the images in a burst.
Basically, I set some blink rate for the array which should allow some number of blinks (LED steps) between shots. Each frame in a burst then captures just a few of the LEDs lit up, depending on exposure time. By noting explicitly which LEDs are lit, I can measure the time interval between shots.
Here’s how the LEDs are arranged in 4 sub-arrays, where the numbers at the bottom of each frame indicate the order in which the columns are lit, and the vertical values (1 through 8) show the order in which the LEDs light up once a column has been activated. At the beginning of this video (
http://photodave.us/ShutterSpeedTesting/LED-Array.mp4) you can see how the LEDs light sequentially as the signal generator sends pulses to the array.
Thus, I consider the lower left corner of sub-array A to be LED #1. The next pulse lights up LED #2 (2nd LED in column A1). After LED #8 at (the top of the first (left-most) column of sub-array A), the next LED to light up is #1 in the left-most column of sub-array B. By noting which LED(s) are lit up in a frame, I can assign a pulse count N to the LED by calculating
N = 8*column value + Row value
By taking the difference between frame-to-frame N values (with proper allowance for when the array starts over again), I know how many counts elapsed between images.
If the pulse rate to the array is frequency F, then the time interval between pulses is just 1/F. So, by multiplying the count between frames by the time interval, I get the actual time difference between frames.
Here are three cropped frames from a ten-frame burst I shot with my K-3 III. The pulse rate is 500 pulses per second, and I used a shutter speed of 1/125 second (really 1/128 second), so I expect about 4 LEDs to be lit within an exposure (depending on exactly when the exposure starts, there may be a fractional LED lit up at one or both ends of the exposure). I used my SMC Pentax-DA 18-135mm F3.5-5.6 ED AL [IF] DC WR lens at about 40mm zoom length and f/8. I used ISO 800 so the LEDs showed up nicely, and shined some light on the overall array so I could readily see which column and row were illuminated. I pre-focused the lens, and autofocus was turned off.
For the first frame, I get column #6, and row #5.5 as the LED that is lit up (estimating the middle of the LEDs), for the second frame, I get column/row 13/2.5, and for the third frame, 20/0.5 (note that two of the sub-arrays have illuminated LEDs). These convert to times of 0.107, 213, and 0.321 seconds, with respect to an arbitrary start time. The delta Ts (time between frames) is 0.106 and 0.108 seconds. These intervals are consistent with all of the 9 intervals that I measured. The average interval is 0.1076 seconds. The corresponding burst rate here is a bit more than 9 frames per second.
The manual for the K-3 III claims a maximum burst rate of 11 frames per second. The actual exposure time must affect the burst rate - you could not possibly have a burst rate of 11 frames per second if your exposures were, say, half a second! If I subtract my exposure time (0.0078 seconds) from the intervals, I get almost exactly 0.100 seconds between end of one exposure and the start of the next one - or, a burst rate of 10 frames per second - not too far from the claimed 11 per second.
I am planning to test my K-3 (original model) as well as my K-1 with this set up. I will also try shorter exposures. I do not know what other camera settings may affect the burst rate, such as f-stop. If the lens opens up and then stops down between exposures, that may well affect things. Therefore, I will also try some sequences at wide-open f-stop.
Any other ideas for testing?