A Trigger for Lightning Photography
My first and IMHO still best lightning picture was taken in the spring of 1980(?) with a Voigtländer `Bessamatic'. The exposure was done the `classical' way, i.e. with a night-time thunderstorm, the camera set on `bulb' and a cable release.
Over time I gained the impression, that most thunderstorms come at times that aren't particularly well suited for the `classical' approach of photographing them, because it's still too bright. I knew there were specialists who waited with a finger on the shutter release, ready to depress it at the slightest hint of a lightning stroke and achieving a hit rate of an astonishing 20% -- it's possible because a single flash of lightning often consists of multiple `sub'flashes -- but that would prevent my enjoying the show.
For the
1999 solar eclipse I had bought myself two winders for my trusted OMs.
Together with my slightly modified (1985!!) Sharp PC-1402 `Pocket Computer',
they should enable me to enjoy all of totality, without having to care too much
about my ambitious photography program. And really, except for one short glance
to reacquire my grip on the speed lever, I could keep my eye on the sun without
a break ...
The remote socket of the winders could, of course, be utilized in many ways. Not long after the eclipse I realized that it could also be used for lightning photography, if I found a `clever' way to wire it to a photosensor.
The `clever wiring' consists of the circuit depicted in the image below.
First the signal from the photo-diode has to pass an operational amplifier (OpAmp)
wired as a differentiator; after all, the circuit is only supposed to react to
changes in the light level happening with (lightning!) speed.
The second OpAmp
amplifies the signal, whereas the third one ensures that it is applied to the base
of the transistor long enough, to result in a certain release of the winder.
The circuit can be quite sensitive. It even reacted to a flash discharged (on a sunny day) in an adjacent room, when the light had to travel out through a doorway, along a hallway and (of course) in through a second doorway (12 meters and two right-angle turns all in all), when NOT behind the camera viewfinder (the light from the flash wasn't noticeable to the naked eye anymore). Unfortunately somebody must have leaked the existence of this circuit to some lightning bolts. As lightning seems to be camera shy, the really picturesque displays have been avoiding me ever since. So all I've got are the images of the trial run -- 5 visible bolts in 8 frames. When I get around to scanning one I'll place it here, but don't expect a memorable image (trial run!!!) ...
Because the trigger only reacts to (extremely) quick changes in illumination levels (it doesn't react at all when I turn on/off the [incandescent] light in my room) it could be used day and night, but there's not much sense in doing the latter -- personally I prefer the ``classical'' approach for nighttime thunderstorms. Sadly, daytime use also has its limits. If it's too bright and the proper exposure time drops below a tenth of a second your chances of catching one of the following `sub-flashes' also drops. Of course you can increase the exposure time by stopping down the aperture, but try not to go beyond f/4 or f/5.6 because the strokes will become less impressing the further you have to stop down ...
Feel free to copy this circuit if you want to, but don't forget that the values
of the components given in the diagram, especially those following the third OpAmp,
are optimized to ascertain release with Olympus winders (both 1 & 2).
The same applies to the polarity of the 2.5 mm (mono) RCA plug.
And it's a pity that the following disclaimer really seems to be neccessary:
Of course it's much simpler to have the photo-diode `look' through the finder
of your camera. I should have had this idea myself, but I have to admit that I
gleaned it from
I took a few test frames with one of my OM-1s on my Winder-1 and also both Winder-2s to determine the time lag between the triggering event (the lightning stroke) and the instant at which the shutter was fully open (X-sync condition). The Winder-1 (Ser. # 1087xx) displayed a lag of a mere (126+/-11) milliseconds, but only made an image at every second `lighning stroke' (giving just a `whirrr' at the first). The older Winder-2 (Ser. # 3476xx) had a lag of (164+/-3) ms and the `newer' unit (Ser. # 4580xx) reached (185+/-10) ms. The batteries employed were four fully charged NiMHs. To have conditions corresponding to those while taking lightning shots, MLU was not used.
Exact numbers (including spread) and details of the measurement will
be posted here as soon as I've got the film developed and made a sketch
of the setup (turntable, laser-pointer and a few odds `n' ends).
The table below summarizes the delay times I obtained with my fastest Winder-2 and one of my OM-1's. My other Winder-2 adds 30 ms in every field with a spread twice as large, my Winder-1 shows erratic behaviour. Contrary to my expectations using MLU (mirror lock up) does shorten the lag! The difference might be `only' 30 ms, but it just could decide between a `hit' or a `miss'.
Using MLU and letting the sensor `see' through the viewfinder are, of course, mutually exclusive -- unless you can or want to use another camera for this. In the latter case I'd recommend a slightly longer lens on the `trigger'-camera, to reduce the number of frames with the lightning bolt on the extreme edges of the frame. If you don't have an OM-1 this point is moot anyway, if you do you could still rig a different kind of `viewfinder' for the sensor.
In case you don't want to give away a single millisecond you could also stop down the lens in such a way, that the aperture lever of the lens is fixed in the `engaged' position. This would steal you yet another 3 or 4 ms, but I wouldn't recommend tampering with the aperture mechanism.
| MLU \ Lens | with lens | without lens |
| with MLU | 126+/-2 ms | 122+/-2 ms |
| without MLU | 154+/-3 ms | 151+/-3 ms |
I gleaned from the Olympus mailing list archives that the governor of the OM-2 S/P was slowed down to enable the `Program' mode to function properly. This results in a maximum frame rate of 3.5 fps in contrast to 5 fps for all other one-digit OM's (with a MotorDrive-1/2). In other words, the OM-2 S/P should take 85 ms longer per frame.
This effect can (unfortunately) also be seen with the ligthning trigger. Using the same winder as for the table above my OM-2 S/P showed a shutter lag time of 225+/-5 ms. That's 71+/-5 ms more than that OM-1 and in sufficient agreement with the 85 ms mentioned above. So if you have a choice, don't use your OM-2 S/P to chase those lightning bolts.
Repeating the measurement with my other OM-1 I found consistently shorter time delays. This (older) camera seemed to be about 30 ms faster than the younger one tested above -- until I noticed that the flash-sync lever was set on `FP-sync'. Setting it on `X-sync' revealed it to differ by only 3 ms from the younger OM-1, but leaning to the longer side. Sadly this older OM-1 shows erratic behaviour when using it with MLU, but only when using the winder (go figure!). So much for taking stereoscopic images of lightning with the shortest possible time-delay ...
The 30 ms mentioned above show that the flash-circuit is closed a 1/60th of a second before the first shutter curtain starts moving at all if the camera is set to `FP-sync'. Maybe someday I'll find a use for this (like triggering a fast event to be photographed with the camera by the camera).
The image below shows my measuring setup for the shutter-lag of different winder-camera combos. The inset (lower left corner) shows a real measurement picture taken with an OM-1 employing MLU, revealing a lag of 123 ms for this exposure. The time added by my lightning-trigger electronics should be below 1 ms and thus completely hidden in the several millisecond spread.
If you're only interested in finding the fastest combination of your Oly equipment (I hate to admit it, but it will work just as well even for N*k*n or C*n*n or somesuch), simply replace the laser-diode and lightning-trigger with a small magnet and a reed switch.
If your turntable runs at 45 rpm it completes a revolution every 1 1/3 seconds or 1333 milliseconds. You just need to copy a meter-stick (yes, `metric' really shines for this ...) with such a scale that the copied `meter' would wrap around 3/4 of the turntable and each `millimeter' will correspond to a millisecond.
To set the zero-point, simply align the beginning of the scale with
the index-mark (making sure the turntable doesn't move) and adjust the
laser-pointer to hit the sensor or move the magnet until it just triggers
the reed switch and you're done. That's all there is to it.