I BOUGHT a shutter speed tester the other day, It’s nothing fancy, just a black box with a light sensor on one side and a cord that plugs into a computer’s audio jack. I got it because I use a lot of old or antique cameras, from a Pentacon Six, to a Mamiya C220 to 70-year-old folders.
I started using antiques to see what the results would be like and as my respect for this equipment grew, I wanted to use it to try to produce high-quality work. But while my old cameras seemed to work well, the results from them varied wildly. I started to prefer one to another without knowing whether my metering by guesstimate was to blame or if there was some deeper cause such as out-of-date film or poor shutter speeds.
Hence the little black box. At last I would find out what those old leaf-shutters (the kind made of little metal blades that sit inside the lens, rather than back next to the film) were doing.
My first two sets of tests astonished – and dismayed – me. They were on a Welta Perle 6×4.5 folding camera made in the early 1930s for amateurs and a professional-level Mamiya C220 twin-lens reflex (TLR) camera from the mid-70s. Both were recent acquisitions and I had shot just one film with each.
The little Perle had produced unsurprisingly shaky results when handheld for late afternoon street shots, which told me it was working but not much more.
However, when I checked the shutter times I found to my astonishment that they were almost perfect, except for the highest speed which is rarely ever close to the advertised timing on old folders. The actual results for the Perle were: 1s is really 1.1s; 0.5s = 0.6s; 1/5s = 1/5.5; 1/10 = 1/11; 1/25 = 1/29; 1/50 = 1/53; 1/100 = 1/95 and 1/250 = 1/160.
I then checked a Mamiya Sekor 80/2.8 lens which came with my Mamiya 220 TLR. These cameras have leaf shutters in the lenses which are interchangeable, That means the shutter speed is for a particular lens and has nothing to do with the camera body.
My Sekor 80, unfortunately, turned out to work at almost exactly half the proper shutter speed at every setting. At least that’s easy to compensate for. Despite this being the standard lens for a professional level camera, there are a number of reports on Flickr’s Mamiya TLR group of 80mm Sekors running slowly.
My other Sekors, the 65mm and 105mm (bought from a different source) turn out to be much better. They both run perfectly at both the high and low speeds. Curiously, there is a midpoint where both of them are out of kilter: one is wrong at 1/125 and one at 1/250, but both of them are accurate at the highest speed. I can only assume that this is the point where new springs are bought into action as the shutter switches from the lower to higher gear trains.
A different problem came to light with my Pentacon Six cameras, both of which have been serviced by Rolf Baier who is one of the world’s leading experts on the maintenance and repair of these cameras.
As I didn’t want to risk snagging the curtain in the body of the Pentacon six, I initially tried to measure the shutter speed by putting the detector in the place where the lens fits, well in front of the shutter. To my astonishment, the exposures measured correctly up to about 1/30s and then stayed the same for all the higher shutter speeds. Was there a major defect with the camera? No! The answer lies in the way a focal plane shutter works.
Instead of opening completely and then shutting, the way a leaf shutter does (where the shutter is incorporated in the lens), a focal plane shutter has two curtains and at high speeds the second shutter starts to close before the first shutter is completely open.
As the shutter speed increases, instead of the shutter moving faster, the gap letting the light through gets smaller. In my Pentacon Six, the shutter curtain moves from one side to the other in 1/34s. To achieve 1/60s, the second shutter starts moving as soon as the first one is half open, at 1/125 the gap between them is 1/4 the width of the film gate, at 1/250th it is 1/8th and at 1/500 the moving slit is just 1/16th of the width of the 58mm gap. At 1/1,000 it is down to 1/32nd, or just about 2mm wide,
This has consequences for any effort to measure the shutter speed on a Pentacon Six (or any other focal plane shutter).
To start with, the measuring device must be on the film plane and it must, as near as possible, mimic the experience of one single grain in the emulsion.
By cutting out a piece of card exactly the same size as a piece of 120 film, making a hole in it for the sensor on the device to “see” through and then taping that to the back of the camera it is possible to measure the shutter speed with reasonable accuracy up to perhaps 1/125s. Beyond that, the experimental error starts to swamp the data.
If the hole for the sensor to see through is 2mm wide, then at 1/1,000s it will start “seeing” light as the 2mm wide slit arrives at the beginning of the hole and will stop “seeing” it as the slit reaches the end of the hole – a combined distance of 4mm. That means the reading will be for the (width of the gap in the curtains + the width of the hole in the card) introducing 100% error.
From the front of the camera, the sensor just measured the minimum time it took for the curtain to move from one side of the camera to the other (1/30s). When I put it into the film gate behind a bit of card, it was able to measure with reasonable accuracy all the shutter speeds up to the point where the size of the hole in the card approached the size of the gap in the curtains. That is about 1/125 (where the gap in the curtains is 15mm and the error is not much more than 10%). At 1/250 the error becomes 20% at 1/500 it is around 40% and at 1/1000 it is close to 100%.
My actual readings were: 1 = 1s; ½ = ½; ¼ = ¼; 1/8 = 1/8; 1/15 = 1/18; 1/30 = 1/28; 1/60 = 1/62; 1/125 = 1/125; 1/250 = 1/229; 1/500 = 1/330; 1/1,000 = 1/560 which shows the speeds to be completely accurate as far as it is possible to measure them by this method.
The lessons from this are: Old leaf-shutter cameras can still be almost perfectly accurate 70+ years after manufacture but they can also be way off their setting after only a few decades; speeds may be accurate at low and high settings but still be out-of-kilter somewhere in the middle range; curtain shutters cannot be accurately measured by this sort of device at higher speeds, because of the nature of the shutter.
Above all, if you want to get accurate exposures from old cameras you must test the shutters properly. Neither condition nor quality provide any guide to how well they work – my most perfect, almost unused, folding camera has speeds that are way out but a tatty Super-Ikonta made five years later and an equally tatty Welta Perle made five years before are both spot on the correct times (except for the highest speed, which usually disappoints).