ridax
10-Jul-2013, 11:58
There were posts on between the lens shutter efficiency recently in another thread discussing a specific shutter speed tester... I usually don't like to start new threads (except when asking an urgent question). When I want to say something that I think other people should know I actually prefer to wait for the topic to be put on by others; that way I feel more confident people really want to explore it. But this time I think the topic deserves way more then a casual reference in a discussion of testing devices (ironically going on in two separate threads already now - where it may get buried down and lost).
The problem: no shutter opens and closes perfectly immediately. Opening and closing of the blades takes time, and the whole working cycle includes this opening and closing time along with the time the shutter stays open completely. Set on the fastest speed, the shutter starts closing as soon as it is fully open. In this case, its efficiency is somewhat about 50% (it's way more complicated really as many factors influence the actual percentage - the number of the blades and their shapes, weight of the blades and other moving parts, etc. but I think we don't need to go that far in this thread). 50% efficiency means that to let the same amount of light through, the shutter has to stay more or less open during the time twice as long as a perfect 100%-efficient shutter (that opens and closes in no time) would.
First, coming back to shutter speed testing - what do we want those tests for? When concerned about the motion blur danger, we naturally should measure the time from the beginning of the opening till the end of the closing processes. But when we are concerned about exposure accuracy, we have to measure the equivalent speed of a perfect shutter, not the actual speed of the shutter tested - i.e. evaluate the total amount of light that gets through. We may want quite different testers for the two tasks. For exposure tests, an integrator is a natural way to go (a flashmeter works great btw. Though I may be too naive a guy to believe there are still a lot of photographers out there carrying flashmeters instead of smartphones :) ).
It's usually said a shutter efficiency is that low only with its fastest speed and only when the aperture used is nearly as large as the maximum shutter opening. While the latter is perfectly true, and stopping the lens down really makes the shutter efficiency greater, the former statement is not true for the vast majority of real world shutters. In textbooks, a between the lens shutters is said to open quickly, then stay open (except on the fastest speed) for the time needed and then quickly close. In reality, ONLY Compurs and Compounds work like this. (And so do Compur clones like the Russian Moment found on the Moskva roll film cameras, and smaller sized shutters on Smena and Liubitel - one of which is seen in the pictures the tester designer has posted in the above mentioned thread btw.)
ALL other shutters that I've put my hands on (including Copal, Copal-Press, Prontor-Press, Prontor Professional, Seiko, Alphax, Betax, and so on) work this way on the really slow speeds only, very often no faster then 1/15. And with the faster speeds, the retarding device does not wait till the blades are open; it just slows down the blades' movement itself. The result is the same really low (circa 50%) shutter efficiency on all the faster speeds (often enough, just from 1/30 or so for shutters as modest in size as #1).
Why do they manufacture them like this? That's simlple: the springs may be weaker; the materials used may be of lesser quality - cheaper and softer and easier to deal with in production. A number or parts may just be pressed out of those softer materials instead of being machined out of hard stainless steel. The shutter doesn't need to be as robust as a Compur. A flimsy shutter will work OK and last long enough if its parts move way more slowly at all the speeds but the fastest one. The shutter comes out cheaper and drives the really costly highly efficient but nevertheless bulletproof Compur out of the market. (Add a 5-blade iris instead of a round one, and you've won the price competition completely.)
Lots of people do not like Compurs though. Some even spend money to have their lenses remounted into Copals, especially when the shutters are large. There really is a reason for that. With their fast moving and abruptly stopping blades and other parts, bigger Compurs can considerably shake the lens and the camera. It's less dangerous for a heavy metal 8x10 Linhof Kardan Color or a Calumet C-1 but can be a disaster for a wooden field. That's why Herr Deckel did not manufacture bigger Compurs in those old days (though the dial-set Compurs were, and still are, far less prone to vibration then the later rim-set ones) but offering only Compounds in the larger sizes. The pneumatic retarding device used in the Compound is also a natural shock absorber - and btw one of the best ever invented. And Compound is a pretty efficient shutter with fast moving blades that open quickly, then stop and stay open, and then quickly close. I'd never trade a good clean Compound for any Copal or a Prontor - except for shooting slide films, as mechanical shutters are more accurate of course.
How does this influence the picture?
First, motion blur looks really bad too often (maybe except when all the motion goes in one direction, as with a steady wind or movement of the subject itself; but I'd say even in those situations the blur does not come out nice every time). A Compur (or a Compound) having 50% efficiency at 1/250 of a second spends 1/500 on opening its blades and 1/500 on closing them - at all speeds alike. Set at 1/30, it would be actually open for a negligible fraction of a second more then 1/30. Its efficiency at 1/30 would be very close to 100%. ANY other shutter - that has its blades moving slower at slower speeds - would maintain 50% efficiency at 1/30 and will actually stay more or less open for 2/30 = 1/15. For maintaining, say, a tree leaves sharp in a slight wind, that difference may turn out to be fatal.
Second, the bigger part of the actual exposure time the shutter blades still get in the way, the more they influence the character of the out of focus parts of the image - just like the aperture blades do. If the opening and closing shutter blades could maintain the opening itself perfectly round all their moving cycle through, that could be good news actually. In that case the picture could possibly look better then with a very efficient shutter - al least with a so-so lens having little ability to make a good looking out of focus blur on its own (though the picture by a good lens that blurs the out of focus image beautifully could be quite a bit spoiled even in this situation). But a half-open shutter is not a round aperture maker at all. What it actually makes is a badly misshapen star-like opening. Imagine a picture made with that kind of aperture in your lens... or better still, stop your shutter blades in the half-open position and actually see the resulting picture. Yes tastes differ of course but I personally was just shocked by the nasty shapes I saw out of focus the day I run a test like this.
Certainly a real picture taken with a slow-blades shutter is never as ugly because the blades are moving and also because the shutter is fully open part of the exposure time. But the resulting picture is still somewhat in-between the normal one and the one visible in the above test. And this resulting picture is not quite as beautiful as the one taken with a highly efficient shutter.
The problem: no shutter opens and closes perfectly immediately. Opening and closing of the blades takes time, and the whole working cycle includes this opening and closing time along with the time the shutter stays open completely. Set on the fastest speed, the shutter starts closing as soon as it is fully open. In this case, its efficiency is somewhat about 50% (it's way more complicated really as many factors influence the actual percentage - the number of the blades and their shapes, weight of the blades and other moving parts, etc. but I think we don't need to go that far in this thread). 50% efficiency means that to let the same amount of light through, the shutter has to stay more or less open during the time twice as long as a perfect 100%-efficient shutter (that opens and closes in no time) would.
First, coming back to shutter speed testing - what do we want those tests for? When concerned about the motion blur danger, we naturally should measure the time from the beginning of the opening till the end of the closing processes. But when we are concerned about exposure accuracy, we have to measure the equivalent speed of a perfect shutter, not the actual speed of the shutter tested - i.e. evaluate the total amount of light that gets through. We may want quite different testers for the two tasks. For exposure tests, an integrator is a natural way to go (a flashmeter works great btw. Though I may be too naive a guy to believe there are still a lot of photographers out there carrying flashmeters instead of smartphones :) ).
It's usually said a shutter efficiency is that low only with its fastest speed and only when the aperture used is nearly as large as the maximum shutter opening. While the latter is perfectly true, and stopping the lens down really makes the shutter efficiency greater, the former statement is not true for the vast majority of real world shutters. In textbooks, a between the lens shutters is said to open quickly, then stay open (except on the fastest speed) for the time needed and then quickly close. In reality, ONLY Compurs and Compounds work like this. (And so do Compur clones like the Russian Moment found on the Moskva roll film cameras, and smaller sized shutters on Smena and Liubitel - one of which is seen in the pictures the tester designer has posted in the above mentioned thread btw.)
ALL other shutters that I've put my hands on (including Copal, Copal-Press, Prontor-Press, Prontor Professional, Seiko, Alphax, Betax, and so on) work this way on the really slow speeds only, very often no faster then 1/15. And with the faster speeds, the retarding device does not wait till the blades are open; it just slows down the blades' movement itself. The result is the same really low (circa 50%) shutter efficiency on all the faster speeds (often enough, just from 1/30 or so for shutters as modest in size as #1).
Why do they manufacture them like this? That's simlple: the springs may be weaker; the materials used may be of lesser quality - cheaper and softer and easier to deal with in production. A number or parts may just be pressed out of those softer materials instead of being machined out of hard stainless steel. The shutter doesn't need to be as robust as a Compur. A flimsy shutter will work OK and last long enough if its parts move way more slowly at all the speeds but the fastest one. The shutter comes out cheaper and drives the really costly highly efficient but nevertheless bulletproof Compur out of the market. (Add a 5-blade iris instead of a round one, and you've won the price competition completely.)
Lots of people do not like Compurs though. Some even spend money to have their lenses remounted into Copals, especially when the shutters are large. There really is a reason for that. With their fast moving and abruptly stopping blades and other parts, bigger Compurs can considerably shake the lens and the camera. It's less dangerous for a heavy metal 8x10 Linhof Kardan Color or a Calumet C-1 but can be a disaster for a wooden field. That's why Herr Deckel did not manufacture bigger Compurs in those old days (though the dial-set Compurs were, and still are, far less prone to vibration then the later rim-set ones) but offering only Compounds in the larger sizes. The pneumatic retarding device used in the Compound is also a natural shock absorber - and btw one of the best ever invented. And Compound is a pretty efficient shutter with fast moving blades that open quickly, then stop and stay open, and then quickly close. I'd never trade a good clean Compound for any Copal or a Prontor - except for shooting slide films, as mechanical shutters are more accurate of course.
How does this influence the picture?
First, motion blur looks really bad too often (maybe except when all the motion goes in one direction, as with a steady wind or movement of the subject itself; but I'd say even in those situations the blur does not come out nice every time). A Compur (or a Compound) having 50% efficiency at 1/250 of a second spends 1/500 on opening its blades and 1/500 on closing them - at all speeds alike. Set at 1/30, it would be actually open for a negligible fraction of a second more then 1/30. Its efficiency at 1/30 would be very close to 100%. ANY other shutter - that has its blades moving slower at slower speeds - would maintain 50% efficiency at 1/30 and will actually stay more or less open for 2/30 = 1/15. For maintaining, say, a tree leaves sharp in a slight wind, that difference may turn out to be fatal.
Second, the bigger part of the actual exposure time the shutter blades still get in the way, the more they influence the character of the out of focus parts of the image - just like the aperture blades do. If the opening and closing shutter blades could maintain the opening itself perfectly round all their moving cycle through, that could be good news actually. In that case the picture could possibly look better then with a very efficient shutter - al least with a so-so lens having little ability to make a good looking out of focus blur on its own (though the picture by a good lens that blurs the out of focus image beautifully could be quite a bit spoiled even in this situation). But a half-open shutter is not a round aperture maker at all. What it actually makes is a badly misshapen star-like opening. Imagine a picture made with that kind of aperture in your lens... or better still, stop your shutter blades in the half-open position and actually see the resulting picture. Yes tastes differ of course but I personally was just shocked by the nasty shapes I saw out of focus the day I run a test like this.
Certainly a real picture taken with a slow-blades shutter is never as ugly because the blades are moving and also because the shutter is fully open part of the exposure time. But the resulting picture is still somewhat in-between the normal one and the one visible in the above test. And this resulting picture is not quite as beautiful as the one taken with a highly efficient shutter.