| Author |
Message |
    
Don Congdon
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, April 20, 2005 - 01:24 pm: | 
|
Since I am playing with my Retina's shutter, I am curious about measuring leaf shutter timings. When the nominal shutter speed is quoted in a camera's service manual, what does it really mean? For example, if 1/500 is nominally supposed to be 2ms, from what point is that number determined? From the moment the shutter starts opening until it is fully closed? Or just the fully open time?
I am thinking of using some light-gate instruments that I have in my lab to do some measurements, but I need to know what part of the shutter cycle is the one on which timing specs are based. Thanks! |
Tony Duell
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, April 20, 2005 - 01:35 pm: |
|
If you consider a 'perfect' leaf shutter, then the light-time curve at the film would be a rectangular pulse. It would open instantaneously, and close instantaneously.
A real shutter will have a light-time curve with sloping edges (maybe not even straight).
One possibility would be to measure the total amount of light coming through when you operate the shutter, and then use that to calculate the equivalent time for a perfect shutter. In other words, calculate the integral of the light-time curve, compare that to the area of a rectangle.
I think (I am doing this in my head) that if the opening and closing edges are straight, then this is the same as taking the time between the half-intensity points on the opening and closing edges
BUT -- and it's a big but -- you'll find that this time varies with the lens aperture. It has to (think about what the shutter actually does). I've yet to find a camera which makes any correction for this.
And I am pretty sure that many of the cheaper shutter testers (which are still not that cheap!) just measure the time while 'some' light is hitting the sensor. Since the threshold, and the intensity of ambient light, is not known, this will presumably not be very meaningful at the higher speeds. |
rick oleson
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, April 20, 2005 - 03:02 pm: |
|
tony's right.... accurately measuring the fastest speeds of a leaf shutter is pretty futile, since even a perfectly accurate one will give a different speed at every aperture. the really right way to test one would be to plot the intensity over time on an oscilloscope and take the area under the curve, but even doing this the aperture variability remains.
i have one of those cheaper shutter testers, and the prescribed method is to set it up with the light intensity adjusted to trigger the tester as the shutter passes the half-open point. this is easier said than done, and i have not gotten consistent readings trying to do this with a leaf shutter.
you can sort of see the problem graphically if you try to evaluate a leaf shutter using a CRT screen as described on my website: with a focal plane shutter in good trim you get a line whose width is proportional to the speed... with a leaf shutter (viewed through the lens) the line bulges, representing different effective speeds in different portions of the lens area (faster at the edges and slower in the center). |
charlie
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, April 20, 2005 - 06:36 pm: |
|
It sens to me that a slower shutter speed, say 1/50 sec. should be fairly independent of the slopes at the opening and closing, that the exposure departure from an instantaneous opening and closing would not be very large. Therefore if you had a fussy film in the camera, like Kodachrome, you could take a series of exposures starting with 1/50 and at each step halving the exposure time and doubling the aperture and compare the resulting slides they should all look alike if the shutter is acting like the indicated speeds. I can't think of any reason why the f stops should accumulate errors over time. And isn't this why we care about shutter accuracy in the first place? |
charlie
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, April 20, 2005 - 06:38 pm: |
|
Make that "seems" |
Henry
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, April 20, 2005 - 07:22 pm: |
|
Yeah, trying to measure shutter efficiency is fun...and frustrating. But the rangefinder designers took it into consideration. If you clean the mechanism well enough all should be as it was intended. And the overall exposures should be close enough.
Adjusting individual speeds is very difficult on most rangefinders.
Henry |
rick
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, April 20, 2005 - 07:38 pm: |
|
hi charlie:
yes, once you're a couple of stops down from the fastest speed in a leaf shutter the variability becomes insignificant. generally you can test the speeds conventionally at the slower speeds up to 1/125 or so and if those are okay generally the top 2 will be too. i think the factory generallyset the fastest speed assuming the lens would be pretty much wide open.... not always a good assumption, but as henry suggests, they generally hit close enough for the film's latitude.
not sure i understand your question about f stops accumulating errors over time.
rick : ) = |
charlie
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, April 21, 2005 - 04:51 am: |
|
Hi Rick - What I meant was that I assume f stops would not change significantly over time and that they would provide a reliable method of comparing shutter speeds when used with a narrow range of exposure film like Kodachrome. |
rick oleson
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, April 21, 2005 - 11:27 am: |
|
Hi Charlie:
The f/stops will be consistent, but at the highest speed or two the speed will not be the same for all f/stops. Due to the travel time of the blades, in order to get (say, for instance) the equivalent of 1/500 second worth of light to the film at full aperture, the total time from when the center begins to open until it's completely closed again might be, say, 1/250 second, with very little light coming through at the beginning and end of the time and only having a really full open condition for that instant as the blades reach their farthest open point and begin to close again. When the lens is wide open, this varying light intensity can be calculated out and adjusted to give the equivalent of a perfect 1/500. But, if the lens was stopped down to f/16, the full area of that small f/16 opening has been exposed very early in the process, and most of the time the blades spend opening and reclosing is lost because they are hidden behind the aperture blades. In this case, the shutter has been fully open at f/16 for the full 1/250 second or very close to it. At intermediate aperture settings, you get a compromise between these two extremes, so that the same, perfectly adjusted shutter gives a different effective speed for each aperture chosen. There is no way to adjust this out, it's just a fact of life with a leaf shutter that there will be roughly a one stop difference in shutter speed from the largest to the smallest aperture at the fastest speed setting.
At slower speeds this effect goes away, as the time spent with the blades stopped in the full open position far exceeds the time spent in a continuously changing condition... the slower the speed relative to the shutter's fastest speed, the more nearly ideal the speeds become.
: ) = |
charlie
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, April 21, 2005 - 02:33 pm: |
|
Something I never thought of. However, now that Kodachrome 25 is off the market, the higher speeds may be used with a larger variety of f stops and it becomes more significant. Sort of reverse reciprocity failure. |
|
