Shutter Speeds - More accurate than we often suppose?

[David Karp]

Over the years, I have read on many occasions that our shutters are less accurate than we would like to believe, especially older shutters.

A while back, I purchased the Shutter Speed app for my Android devices. This is the one created by Lukas Fritz (http://www.photoplug.de). I first used it without his PhotoPlug accessory and was very surprised that most of the tested shutters were quite accurate (assuming, as I do, that the app is accurate). Below 1/125, most of them are almost right on.

I was worried that using the app without the PhotoPlug was leading me to bad results, even though my results were repeatable.

So, I purchased the PhotoPlug and it sat around for months and months. This weekend I pulled it out and started making some measurements with my Nexus 7 tablet. I tested older style (jagged edge shutter speed ring), silver block style, and all black style Copal shutters. It did not seem to matter how old they were. I also tested versions used by Rodenstock, Nikon, and Fuji. Again, it did not matter which ones I chose. They are all mostly right on target. Of course, the higher speeds (above 1/60) are all off. However, they are not off by much. Usually, 1/125 and 1/250 are off by 1/3 or 2/3 of a stop. The top speed, 1/500 or 1/400, are, as expected, off by much more. Usually 1 1/3 stop.

All but one of my lenses are used. I have not yet tested the one I purchased new. None have been serviced since purchased by me! The one I sent to Carol, she sent back after making a minor fix for free!!

Based on what I have read, I expected more deviation from the marked speed, but I found that with the PhotoPlug the shutters were more accurate than I expected after using the app without it. So, I am wondering what others have experienced. Am I lucky in the shutters I have tested so far? Or is this a typical experience?

I have yet to test some Seikos, an Ilex, and some Wollensak shutters. We will see how that goes.

[Jac@stafford.net]

Based on what I have read, I expected more deviation from the marked speed, but I found that with the photoPlug the shutters were more accurate than I expected after using the app without it.

Your assertion certainly has my attention and I look forward to more comments from our readers. I've worked for forty years by finding a couple shutter speeds I could trust, and adjusting aperture or film speed to accommodate the one or two shutter speeds that work.

[photog_ed]

Over the years, I have read on many occasions that our shutters are less accurate than we would like to believe, especially older shutters.

A while back, I purchased the Shutter Speed app for my Android devices. This is the one created by Lukas Fritz (http://www.photoplug.de). I first used it without his PhotoPlug accessory and was very surprised that most of the tested shutters were quite accurate (assuming, as I do, that the app is accurate). Below 1/125, most of them are almost right on.

I was worried that using the app without the PhotoPlug was leading me to bad results, even though my results were repeatable.

So, I purchased the PhotoPlug and it sat around for months and months. This weekend I pulled it out and started making some measurements with my Nexus 7 tablet. I tested older style (jagged edge shutter speed ring), silver block style, and all black style Copal shutters. It did not seem to matter how old they were. I also tested versions used by Rodenstock, Nikon, and Fuji. Again, it did not matter which ones I chose. They are all mostly right on target. Of course, the higher speeds (above 1/60) are all off. However, they are not off by much. Usually, 1/125 and 1/250 are off by 1/3 or 2/3 of a stop. The top speed, 1/500 or 1/400, are, as expected, off by much more. Usually 1 1/3 stop.

All but one of my lenses are used. I have not yet tested the one I purchased new. None have been serviced since purchased by me! The one I sent to Carol, she sent back after making a minor fix for free!!

Based on what I have read, I expected more deviation from the marked speed, but I found that with the PhotoPlug the shutters were more accurate than I expected after using the app without it. So, I am wondering what others have experienced. Am I lucky in the shutters I have tested so far? Or is this a typical experience?

I have yet to test some Seikos, an Ilex, and some Wollensak shutters. We will see how that goes.

I bought Lucas's app and photocell and tested all my shutters a couple of months ago. I got a similar result; they were all very accurate at 1/30 and slower. Performance degraded at higher speeds, usually about one stop slow at 1/400. They are all Copal shutters, and just as I received them, no CLA on any of them. A total of six shutters, I think.

Ed

[Leszek Vogt]

Most of my shutters seem OK....till they were properly tested. One Ilex didn't work at all (I bought the lens at v. reasonable), couple of them seemed slow at 1/2 and 1 second. Decided to CLA (all 6), so I wouldn't have to look over my shoulder. Only one Copal seem perfect, so I didn't send it. Each person will likely will have a different story.

Les

[morecfm]

I use the app with the photoplug. I don't think in terms of speeds being off by x/y stops but rather matching a real speed vs. an indicated speed. That way I just meter to the real speed of the shutter and don't bother with the additional math after to make adjustments to the marked speed. I make a chart for each lens. I also do this on my mechanical medium format leaf shutter lenses.

[David Karp]

I am making charts for each lens in Excel. I will put them on labels or tape them onto the lensboard with each lens. I may also make note cards and laminate them.

[drew.saunders]

I am making charts for each lens in Excel. I will put them on labels or tape them onto the lensboard with each lens. I may also make note cards and laminate them.

I have a Calumet shutter tester (bought new long ago), and used it to test every lens and create a cheat sheet that's taped to each lensboard with the shutter speed corrections. I also have the bellows extensions corrections for each lens as well on the sheet (i.e.: infinity at Xmm, 1/3 stop at Ymm, etc.), all of these created in Excel. I made a master sheet that I printed small and carry with my grey card (which came with a nice ziplock plastic bag, so it stays dry and easily found without needing lamination).

[David Karp]

Drew,

Did you find your shutters were accurate? Inaccurate?

[pjd]

I use that shutter speed tester app with the photo diode widget, it showed that an Ilex Universal I've got is fast at slow speeds and the top speeds are all 1/25th. That Ilex is the shutter type that SK Grimes won't attempt to repair according to their site, by the way.

I've got a few old dial set compurs that test consistently accurately. Also a couple that are less reliable, and a couple of outright dogs. I use a couple of Volute shutters one of which has speeds a bit slower than the dial speeds, but consistently so. The other Volute runs nicely most of the time but is a bit awkward in cold weather.

Compound shutters are generally pretty good in my experience, usually a bit fast at 1 second, accurate through most of the mid range and then a bit slower than claimed at top speed.

[IanG]

I tested quite a few of my lenses 3 or 4 years ago and found all the shutter speeds on Copals, Compurs and Compounds were well in tolerance with little deviation less than 10%, not enough to compensate for. What surprised me most was the smoothest were the early Dialset Compurs and Compounds. I was most interested in the shutter speed range I was typically using 1 second to 1/30, although I do use faster when working hand held.

The Ilex and Wollensak shutters showed much greater variability while just in tolerance which is +/- 1/3 stop they aren't as accurate.

Ian

[drew.saunders]

Drew,

Did you find your shutters were accurate? Inaccurate?

None are accurate at the highest speeds, even the ones I bought new within the past 10 years, and most are pretty good in the middle and lower speeds. I have one that's 1 stop off on all speeds, which works out just fine!

[Ken Lee]

I never tested at high speeds but a few years back, using the Calumet tester and the aforementioned app, found them all much too close to repair.

One of my old un-coated Tessars eventually (after 80 years) needed CLA because it refused to operate at low speeds. After a trip to SK Grimes it's good as new. I'd be surprised if my smart phone will work as well in the year 2100 :rolleyes:

[Kevin Crisp]

I have one of the Calumet testers. I also wondered when I bought it how accurate it was. I tested it on an electronically controlled shutter (I think it was on an F3, I had to fool the camera into thinking the back was closed). It was accurate, dead on, to many places. I stopped worrying about it after that.

[Bob Salomon]

I have one of the Calumet testers. I also wondered when I bought it how accurate it was. I tested it on an electronically controlled shutter (I think it was on an F3, I had to fool the camera into thinking the back was closed). It was accurate, dead on, to many places. I stopped worrying about it after that.

Everything has tolerances. The shutter speeds have tolerances, so does your shutter tester. The test for the photographer is the results on film. Nothing else matters. If you are happy with your exposures then everything is fine. If not, have your equipment checked.
And the camera technician that tests that shutter does not use a consumer grade shutter tester.

[neil poulsen]

Years ago, I purchased Calumet's shutter tester for about $90 or so. It was an excellent purchase. After purchasing a new lens (new to me), I use this tester and take six readings on each speed and load the data into an Excel spreadsheet.

Using a pretty straight-forward logarithmic calculation, I average and determine the discrepancy between actual speed and nominal speed in stops, record this on a 4x6 index card, and make corrections in the aperture at the time of the exposure.

Since I can always correct for inaccuracy, my concern is whether or not a shutter speed is consistent. For consistency, I look at the range, which is range = max - min. (Also in stops.) Many times, the difference between the fastest and slowest reading (among the six that I take) on a given speed is within 1/10th stop.

If a shutter speed isn't consistent, I don't use that shutter speed. My experience is that Copal shutter speeds are surprisingly consistent. However, I recently purchased a Nikon 450mm Q lens where the fastest speed isn't usable. An unusable speed is rare.

Here's a copy of the card that I carry. These stop deltas are differences between actual an nominal speeds. The 450mm lens shown on the card is different from the Q above. Otherwise, the 125th sec speed would have been displayed as "na".

[Dan Fromm]

Neil, thanks for mentioning speeds' consistency. I've been following this thread and feeling sorry for myself because not all of my shutters' speeds are real, real close to nominal. I went back to the latest spreadsheet and found that nearly all are very consistent. Small comfort, perhaps, but comfort.

Oh, yeah, I use a Calumet tester.

[Jason Greenberg Motamedi]

I also have the Calumet shutter tester, and have tested all of the shutters I have had over the past 15-20 years. I would say that most (say 75 percent of 125 shutters) were within a 1/6 of a stop at speeds under 1/30. Faster speeds were rarely accurate, and were often 1/2 to a full stop too slow. Like Ian, I never really noticed much difference between Copal, Compur and Compound shutters, but Ilex shutters were more likely to be slow. I don't think was a function of age since many of my Compur or Compound shutters were as old.

As an addendum, I have found that almost all of my shutter speeds are consistent. When they are not I figure it is time for a CLA.

[David Karp]

Thanks for the information everyone. It is interesting.

I tested my Ilex. At the lower speeds it is with in 1/3 of a stop. At 1/50 it is right on. Then it goes off again. I tested some other Copals of various vintage and the results were similar to my initial results reported above. I guess I am lucky when it comes to used shutters!

[LabRat]

Thanks for the information everyone. It is interesting.

I tested my Ilex. At the lower speeds it is with in 1/3 of a stop. At 1/50 it is right on. Then it goes off again. I tested some other Copals of various vintage and the results were similar to my initial results reported above. I guess I am lucky when it comes to used shutters!

But one trick would be to cool your shutters in the fridge and test again... Or a really hot day and see what you get... You know that under the worst conditions the shutter might act up... ;-*

Steve K

[IanG]

Neil Poulsen's chart indicate what many expect the fastest shutter speeds to be slow, of course he's not indicated what lens/shutters but thats not particularly important.

The very nature of the way a leaf shutter opens and closes affects the exposure accuracy of the highest speeds due to it effectively being an opening and closing aperture, so as Bob Salomon says it's the results in practice with film that are more important than the electronically tested, So a leaf shutter at 1/500 or even 1/1000 is less accurate in terms of actual exposure wide open (full aperture) compared to stopped right down.

So in practice those slow top speeds might be compensation and actually giving correct exposure. It's also why most larger leaf shutters have relatively slow fast speeds.

Ian

[neil poulsen]

Neil Poulsen's chart indicate what many expect the fastest shutter speeds to be slow, of course he's not indicated what lens/shutters but thats not particularly important. . .

The 48mm, 58mm, 75mm, and 90mm are Copal 0 shutters. The 105mm is a Fuji Seiko shutter. The 120mm, 121mm, and 150mm are Copal 0 shutters. The 180mm is a Copal 1 shutter. The 250mm (f6.7) is a Fuji Seiko shutter. And, the 355mm and 450mm are Copal 3 shutters.


. . . Bob Salomon says it's the results in practice with film that are more important than the electronically tested, . . .

This is definitely a good point to consider. (I know that its occurred to me.)

But practically speaking, how often do those faster speeds actually get used in VC photography, and then at larger apertures? For me, not at all. I rarely photograph much above 1/60th sec, if even at 1/60th sec, and just about always at apertures smaller than f16.

For me, my card works in practice. But thinking about it, what hasn't occurred to me to consider is, at what aperture (as the shutter blades open) does my Calumet shutter tester actually begin to record time? Hmm. Not sure on that one.

And then, as the shutter closes, at what aperture does the timer stop recording time? It's not clear to me that those two apertures are necessarily the same.

It's also worth thinking about, at what apertures do I do my film speed tests? Again, Hmm.

Oh dear!

[Doremus Scudder]

Neil,

Your chart is interesting, but do you really adjust your aperture 0.08-stop? :)

I test my shutters in a similar way to Neil, but round off the speeds to the nearest 1/3 stop; that's close enough for me. I fire the shutter when testing at f/22 (my most common aperture, although I'm not sure it makes a real difference with the tester) and note the millisecond times for six or so firings at the same speed. I'll dry-fire the shutter once before testing to imitate my field practice. I then figure the speeds for each setting, rounding off to the nearest 1/3 stop and print these out on a sticker which goes onto the lensboard. I indicate the 1/3-stop variances from standard speeds with + or - signs. I shoot black-and-white only, so I'm more than happy to be within 1/3 stop with shutter speeds. The only instance where this practice is a bit inconvenient is with one Supramatic shutter I have, for which the 1-second setting is almost 2/3-stop slow, and the 1/2-second setting is 1/3-stop too fast, which results in a 2-stop spread between the two settings. Using speeds in this range sometimes forces me to use a less-than-optimum aperture. Otherwise, I feel confident with the consistency of my exposures.

Doremus

[neil poulsen]

Neil,

Your chart is interesting, but do you really adjust your aperture 0.08-stop? :)

Hi Doremus,

I don't begin to worry until the correction is up to about 0.3 stops. But, I do like being able to distinguish 1/10th stops. For example, 0.5 stops is easy enough to find on a shutter, versus 1/3rd or 2/3rd.

When testing, I've always left my lens wide open. But even then, it's possible the tester doesn't being to count time until the blades reach f8, f16, f22, etc? I don't know now, but it's easy enough to check.

The optimum would be for it to begin timing right away, after the shutter is released. I'm thinking that the relevant detail isn't so much the f-stop setting while testing as the brightness of the light that I use. The brighter the light, the quicker the tester begins to count time.

Years ago, I made a jig for taking readings that's shown in the photos with a lens mounted in a Copal 3 shutter. It helps illustrate the point. Perhaps I should use a quartz-halogen lamp on top, or something stronger? As I say, the sooner the tester begins to record time, the better.

[David Karp]

So, I am done testing all of my lenses.

Interestingly, I did not have a single lens that was wildly out of whack. All are easily usable.

The champs are a 150mm Fujinon NW in an all black Copal 0 shutter (It is good at all speeds except 1/500. I did not believe it, but that was repeatable) and a 210mm Caltar II-E in an all black Copal 1 shutter (It too was good all the way until I hit 1/400).

Interestingly, I have two shutters that I purchased new. One was on the aforementioned Caltar II-E, which is the only lens I ever purchased new (16 years ago). The other is a Nikon branded Copal shutter that was ordered to use with a 75mm f/4.5 Grandagon-N that wa tsaken off of a DB shutter mount. That one is easily usable, but half of the speeds are off +1/3 or -1/3 stop.

Most of my Fuji Copals are pretty darn accurate (assuming the tester is accurate). The Graphex shutter on my 135mm Optar is surprisingly accurate. The Seiko shutters on a couple of old Fujinons are good, but most speeds are off by 1/3 stop in one direction or another.

So, my takeaway is that all of my shutters are usable and really don't need to be serviced. Age does not really seem to matter. Silver notched ring Copals and jaggy edged silver ring Copals are among my more accurate. Another takeaway from my results and reading about the experience of others is that there is no rhyme or reason to it. If you start out with a good shutter and are good to it, it should be pretty good to you in return. If you purchase a lens used, you might or might not get a good one, but don't assume it will need a servicing. Short of the two lenses mentioned above, I have a number of used lenses that are really very accurate at the speeds I normally use. That little PhotoPlug and the Shutter Speed app were well worth the investment.