Speed accuracy?

[Calamity Jane]

Over the past couple days I have tested all the shutters I own using a photovoltaic cell and my digital storage scope. The scope is a very accurate time base and captures a time v.s. light intensity trace.

I was shocked at the results!

Even in my best, recently services shutter showed speed errors of up to 25% at the high speeds. Some shutters had errors up to 50% !!!!

Of the 5 shutters I tested, all were consistent in their speed and varied less than 5% from one exposure to the next. Shutters with low speeds (slower than 1/5 Second) usually were the most accurate at slow speeds.

I was surprised (and somewhat appauled) to find shutter speeds were so much different than the set speed, even on the recently serviced shutter. My faith in mechanical shutters has been shaken! I look forward to getting the Compur electronic to see if it is more accurate.

Has anybody else tested a number of shutters? If so, what were the results?

I did note the set speed v.s. the actual speed in my notebook for each shutter. Maybe I should put this info on a sticky label and attach it to each lens board.....

[Nick_3536]

Or do a film test? I've always thought some of the griping about film speeds really was related to equipment and not the film.

OTOH 25% sounds like a lot but it takes 100% to be off one stop.

[Brian Ellis]

I've tested mine several times using the Metrolux II darkroom controller. My experience was similar to yours, most speeds were off by a third to a half stop. Oddly enough, the Copal shutters on lenses I bought used were more accurate than the Copal shutters on lenses I bought new. I kept a list of the actual speeds with each shutter and used it for a while but eventually got tired of consulting it every time I made a photograph. A third to a half stop isn't a real problem with the b&w film I use. The most important thing is consistency I think.

[Ernest Purdum]

I think your results were fairly typical. Of course, if you had just purchased all your shutters new, you would be entitled to better results.

One factor that is usually ignored because there isn't all that mmuch to do about it is shutter efficiency. At f32, you're getting a lot more effective exposure than at f5.6, because at the latter opening the shutter blades are partially covering the opening for a longer period of time. Similarly, at top speed, the shutter blades are only partially open essentially all the time, while at 1/2 second the blade movement time is trivial.

As Nick pointed out, these problems are reduced because of doubling and halving from one stop to another. Your approach of keeping note of actual exposure time is a good one.

[Bobby Sandstrom]

From what I remember reading somewhere, shutter speeds vary with temperature so even if you write the difference on the lensboard it can change! I believe it was an article on the futility of trying to be overly precise with regard to Personal EI. I stress OVERLY as I realize you need to be centered somewhere close.

[Robert A. Zeichner]

The speed accuracy of mechanical shutters is never perfect in the LF world. It's quite common to see speeds marked 1/400th sec actually measure at 1/165th sec for example. Those high speeds are almost never used when stopping down to f22 or even smaller apertures as we commonly do. The slow speeds on the other hand are ones we use a lot and need to be known and consistant. I usually measure all my shutters before a big trip. The procedure I use is to first exercise the shutter at the speed I am measuring 2 or 3 times. Then I make 3 measurements and average. Each measurement should not vary by much (+/- 5%). If the measurements are all over the place, you've got trouble. When you're done, make a little correction chart on a label and stick to the lens board. Note that on many shutters there is a crossover point below which slower speeds are a bit faster than they should be and above which faster speeds are slower. This is normal. When you take your measurements, do so at an aperture that is typical for the type of work you do. Particularly at faster speeds, a factor known as transit time enters into the picture. This is time it takes for the shutter blades to fully open, then close again. Obviously at faster speeds, that transit time, while pretty quick, can add into the total exposure time and slow it down more significantly than at slow speeds. I hope this is of help to you.

[d.s.]

I put a piece of tape on my lens barrels and marked the speeds as 250 = +30 % (250 is overexposing by 30%). I only note differences of 25% or greater. On one lens 400 = 200. Thats the worst one. I also exercise my shutters several times when I'm testing them and before I make the actual exposure.

dee

[Pete Watkins]

Metal expands and contracts with changes in temprature, and this will affect the accuracy of your shutters. Having read and enjoyed most of your earlier posts I'm amazed to read that you actually use shutters, does this mean that you have progressed above 6 ASA? ;-). Keep up the good work.

[Paul Fitzgerald]

Hi C.J.,

for Copal #1 tolerance is +/- 30%

for Rapax #1,2,3 tolerance is +/- 20%

for Synchro-Compur tolerances are +/- 15% slow speeds ; +/- 20% fast speeds

Those are from the factory service manuals. Absolute accuracy is not the worry, consistent repeatability is the worry.

Good luck with the electric shutter but it has tolerances too.

[windpointphoto]

How often do you use the high speeds? If you don't use them does it matter? I don't think I've used a speed over 1/15 on the 4x5 much less the 8x10.

[Paul Coppin]

At the risk of setting off another 150 post thread, why are you surprised at the results obtained using pre-WWII inexpensive technology? Until I mention the D word, you're playing with hardware designed in the 30's, in which the process is self-calibrated, less so the equipment. For most, LF remains a craft, rather than a technology, wherein the results are driven by the skill of the operator to understand and learn to use his tools, not the tools themselves. Exactly the same methodology you use in your workshop. As has been mentioned, repeatibility is the holy grail. The numbers on the dials really mean squat. Your job is to find the right combination of click time over the right sized hole that works in your process to give you the end result you want. LF is a craft of first principles, not technology (unless of course you have enough money to afford the D word :).

As has been pointed out, even on new lens-shutters, the manufacturers generally quote +/- 25-30%, because they're still building 1930s shutters.

[Paul Coppin]

I did note the set speed v.s. the actual speed in my notebook for each shutter. Maybe I should put this info on a sticky label and attach it to each lens board....

You're on the right track. I made up a table of corrections for various things, printed it out in a text size I can just make out and laminated it into a small card that clips to the tripod for ready reference when I'm setting up a shot.

[Calamity Jane]

Paul: Thru my 56 years I have worked ALOT with antique machinery. I have a great appreciation for the craftsmanship, quality of materials, and ingenuity of old designs. Though I have restored/rebuilt many antiques, from steam engines to clocks, I had not worked on shutters until recently.

Certainly antique clocks are capeable of delivering good accuracy and I guess I was expecting shutters to be of similar quality. Non-pneumatic shutters certainly do well on slow speeds, there more sophisticated timing delays are used. Pneumatic shutters, of course, are effected by wear of the pneumatic delay cylinders, so I expected to find some shift with age.

It was while working on a Wollesak pneumatic from about 1901 that I began to understand the higher speeds are controlled by spring tension v.s. inerta and are therefore subject to changes due to wear, fatigue, dirt, and numerous other factors.

I posed the original question simply to find out if my group of shutters was typical or if they all needed work. It was not ment as a criticism of old shutters or their manufacturers. I have always been in awe of the complexity and design of shutters!

I refuse to even mention the "D" word!

[Paul Coppin]

And I wasn't trying to be critical in my response either. Its just that in this age there is a tendency to expect higher precision from the tool and less from the tool user... Within the scope of practice of "ordinary" LF, there is enough controllable variabilility in the process to accommodate the variability of the shutter, and it was pretty much always so... Clock precision is another matter. There was a necessity for a high degree of accuracy from the beginning, and so the mechanics evolved to reflect that accuracy. Shutters are not clocks, they are short timing devices that need to be far more robust than clocks, in order to drive a mechanism, that in and of itself has significant restraints due to a need for compactness and affordability. An interesting parallel would be to measure the variability in the precision of modest small mechanical alarm clocks - how precise do they work when you set them to go off at 8 o'clock...:) Wouldn't surprise me that you would see similar variability. Again, consistency is the higher goal in shutters, rather that accuracy. Accuracy is nice, but at what cost? In the trade-off, consistency carries much more weight than accuracy. Like when you're running your lathe (unless you are blessed with a multi-stage NC mill:), it matters little that the speeds it turns are precise, only that they are adequate for the job, and repeatable, for consistent performance.

[Neal Shields]

Last time that I had S K Grimes service a shutter it came back dead on at all speeds.

I checked against my tester and against a 35mm camera with an electronic microprocessor controled shutter.

However, I keep a plastic lamanited card in my camera case for the rest of my lenses with marked and actual.

Note if you ever take an old shutter apart you will realize why some speeds can be slow and others fast. There are two gear mechanisms and a high speed spring. At normal speeds you use one gear drive and the main spring, at slow speeds an additional retarding gear train is shifted in, and at the highest speed there is an additional spring engaged.

On many shutters nothng but the front cover holds the high speed spring in place and it loves to jump out and run off and hide in dark corners.

[Bill_1856]

Mein Gott, you're 56 years old? How did you get...? Oh, never mind.

[Calamity Jane]

"Bill", as in "Wild Bill" ?? !! I thought you waz dead!

Guess, in a way, I'm older than 56, seein as how I was born in 1852 the first time ;-)

[Michael S. Briggs]

Understanding shutter efficiency is important in measuring fast shutter speeds. Calamity, specifically what time did you measure? The total time opening time, from when the shutter blades just admit some light to just before they block all light, is not the correct time to measure. You want to allow for shutter efficiency, which allows for the transient time of the blades. If the lens is set at f16, as the blades just start opening, they will admit light at f64, so you have to treat that delta-t of the blades at that position as counting for 1/16 of delta-t of exposure time.





Chapter V of the Handbook of Photography by Henney and Dudley (1939) explains this with figures.





Simpliest is to realize that LF photographers rarely use the fastest speeds.





You did find your shutters to be very consistent.





Previous discusssions:
http://largeformatphotography.info/lfforum/topic/287113.html

and http://largeformatphotography.info/lfforum/topic/113993.html.

[Calamity Jane]

Thanks for the reference to those old threads Michael - VERY interesting.

Indeed, I was measuring the time from when the light started to pass until the light was cut off which, as you state, is not accurate. I don't know the proper measurement but it seems logical that the exact exposure would be "the area under the curve". The waveshape, at high speeds, begins to look like an inverted V, where the shutter transition times are a large part (most) of the waveshape.

When I have time, I should go back and look at the area under the curve - it is likely that the shutters are more acccurate than my first impression.

(When it comes to "digital capture device", I LOVE my digital storage scope!)

[John Cook]

Jane, I believe your results are typical. The instruction sheet for Ethol T.E.C. begins with the statement: “Photography is not an exact science”.

The actual exposure is, of course, further distorted by the slop in the f-stop lever. The diameter of the aperture is often quite different between opening up to a setting versus closing down to that setting.

Mechanical shutters also get slow and sticky with age and inactivity. Including those in a Hasselblad. Part of my first job as assistant/studio manager in Hollywood in the old days was to get out all the lenses and work the shutters every Friday afternoon.

At the risk of sounding like a heretic, this is why I do my film exposure/development tests (as much as possible) on the EOS with the electronically-set shutter and f-stops. Try exposing a gray card over seven stops in perfect third-stop intervals on a view camera.

Also why I shy away from emulsions which do not come in both 35mm and sheet sizes.

But don't tell anybody my secret...