Are "fast" speeds on old Compur/Prontor shutters always this... slow?

[ProfessorC1983]

So I recently built a DIY leaf shutter speed tester from a microprocessor, LED and photodiode. To my surprise it seems quite accurate within about 16 μs (so 64khz) as calibrated with a second microcontroller as timing source, so I'm fairly confident in the results below, at least within 1 order of magnitude.

I tested several 1950s era mechanical leaf shutters, all Prontor-SVs and Synchro-Compurs. While I was expecting the higher speeds to be less than perfectly accurate, I was surprised that in many cases, the top speed was slow by over 1 stop. The biggest offender (Prontor-SV #3) was on a Zeiss camera I recently bought in part because the seller claimed it was recently CLA'd... and it may well have been, since it's quite consistent, but the 1/300 fires at almost exactly 1/100! Definitely explains some persistent overexposure issues...

Just thought these results might be interesting to some folks. Enjoy!


Prontor-SV #1

"1/100": 12064, 12060, 12092, 11428, 11704, 11432, 12096, 12124, 12076, 12176
Avg: 11925 (1/84)

"1/300": 4792, 4524, 4388, 4300, 4320, 4520, 4624, 4820, 4856, 4852
Avg: 4600 (1/217)


Prontor-SV #2

"1/100": 14472, 14492, 14572, 14744, 14616, 14512, 14376, 13904, 13656, 13484
Avg: 14283 (1/70)

"1/300": 5548, 5584, 5604, 5600, 4684, 4672, 4704, 4596, 5636, 5612
Avg: 5224 (1/191)


Prontor-SV #3

"1/100": 14284, 15192, 15728, 15196, 14504, 16128, 15824, 15400, 15732, 14288
Avg: 15228 (1/66)

"1/300": 9196, 9112, 9320, 9156, 9040, 9124, 9068, 9312, 8984, 9028
Avg: 9134 (1/109)


Synchro-Compur #1

"1/100": 16512, 17348, 17472, 15084, 16096, 16908, 17044, 15844, 15784, 15852
Avg: 16394 (1/61)

"1/250": 8024, 8236, 7072, 7468, 8032, 8104, 8360, 7628, 8500, 7884
Avg: 7931 (1/126)

"1/500": 4520, 4576, 4496, 4840, 3568, 4860, 4044, 4380, 4156, 4680
Avg: 4412 (1/227)


Synchro-Compur #2

"1/100": 12732, 12874, 12680, 15012, 12836, 15156, 14568, 14388, 13220, 14460
Avg: 13793 (1/73)

"1/250": 7080, 6880, 7128, 7304, 7124, 6656, 6492, 7044, 6348, 7396
Avg: 6945 (1/144)

"1/500": 3876, 3956, 4004, 3908, 4256, 4052, 4448, 4084, 3984, 3924
Avg: 4049 (1/247)


Synchro-Compur #3

"1/100": 9340, 12384, 11304, 11252, 12104, 12600, 12332, 12396, 12164, 11220
Avg: 11710 (1/85)

"1/250": 7592, 6128, 7248, 6788, 7300, 7504, 7640, 7748, 6588, 7640
Avg: 7218 (1/139)

"1/500": 3520, 3392, 3256, 3864, 3868, 3908, 3900, 3528, 3464, 3844
Avg: 3654 (1/274)

 

[DWThomas]

Interesting! Of course, I'm older than they are and my top speed isn't what it used to be either! I'm not familiar with the specifics of those shutters, but I seem to recall that some have an additional booster spring for the highest shutter speed. If these are of that type, perhaps the booster spring is fatigued, as it likely operates under the highest general stress. Somewhere around here I have had the parts to build a shutter timer for several years, but have not done so. I suspect something in the recesses of my aging brain doesn't want to know!

Anyway, thanks for the data.

 

[ic-racer]

What aperture are you testing at. Are you taking into account the shutter efficiency?

 

[Ian Grant]

I'd like to see the slower speeds as well, I have two early pre-WWI (1912-14) Dial-set Compur shutters and slow speeds up to 1/30 are remarkably accurate and smooth.

Ian

 

[ProfessorC1983]

I should've mentioned that all tests were conducted at 2 stops down from wide open (which was f/8 for the Prontor-SVs in my Zeiss folders and TLR, and f/11 for the Synchro-Compur LF lenses) in order to semi-control for shutter efficiency. The LED was placed about 1" in front of the lens and the photodiode nearly touching the rear element, and as centered as I could get it by eyeballing. I just used a basic "helping hands" tool to position the sensors so there could've been drift between tests as I recocked and fired the shutter.

But, good point, and I may re-test at the smallest aperture with the photodiode closer to the "film plane" to see if that makes any difference.

 

[Arvee]

Henry, in the 2nd ed. of "Controls for B/W...", published similar numbers for the shutters in the Hasselblad and Nikon FE he used in his research. Very interested in the results from subsequent tests on your shutters. Please keep us informed of your progress.

 

[Dan Fromm]

OP, research "shutter efficiency."

 

[Donald Qualls]

Those shutters are 60-70 years old. They probably weren't as fast as claimed even when new (and that doesn't go only for leaf shutters; Contax put 1250 on the Contax II focal plane shutter, and I've read that they virtually never beat 1/750 even with a freshly rebuilt shutter, including springs). Speeds above about 1/200 were as much about marketing as actual exposure. Photographers who needed to use those speeds would calibrate their exposures and know that, for instance, they needed to stop down the aperture one stop for top speed.

As suggested above, there's also a likelihood that a booster spring, if present, has weakened over time, especially if the shutter was left for decades set to highest speed and cocked.

With shutter speed testing having gotten easy enough that many smart phones can do it with no other tools needed, it's probably easiest just to print a chart of actual speeds and keep it with the camera or lens...

 

[SalveSlog]

A point with leaf shutter speed testing that is important to me is that for the effective exposure, the shutter times will be different with a small and a large aperture. That's because the leafs takes time to move from just giving a small hole to open fully up. And the same when closing again. That means that if you need a wide open lens to get enough light while using the fastest shutter speed, you will only have enough light coming through the lens while the aperture is practically fully opened (by the shutter blades). So the film will be effectively exposed for a shorter time than if you have more intense light and the aperture only need to be small.

On my Tempor shutters this gives a very significant difference. The effective exposure at top speed with the aperture wide open is approximately matching the nominal EV. At strong light with the smallest aperture, effective exposure is close to one step more. This is because the Tempor shutter moves the blades slower than i.e. a Prontor or a Compur. But I also find a difference with those other shutters, only not so drastic.

 

[wiltw]

Keep in mind that those shutters are 50-70 years old. Mechanical shutters used OIL to keep the works moving properly, and OIL GETS GUMMY. That is why photographers accepted that they needed to send in their lenses with shutters or send in the entire camera body for a CLA Clean Lubricate Adjust. I rather doubt those lenses have visited a technician in a very long time -- if EVER! Mechanical shutters were like pocket watches or wrist watches...the needed periodic visits to the camera repair guy, just as watches went to the jeweler.

 

[cmacd123]

Keep in mind that those shutters are 50-70 years old. Mechanical shutters used OIL to keep the works moving properly, and OIL GETS GUMMY. That is why photographers accepted that they needed to send in their lenses with shutters or send in the entire camera body for a CLA Clean Lubricate Adjust. I rather doubt those lenses have visited a technician in a very long time -- if EVER! Mechanical shutters were like pocket watches or wrist watches...the needed periodic visits to the camera repair guy, just as watches went to the jeweler.

although molybdenum disulfide Powder is preferred to oil, which often will gum up a shutter before it is assembled.

 

[shutterfinger]

Two points not yet covered, the wear on the moving parts affects the speed even if the springs are good. A leaf shutter is a variable aperture and as such will close down to more than 2 stops below the selected EV rendering any light coming through the lens to be at such a low level that it will not affect overall exposure. The sensor at center of the shutter opening measures all the light coming through the shutter from the instant the blades move to allow light, however faint, to pass through until they close and prevent any light from coming through.

I test shutters with a light source and a IR led connected to a computer running Audicaty audio recording software. On rare occasion I have encountered a shutter whose top speed measured 1/2 stop slow wide open, sensor at center but the average is 1 stop slow with well worn shutters 2 stops slow.
https://www.flutotscamerarepair.com/Shutterspeed.htm The tolerance for speeds below 1/125 is 20%, 1/125 and higher 30%. One needs a 1/3 stop change in exposure to see the difference with film.

 

[Donald Qualls]

although molybdenum disulfide Powder is preferred to oil, which often will gum up a shutter before it is assembled.

Oil has two advantages over moly disulfide: it was available in 1950 (pretty sure MoS2 came out in the 1980s, certainly no earlier for civilian use than the 1970s), and it stays mostly where you put it, if you use the correct microdrop quantity. Correctly applied, the correct (very light) oil will do its job for decades without causing trouble -- as witness the number of 1920s vintage shutter that still mostly work when purchased on eBay in a "looks new" folding camera or plate camera.

 

[takilmaboxer]

With shutter speeds, consistency is more important that accuracy. You're better off with a shutter that is always 1/100 instead of the 1/250 reading on the speed ring than you'd be with one that is sometimes 1/250 and sometimes 1/100. I have several old folders that have been CLA'd and none of them reach 1/250. However they are consistent and that enables me to predict negative density quite well.

 

[RalphLambrecht]

What aperture are you testing at. Are you taking into account the shutter efficiency?

+1;easily overlooked!