Rolleiflex Lazy at Faster Speeds??

[Zarraboy]

I have just received one of those homemade Audacity shutter testers, bought to test a Pentax 67 which i found had a faulty shutter. Out of curiosity, i grabbed my best Rolleiflex (which happens to be a 3.5E Planar) to test it too. I did a very detailed testing and found all speeds to be accurate at f11, except for 1/250 (found it to be at 1/180) and 1/500 (found it to be 1/300). Isn´t it weird that all the slower speeds are spot on and these two are lazy? Should i try a new batch of tests at another aperture? Do i need a CLA?

 

[Slixtiesix]

It is not uncommon that the 1/500 is in reality only 1/300. Even after the cameras had a CLA. I do not think that the aperture does play a role here. I would not bother about it.

Best, Benjamin

 

[bsdunek]

It is not uncommon that the 1/500 is in reality only 1/300. Even after the cameras had a CLA. I do not think that the aperture does play a role here. I would not bother about it.

Best, Benjamin

Right, not uncommon at all. By the old ASA spec. 1/500 can be off by 50%, so it could be as slow as 1/250. Most mechanical leaf shutters are slower at higher speeds. The few that had 1/1000 were lucky to make 1/750. Doesn't help the motion stopping but gives you rich negatives!

 

[Steve Smith]

I built a shutter speed tester a while ago and tested everything I owned. My findings with in lens shutters were similar. Higher than 1/250 and the marked speed is a bit optimistic!


Steve.

 

[Zarraboy]

I am glad to hear that. I knew the speed could be "off by default" but i did not think by such a large margin. Not bad for a 40-years old camera! Now something that comes to mind: When doing EI zone system tests should i take this into account or should i simply do the tests as if i was actually testing 1/500?

 

[Jeff Kubach]

I'm lazy at those speeds also!

Jeff

 

[elekm]

Even if a shutter is off by 25%, that still roughly 1/2 stop, I think.

I think the variance in shutter also applies to shutters in all-manual cameras, including rangefinders and SLRs. I read about the variance years and years ago. Does anyone remember if camera tests showed how close the actual shutter speed was to the marked speed?

 

[Q.G.]

Yes. Camera tests (obviously not all are the same) did show the shutter accuracy at all speeds.
Shutters aren't as accurate at fast speeds as at slower ones, but deviations as large as the ones reported here are not (!) normal. Leaf and focal plane shutters alike should be off no more than about 10%.

It's important to measure the shutter speed at the right distance behind te shutter, i.e. at the focal plane.
And - to be able to take efficiency into account - to know when the metering device is 'triggered', how low the light level is at which it says the shutter is open. You do not want to measure, for instance, the time lapsed between the moment the shutter first begins to move and the moment it has come to a full rest again, but the effective exposure the shutter provides.

And, of course, you have to know the efficiency of your tester and testing procedure. 'Faults' there will show most at short speeds, so make sure that what you are coming across is a long shutterspeed, and not something introduced by your metering device and/or method.

 

[piu58]

Don't forget: Shutter speed at the fast end depends on the aperture setting! The shutter opens and thereafter closes immediately. This is what you see at f/4. If you use f/22 the shutter opens, stays open (in reality it opens and closes the outer part of the lens) and then closes. Quite different.

 

[jscott]

I've tested my collection of 17 Rolleicords and 16 Rolleiflexes (all but two bought from ePay) in some detail, with discouraging results. Only three cameras (one Art Deco, one Rollei Std, one Rollei T) had measured speeds within 10% of normal. All others, including five Automats and a 2.8C were too slow to be useable, running an average of 180% of normal speed. I've sent the three best cameras off for CLA's.

I was a bit shocked because I paid market rate for one of these Rollei Automats at a professional camera store (Kenmore Camera) and it turned out to be 50% slow! Apparently they had never even tested or cleaned it; they just sold for a tidy sum with unusably slow shutter speeds. Nice looking camera--but I didn't buy it to look at it.

I used a homemade Radio Shack phototransistor tester, and used Audacity to measure speeds. I did ten shutter tests for each speed on each shutter, then randomly took 4 and averaged the results in Excel.

I attempted CLA's on some of the older rimset Compur shutters and although they speeded up slightly they still were 20-30% slow, maybe they need a new mainspring.

My conclusion is that buying older Rolleis from eBay and expecting them to perform is unrealistic. Chances are that nine out of ten of them will have serious slow shutter issues. Sadly, buying from a professional shop is not much better (and tends to cost a lot more).

 

[Ben 4]

Reasonably Accurate Results

I had to have the shutter of my Rollei MX serviced about 5 years ago (I suddenly lost all speeds above 1/25). I sent it to Paul Ebel who diagnosed a faulty main spring, which he repaired. He returned the camera with a test sheet showing the results of his shutter tester: 1/500 tested as 0.0022 seconds (about 1/450), and 1/250 tested as 0.0035 seconds (about 1/300). All the other speeds tested within 10% of what they should have been.

 

[outwest]

It doesn't really matter what the shutter speeds are as long as they are consistent and you know what they are. One of the joys of older mechanical cameras.

 

[Diapositivo]

I think it was normal for mechanical shutters to consider 20% mistake an acceptable one. That would be a bit less than 1/4 EV. When buying second hand, the best would be buying from those repairers who make a business of buying broken old cameras and repair and cannibalise them.

I bought a Nikon FE2 second-hand from such one repairer on eBay. He gave me the sheet of all actual shutter speeds. Seeing how serious his work was, I also brought him four Minolta bodies (SRT 100x, SRT 101b, X-700 and XM) and he did the cleaning and gave me a sheet with actual speed for each.

When properly serviced, discrepancies are quite contained. I am surprised by the X-700 which at 1/1000 (maximum speed) overexposes by 36%. (1/3 EV I presume). The second worse time is 1/500, 9% mistake.

The Minolta XM (bought very old and used, probably by a professional) has the maximum mistakes at 1/500 (15%) and 1/1000 (14%). That's a very small mistake anyway, probably 1/6 of an EV.

My Nikon FE2 has the maximum mistakes in 1/250 (11,25%) and 1/4000 (10%).

My Minolta SRT 100x has the maximum mistakes at 1/30 (23%) and 1/8 (12%). Curiously, the fast times (1/1000, 1/500, 1/250) are extremely accurate (mistakes resp. 1,5%, 0,5%, 3%).

My Minolta SRT 101b has the worst three times at 1/30, 1/8, 1/15 (resp. 20%, 20%, 11,6%). Again fast times very accurate.

The tests were carried on with an oscilloscope by a person who has a laboratory for medical appliances maintenance.

General conclusion: A mistake within 20% is generally speaking quite acceptable, you are not going to suffer it at all shutter speeds.

Diaphragms and film speed also introduce discrepancies between theoretical exposure and final effect.

As far as I know, modern film, also "unprofessional" one, is produced to a 1/6 EV tolerance or so. Your shutter might give you a mistake of another 1/6 EV and your diaphragm maybe the same. So if you are unlucky, and all three mistakes have the same "direction", you end up with 1/2 EV, which is certainly noticeable on a slide.

Nonetheless, using auto exposure certainly can cause a much higher mistake

Fabrizio

EDIT So if one uses negative film, exposing for theoretical ISO will place the mistake by a slow central shutter in the "safety zone" and all is fine. If one uses slide film, I would keep in mind the real times where there is a big discrepancy and have also the diaphragm checked.

 

[BrianShaw]

I thought the tolerance for shutter speed was generally greater than 10%. Graflex was +/- 20% per their official repair manual.

 

[jp498]

On my rolleiflex automat mx, the 1/250 and 1/500 feel like they are on a different mechanism or spring or something and you feel it and listen to it as you turn up the shutter speed. The instruction manual says 1/500 is indeed different and you should not put into 1/500 while it's cocked, which happens automatically when you wind the film. Unless you had a recent high quality CLA, such speed variations are common, it's probably a very old camera right now. Even after a CLA, it's apt to be off a little but should be consistent and calculated. The shutter speeds are variable as well, so I think perhaps the dial could be off a little too.

 

[JPD]

On my rolleiflex automat mx, the 1/250 and 1/500 feel like they are on a different mechanism or spring or something and you feel it and listen to it as you turn up the shutter speed. The instruction manual says 1/500 is indeed different and you should not put into 1/500 while it's cocked, which happens automatically when you wind the film.

It has the Compur-Rapid, or the early Synchro-Compur with a booster spring for the fastest speed. The Rolleiflex after that, the 3,5 B (MX-EVS) has the newer Synchro-Compur on which you can set 1/500 when the shutter is cocked.

 

[Ian C]

Compensating for a Slow Shutter Speed

The OP stated a 1/300 second shutter time at a marked time of 1/500 second.

That’s an increase of 0.74 stops. Most of us could accept that on negative film.

For transparency film, the user could compensate by closing the aperture by about 0.7 stops.

A table of exposure discrepancies in f-stops each time found significantly off could be kept in a small notebook kept in the camera bag or a pocket for critical applications. Then the correction could be made by altering the aperture by the amount needed.