Shutter speeds on Agfa Isolette III. Calculating error in stops

AndyNC · Dec 30, 2022

Hi,
I've posted this here because it seems the most appropriate.
I have an Agfa Isolette III which I have stripped and cleaned and replaced the bellows. I do need to align the rangefinder because the images are shifted vertically.

I've made myself a little device to measure shutter speed and, of course they are not all good.
But my real question is how I calculate the approximate stop error from the timings.
So here's a table of speeds:
Dial Measured My opinion My error calculation
1 1/0.99 OK Good
1/2 1/1.7 OK 1.12 times the exposure
1/5 1/4.8 OK 1.04 times the exposure
1/10 1/12 OK 1.15 times the exposure
1/25 1/17 Not good 1.5 times the exposure
1/50 1/31 Not good 1.59 times the exposure
1/100 1/67 Not good 1.48 times the exposure
1/300 1/170 Not good 1.76 times the exposure

So my estimate is that say in case for 1/300 the exposure is 1.76 time the expected so approx 0.76 stops more.
Am I correct in this.

I know that all the these values will not make a great deal of different to the exposure and so are not a problem, I'm just getting a base line of performance. I would only use this for B&W

Kind regards
Andy

gone · Dec 30, 2022

Your shutter is "normal" for an old folder, as it appears its essentially spot on until you get to 1/25 and up. I would just make it a stop and a half and be done w/ it rather than making tiny adjustments w/ the aperture.

But you're only halfway there, you need to verify that your meter is dead on. Maybe ck it against a known electronic camera? A lot also depends on what you are using for film and developers too. Tri-X in just about anything, you're good to go as it is or adding that 1.5 stop. My experience w/ handheld meters is that they can vary a lot depending on what exactly you're pointing it at, and at what distances.

My way around all this micro managing was to just sacrifice one roll of film and bracket all the shots. Keep notes, because you may prefer the look of the negs a little under or over exposed. There is no way to know until you shoot that one roll and examine things w/ a loupe on a light table.

AndyNC · Dec 30, 2022

Momus
Many thanks for your response.
Please note that when I said 1.5 times the exposure I think this means an extra 1/2 stop. That in general was what I was trying to find out in my convoluted way.

I agree in principle that there are many other variables so 1/2 a stop is not going to matter.

I have been using Fomopan 400 for two films and on came out a bit overexposed but could have been a bit overdeveloped that's why I checked the shutter speed,

The accuracy of the shutter measurement will be very good. I built a fast opto cct and hang an oscilloscope on the output to measure the timings. I checked it against a Pentax MX-5n and after I realised that the size of the measuring opening is very important ( as it's only a small slit of light that progresses across the film) I found it correlated very well with the dial speed. (didn't write anything down though)

Also I think using a film and bracketing and decide on the developer/times will be the best way forward.

Many thanks
Andy

MattKing · Dec 30, 2022

AndyNC said:
So my estimate is that say in case for 1/300 the exposure is 1.76 time the expected so approx 0.76 stops more.
Am I correct in this.

Not 0.76 stops.
You need to use square roots if you are going to calculate the difference in stops.
The square root of 1.76 is ~1.32 - so basically .32 (or 1/3) of a stop.

wiltw · Dec 30, 2022

About ratios vs. EV...

+1EV will DOUBLE the size of the aperture or make shutter speed TWICE as long.
-0.5EV will HALF the size of the aperture or make the shutter speed half as long

Doubling the size of the aperture changes it by NOT DOUBLE the number...f/4 admits twice the light as f/5.6
Doubling the shutter speed makes for half the amount of light...inverse! 1/500 admits half the light as 1/250

For aperture numbers, it is better to consult tables of fractional f/stops, and count steps

My calculations, compared to what you posted

1/2 1/1.7 OK 1.12 times the exposure...my calculation 1.18, adjust aperture by -0.25EV or just plain ignore the error!
1/5 1/4.8 OK 1.04 times the exposure...my calculation 1.04, adjust aperture by 0EV
1/10 1/12 OK 1.15 times the exposure...my calculation 0.83, adjust aperture by +0.25EV
1/25 1/17 Not good 1.5 times the exposure...my calculation 1.47, adjust aperture by -0.5EV
1/50 1/31 Not good 1.59 times the exposure...my calculation 1.61, adjust aperture by -0.66EV
1/100 1/67 Not good 1.48 times the exposure...my calculation 1.49, adjust aperture by -0.5EV
1/300 1/170 Not good 1.76 times the exposure...my calculation 1.76, aperture adjust by -0.75EV

MattKing · Dec 30, 2022

wiltw said:
1/5 1/4.8 OK 1.04 times the exposure...my calculation 1.04, adjust aperture by -1.0EV

Using this as an example, I think that you may be misinterpreting AndyNC's results.
I don't think he is saying that the difference between the set speed and measured speed is 1.04 x the set speed. I think that he is saying that the measured speed is 1.04 times the set speed, or so close to being the same as the set speed, as to not be of concern.
So, the necessary adjustment is 0.0 EV.

bernard_L · Dec 30, 2022

Not to be confrontational, but all numbers so far are incorrect.
+ 1 stop (or +1EV) is double the exposure. So far every body agrees (I think).
Assume from now on that time is the only variable

Time multiple (Tm)	EV error
1.0	0
2	+1
4	+2

and so on for powers of 2. What is the time multiple is not a power of 2 ??
EV_err = log2(Tm) =log10(Tm)/log10(2)
and to a good enough approximation, log10(2)=0.3
so EV_err = log10(Tm) / 0.3

Going back to the nominal 1/300 setting: Tm= 1.76
EV_err = 0.2455 /0.3 ≈ 0.82
EV correction is, of course, the opposite, i.e. -0.82, i.e. compensate that slow shutter by closing the diaohragm by almost one stop. But a negative film would not complain about one stop over...

Another worked out example for 1/2 setting
Tm = 2/1.7 = 1.18
EV_err = 0.23

albada · Dec 30, 2022

The formula @bernard_L posted is correct. An EV is a base-2 logarithm.
Here's another example using the 1/100 numbers:
log2(100/67) = log10(100/67) / 0.3 = 0.6
Thus, at 1/100 sec, the shutter is overexposing by 0.6 stops.

Mark

wiltw · Dec 30, 2022

MattKing said:
Using this as an example, I think that you may be misinterpreting AndyNC's results.
I don't think he is saying that the difference between the set speed and measured speed is 1.04 x the set speed. I think that he is saying that the measured speed is 1.04 times the set speed, or so close to being the same as the set speed, as to not be of concern.
So, the necessary adjustment is 0.0 EV.

I agree. I should have realized that, even without your post. <doh>
I altered the values in post #5 suitably to make the adjustments apparent.

Randy Stewart · Dec 31, 2022

While I like to slice the technical salami as thin as the next guy, maybe more than most, the reality here is that the faster speeds are running slow in a 75 year old camera, and an aperture closing down by a half stop is more than an adequate user adjustment for either B&W or color negative film. If the top speed is 1/300, then this is one of the lesser (Prontor) shutters offered in the Isolette, and it is probably not adjustable. It just needs a new spring set, and it isn't going to get one. You could pirate another shutter as a replacement, but is it likely to be any more accurate? Since it doesn't have a faster shutter, it doesn't have the Solinar lens, so you are rocking one of the triplets, and therefore you've done all that the value of the camera could justify.

AndyNC · Dec 31, 2022

H All,

Thank you all for your responses.

I now understand more about how to calculate EV errors from shutter times. I've put the calcs into a spread sheet and now I have a feel for the shutter's accuracy.

Well done Randy. Yes it's a Prontor with an Apotar F4.5 85mm. New bellows and a good clean out of all the green "glue". The Range finder mechanism is dated 1952 which makes it an early one and 70 years old, a just few years (3) older than me.

The main reason for the test was to establish whether it was the camera or my developing that made the negs a bit overexposed. Now I know it's my developing.
Given this info I don't need to concern myself with the shutter speeds given the tolerance of B&W film.

Many thanks
Andy

bernard_L · Dec 31, 2022

A slight aside (is it an aside?) about old shutters being slow. Apart from the top speed, which is maybe the result of the marketing /advertising depatment making promises that engineering can't fulfil, the other speeds are also slow, typically by 1/2 EV. I've noted this on several old shutters (Zeiss, Retina...) despite using all available adjustments and repeatedly cleaning in naphta.

Could it be that the main spring is "tired"? Initially I was skeptical, since it is never deformed beyond the elastic regime. But I googled "spring creep". FYI, "creep" is permanent defromation occurring below the elastic limit.

Do Springs Lose Tension When Compressed? | IDC Spring

Many people rely on the operation of springs in their daily lives, both in their homes and businesses. Usually, the only time you notice these springs is when they break or are not working correctly. You may wonder, dosprings wear out from being compressed? What causes a spring to lose energy...

idcspring.com

Mechanical Spring Creep: Does it Matter? - Yost Superior

Does spring wire creep? Will your mechanical springs lose their elasticity when placed under stress for long periods of time?

www.yostsuperior.com

and a number of more academic texts.

So, yes, springs can lose tension over time, just from sitting there. Just about makes up the difference between ISO and ZoneSystem film ratings

AndyNC · Dec 31, 2022

Hi,
bernard: It's not an aside, it's very relevant. I think this shutter mechanism has two stages: 1sec to 1/10sec and 1/25 upwards. I don't remember where this comes from but it could be when I stripped it down last year. So the spring on the uppers speeds has possibly softened.
Andy

bernard_L · Jan 1, 2023

AndyNC said:
Hi,
bernard: It's not an aside, it's very relevant. I think this shutter mechanism has two stages: 1sec to 1/10sec and 1/25 upwards. I don't remember where this comes from but it could be when I stripped it down last year. So the spring on the uppers speeds has possibly softened.
Andy

Just a slight correction. Unless the Isolette III is different from most other cameras of that era, there is just one main spring that actuates the opening and closing of shutter blades. The different speeds are obtained by slowing down more or less that motion. Fast speeds are slowed down by having the main spring push aside a lever connected to a train of gears. For slow speeds, additionally, there is a so-called retard mechanism (sprocket wheel + rocker arm) that produces that audible "whizzz" sound.

For the same amount of main spring fatigue, possibly, the two types of timing mechanism are affected differently; and maybe they also wear down on their own. Plus, as already stated, the highest speed is there for the benefit of the spec sheet and advertisement leaflet.

cramej · Jan 2, 2023

I just found this resource in one of the SPT Journals on learncamerarepair.com. Should be helpful.

AndyNC · Jan 3, 2023

Magic!!

MattKing · Jan 3, 2023

So - how many interpretations can you think of for what is meant by, as an example, "65% slow"?
I can think of at least two.
The way that AndyNC set up his data is much more clear.

Shutter speeds on Agfa Isolette III. Calculating error in stops

AndyNC

gone

AndyNC

MattKing

Moderator

wiltw

MattKing

Moderator

bernard_L

albada

wiltw

Randy Stewart

AndyNC

bernard_L

Do Springs Lose Tension When Compressed? | IDC Spring

Mechanical Spring Creep: Does it Matter? - Yost Superior

AndyNC

bernard_L

cramej

AndyNC

MattKing

Moderator

Shutter speeds on Agfa Isolette III. Calculating error in stops

Moderator

Moderator

Moderator

Privacy & Transparency

Privacy & Transparency