EV fractions matter in practical terms.
When you decide you need a certain exposure, and you set your camera at say 1/125@f/11, the actual exposure will depend on three kind of errors, three differences between real value and expected value:
The error in the shutter speed;
The error in the diaphragm opening;
The error in the film speed.
Decades ago the film speed was not that precise. Your roll could easily be 1/3EV off and even more. Professionals used to buy film in big lots and test a film to determine real speed for that batch.
Diaphragm opening of 1/4 EV of error are pretty much normal, and also 1/3 EV is seen, on certain apertures, when the lens is new, supposing we don't use TTL metering, as when using a separate lightmeter, or we use TTL metering but not in "stop-down" mode so that the actual aperture is not measured. We should add that a lens can have lubrication problem, the blades can close a bit slower when expected when you press the shutter. That might make the error worse.
Shutter mistake, again, can add another 1/4, or 1/3 of an EV
If you are "lucky" those mistakes will offset each other and your final exposure will be very near as expected.
When you are unlucky, the three mistakes will be in the same direction, and summing them up you will end up with even a full EV stop of exposure mistake. Doesn't happen every day, but it could happen in the past and up to a smaller extent, in the present. If you have a random 50% of mistake in each direction, you have on average 1/8 of cases when the three mistake sum up in the same direction.
The practice of bracketing in difficult situations (or in professional productions) is probably borne out of this kind of camera mistake. If there was no mistake, in a studio situation nailing exposure down to minute fractions would be easy.
So everything that reduces exposure deviance is welcome, especially for the non-professional. It's true that a single 1/3EV in shutter mistake, in diaphragm mistake or in film speed mistake it is not noticeable, but when the three sum up against you, it can ruin your photo.
Nowadays films easily reach no more than 1/4 EV in "mistake". "Professional" films are probably around 1/6 EV (that was the guaranteed accuracy margin of Agfa professional film in the early Nineties). I suspect also non-professional film are more or less in that league.
The fact remains that most of us just use a separate light meter and then set the values on our camera in Manual mode, so quality and precision matter.
Sometimes I find my slides slightly off what I would have expected. I suspect it is diaphragm error (I mostly use manual mode). I would like to have an electronic device to test diaphragm precision, like those that exist for focal-plane shutters.
I had to discard yesterday an overexposed slide, I exposed it using an incident light meter, and it was overexposed (Canon Canonet). I presume it is camera mistake. It's probably half EV but it was enough to wash out too much details where there should have been.
Ultimately quality and precisions are welcome. Electronic controlled shutters are better, but we don't notice 95% of the time. Mechanical shutters are good enough, slides have always been taken with them, but in my photographer's life I saw some "surprises".
Many seem not to trust in-camera lightmeters. If the subject is not backlit or sidelit, if the subject is average tone, if there is no sky in the frame, a modern in-camera lightmeter should be just dead on. Light meters are the smaller source of exposure mistake in my opinion. User error in light metering is another matter.
When you decide you need a certain exposure, and you set your camera at say 1/125@f/11, the actual exposure will depend on three kind of errors, three differences between real value and expected value:
The error in the shutter speed;
The error in the diaphragm opening;
The error in the film speed.
Decades ago the film speed was not that precise. Your roll could easily be 1/3EV off and even more. Professionals used to buy film in big lots and test a film to determine real speed for that batch.
Diaphragm opening of 1/4 EV of error are pretty much normal, and also 1/3 EV is seen, on certain apertures, when the lens is new, supposing we don't use TTL metering, as when using a separate lightmeter, or we use TTL metering but not in "stop-down" mode so that the actual aperture is not measured. We should add that a lens can have lubrication problem, the blades can close a bit slower when expected when you press the shutter. That might make the error worse.
Shutter mistake, again, can add another 1/4, or 1/3 of an EV
If you are "lucky" those mistakes will offset each other and your final exposure will be very near as expected.
When you are unlucky, the three mistakes will be in the same direction, and summing them up you will end up with even a full EV stop of exposure mistake. Doesn't happen every day, but it could happen in the past and up to a smaller extent, in the present. If you have a random 50% of mistake in each direction, you have on average 1/8 of cases when the three mistake sum up in the same direction.
The practice of bracketing in difficult situations (or in professional productions) is probably borne out of this kind of camera mistake. If there was no mistake, in a studio situation nailing exposure down to minute fractions would be easy.
So everything that reduces exposure deviance is welcome, especially for the non-professional. It's true that a single 1/3EV in shutter mistake, in diaphragm mistake or in film speed mistake it is not noticeable, but when the three sum up against you, it can ruin your photo.
Nowadays films easily reach no more than 1/4 EV in "mistake". "Professional" films are probably around 1/6 EV (that was the guaranteed accuracy margin of Agfa professional film in the early Nineties). I suspect also non-professional film are more or less in that league.
The fact remains that most of us just use a separate light meter and then set the values on our camera in Manual mode, so quality and precision matter.
Sometimes I find my slides slightly off what I would have expected. I suspect it is diaphragm error (I mostly use manual mode). I would like to have an electronic device to test diaphragm precision, like those that exist for focal-plane shutters.
I had to discard yesterday an overexposed slide, I exposed it using an incident light meter, and it was overexposed (Canon Canonet). I presume it is camera mistake. It's probably half EV but it was enough to wash out too much details where there should have been.
Ultimately quality and precisions are welcome. Electronic controlled shutters are better, but we don't notice 95% of the time. Mechanical shutters are good enough, slides have always been taken with them, but in my photographer's life I saw some "surprises".
Many seem not to trust in-camera lightmeters. If the subject is not backlit or sidelit, if the subject is average tone, if there is no sky in the frame, a modern in-camera lightmeter should be just dead on. Light meters are the smaller source of exposure mistake in my opinion. User error in light metering is another matter.
Last edited by a moderator:
