In-Camera Metering Accuracy

Dave in Kansas · Jul 11, 2011

I've had a Nikon FE2 for several years that wants to overexpose my pictures so I usually shoot on manual mode using a separate hand-held incident meter. Checking the Nikon's meter against my incident meter and a grey card, and comparing it to my Olympus OM-2 and a couple of Pentax cameras with accurate meters, the FE2 is overexposing by exactly one stop.

I took the FE2 to a local repair shop and the fellow checked it out and said the meter was right, but it might be a 1/4 stop off. That didn't seem right, so I took it to a different shop and that fellow said it's close, but about 1/2 stop off. I still think it's a full stop off.

So my questions:
Is the testing equipment used in camera repair shops only accurate within 1/2 or 3/4 stop? I have no idea how their equipment functions, but presume there must be some sort of calibrated light source.

Does the testing equipment need to be calibrated periodically?

Why is the FE2 off when it is probably my newest camera?

And for those of you who also have inaccurate camera meters do you get it fixed, or do you just determine how much off it is and adjust ISO accordingly, or simply shoot in manual mode using a separate meter?

Thanks much,

Dave

Monito · Jul 11, 2011

Bench equipment tends to be more accurate than microminiaturized portable equipment.

However, when you have your own equipment all agreeing within some unstated tolerance, except the FE2, trust your four pieces over the two shop pieces.

Just how close are your four pieces agreeing? (HH meter, OM-2, two pentax)

Measurement is fundamentally a statistical activity and that needs to be fundamentally understood before progressing.

2F/2F · Jul 12, 2011

Hi,

An incident meter should measure an amount of light that is 1/2- to 2/3-stop lower than the amount measured off of a gray card, so it will recommend an exposure that lets in that much more light than the gray card. In other words, the way to use a gray card in place of an incident meter is to take a reading from the card, and then open up 1/2- or 2/3-stop from that reading. I don't like doing this, so when I shoot using a gray card, I move my EI down two notches.

Aside from the possibility of flat-out inaccuracy, different cameras will also give different readings because they may use different metering patterns, and because they may use different values for middle gray.

In general, you are going to get "correct" readings all the time with an incident meter. But an in-camera meter compromises ideal readings for convenience.

The easiest way around all these variables is to simply pick one meter that will be the "master." The incident meter is a good choice to be the "master" meter. Have it serviced by a competent technician who specializes in meters specifically. (Industrial Light Metric is a good place to have this done in Los Angeles.) Then figure how to make all the other cameras conform to that meter. You can use the EC dial if your camera has it. If not, you can change the EI.

tomalophicon · Jul 12, 2011

I have an OM2 that reads 3 stops under my main meter, and my other OM2's meter, and various other meters I have lying around.
I just make sure I adjust the film speed dial when I put film in it.

brucemuir · Jul 12, 2011

The problem with adjusting the iso dial or some other standard is older Cds meters (not sure about selenium) can respond non linearly (may be accurate in bright daylight but off in low light).

Imo it's better to use a meter you know is at least reasonably accurate throughout it's entire range.
You can get Cds cells replaced fairly easily these days...

Rol_Lei Nut · Jul 12, 2011

For calibrating/testing, I find it very useful to tape a piece of white paper to a window, then hold a camera's lens up directly on the paper to measure.
Put a hand-held (non-incident) meter as close to the paper as possible.
This assumes that the outside light isn't changing rapidly, such as on a windy, cloudy day, but gives very accurate and repeatble results for comparing different meters, regardless of their measuring angles or patterns.

Diapositivo · Jul 12, 2011

I think the problem lies in the different reading patterns of the different cameras. How do you compare the cameras? Do you just point them in front of you and see how they measure the scene? The camera which weights more the sky (less prevalence to the centre) will give you a reading which is "closer" (higher EV).

In order to compare the two readings you should find a uniform wall in uniform light (uniform shade, uniform sun light) and then measure and compare (the wall must cover the entire frame).

If the error is linear, compensating the ISO should work as most lightmeters don't actually change anything during measurement when you change the ISO, they just measure a light intensity, which is then converted in an exposure value which depends on the ISO.

Chan Tran · Jul 12, 2011

First it was my newest camera that has the exposure meter off so newer doesn't mean much. My F5 that I bought new had to be sent back to Nikon for calibration twice before it's accurate. The first time to a local repair facility that work for Nikon and the second time to Nikon Mellville, NY.
I only have access to service manual for the Nikon F5 and F3 and in the F5 manual, Nikon recommends the Kyoritsu EF-8000 and it's a shutter, exposure and meter tester all in one. The calibrated light source is relatively small and the camera under test has the lens literaly touch the light source. The light source is adjustable for a range of brightness but I am not sure how much adjustment is available.
I forgot what's in the F5 manual but in the F3 the calibration is done at LV14 and LV9
Although we don't have calibrated light source we can say simply assume that one or more of our meters are accurate. To compare meter like others have said we must eliminate the difference in metering pattern. To do that I would point the camera or meter to an evenly illuminated surface and preferably of neutral color. I have the camera close to the test surface to make sure the entire screen is filled with the test surface. I have the lens focused at infinity, I don't want to focus on the test surface. Do this at the least 2 different brightness levels to check for linearity. This is eaiser done with the test surface being backlit like paper on window because the camera or meter won't cast shadow on the test surface.
Of course this will be a problem if the meter used as standard is an incident meter. If this is the case then you will need a larger surface and it would be difficult to get it lighted evenly and the surface should be of that a good gray card and the diffuser on the meter should be a flat diffuser and not a dome.

John Hermanson · Jul 12, 2011

There's no saying that any of your equipment is dead-on, so you're comparing it to camera that would likely show their own errors once they hit the repair shop. I calibrate meters to my shutter tester and advise customers that if they have a box full of cameras (that they never use) the overhauled camera will likely not match them, and not matching all his other "stuff" is not grounds for a "rework". Meters are check at 4 light levels. John

Dave in Kansas · Jul 12, 2011

Thanks for the replies and comments. My primary meter is a year old Sekonic L-308 incident meter, which appears to be accurate. Slide film (and b&w negative film) processed using this meter seem to be correctly exposed.

I walked out in the back yard one day last week and thought "This looks like a sunny 16 day" then got the meter and confirmed it was sunny 16 on the nose. That's a good test, right?

At that same time, I aimed the FE2 around and couldn't point it at anything in my yard to display a sunny 16 reading. The ISO needed to be adjusted anywhere from one to two stops higher depending on the subject. My OM2 and ME Super easily returned a sunny 16 reading. These are my primary slide film cameras and have historically produced properly exposed slides under normal conditions, except when expected such as backlit scenes.

I suppose the best way to confirm the FE2 metering is to load a roll of slide film and make test pictures under a variety of lighting conditions and various EI settings and then look at the results.

Dave

ic-racer · Jul 12, 2011

Dave in Kansas said:
the FE2 is overexposing by exactly one stop.

Multiply your ISO/ASA setting you had been using by 2 and set the camera to the product.

BrianShaw · Jul 12, 2011

John Hermanson said:
I calibrate meters to my shutter tester and ...

Out of curiousity, how often does a shutter tester in a camera repair environment get calibrated? I have vague recollection from a job long ago in which we were doing photometric light measurement in a scientific environment that our photometers were factory calibrated and certified annually, and before each use we "re-calibrated" using a known standard -- something like calcium carbonate.

Monito · Jul 12, 2011

Dave in Kansas said:
I suppose the best way to confirm the FE2 metering is to load a roll of slide film and make test pictures under a variety of lighting conditions and various EI settings and then look at the results.

It may show some useful differences.

Cinematography lenses are calibrated in T-stops: Transmission stops, so that variations in density of glass and other aspects are accounted for. I assume no UV or other filter is involved with the FE2 metering.

It could be question of Nikon linkages or electrical connectivity, depending on the camera and the vintage / series of lenses involved. Do you have multiple lenses for the camera or can you borrow some? If you have just one lens, can you borrow another FE2 to see if it is the camera or the lens or both that are out of adjustment or somehow dim?

Dave in Kansas · Jul 12, 2011

John,

I didn't see your reply until after my last post. You overhauled my OM-2 last year and the meter seemed to be accurate when I sent it to you and now I think it is even better, or at least as good. I'm very pleased with your work and glad to know you tested it under multiple light levels.

Thanks,

Dave

Dave in Kansas · Jul 12, 2011

Brian,

I am glad you asked about calibration of testing equipment and I have been wondering about that as well. I have suspicions about the testing equipment in one of my local shops. I took a medium format in for some focusing problems and the fellow told me he adjusted the metering. The metering on that camera is most definitely off now, overexposing by at least one stop.

Honestly, I only used a zoom lens when experimenting with the Nikon, but was recently given a 50mm so I should really check it with this one as well.

Dave

Chan Tran · Jul 12, 2011

I don't judge a meter accuracy by simply shoot film in the camera. There are many factors that would influence the result and not just the meter.
1. Is the published film speed accurate and/or is it what you want?
2. Is the shutter acccurate?
3. Metering pattern and metering technique affect exposure a lot more than anything else.
4. In Automode, Indicated shutter speed and acctually shutter speed may not be the same or even close sometimes.
5. The T stop thingy.

John Koehrer · Jul 12, 2011

In theory the tester should be calibrated annually. Most shops can't or won't spend the money or more importantly can't afford the "down time" to have it done.
Some of the testers also have "K" factors for different camera brands. We usually ignored it.
The older Nikons used a 60/40 pattern for metering. 60% in the 12mm circle and 40% on the rest of the finder. It's a center weighted system and probably won't agree with any camera that uses an averaging or spot metering system. Gray card won't read correctly unless you carefully read the instructions that came with it.
As I recall the card should be held at an angle to the subject for a proper reading.(60 degrees?)
Apples & oranges. you can't compare normal and telephoto lenses and expect the same result.
You can have proper automatic exposure with an incorrect indicated exposure as Chan Tran said.

Use one meter and ignore the readings on the other three or remove the batteries or as was suggested earlier, just live with it.

Rol_Lei Nut · Jul 12, 2011

John Koehrer said:
The older Nikons used a 60/40 pattern for metering. 60% in the 12mm circle and 40% on the rest of the finder. It's a center weighted system and probably won't agree with any camera that uses an averaging or spot metering system. Gray card won't read correctly unless you carefully read the instructions that came with it.

Using the piece-of-white-paper-taped-on-a-window method I described earlier, I get consistent readings from lenses ranging from 15mm to 600mm (1200mm with a 2x) and camera bodies of all kinds (spot, semi-spot, matrix, center weighted, averaging), naturally within the limits of any individual meter accuracies and lens T-stop variations.

Chan Tran · Jul 12, 2011

Yes Rol_Lei! Your technique eliminates the differences between metering patterns.

Pumalite · Jul 13, 2011

Matrix Metering in my F4 and F100 are excellent. Other than that; I prefer a Hand Held Meter

CGW · Jul 13, 2011

Dave in Kansas said:
I've had a Nikon FE2 for several years that wants to overexpose my pictures so I usually shoot on manual mode using a separate hand-held incident meter. Checking the Nikon's meter against my incident meter and a grey card, and comparing it to my Olympus OM-2 and a couple of Pentax cameras with accurate meters, the FE2 is overexposing by exactly one stop.

I took the FE2 to a local repair shop and the fellow checked it out and said the meter was right, but it might be a 1/4 stop off. That didn't seem right, so I took it to a different shop and that fellow said it's close, but about 1/2 stop off. I still think it's a full stop off.

So my questions:
Is the testing equipment used in camera repair shops only accurate within 1/2 or 3/4 stop? I have no idea how their equipment functions, but presume there must be some sort of calibrated light source.

Does the testing equipment need to be calibrated periodically?

Why is the FE2 off when it is probably my newest camera?

And for those of you who also have inaccurate camera meters do you get it fixed, or do you just determine how much off it is and adjust ISO accordingly, or simply shoot in manual mode using a separate meter?

Thanks much,

Dave

This always reminds me of the old saying: "A man with one clock always knows the time; a man with several clocks is never sure."

All my camera meters vary somewhat, most of them in linear fashion +/- across the range. But they are reflected light meters with somewhat different metering patterns. Most of my Nikons allow compensation adjustment. If your FE2 is over-exposing, just dial-in the adjustment. Incident and reflected in-camera readings won't be identical. Taking a reflected handheld meter reading off a gray card might get you closer to baseline accuracy for comparison's sake.

Dave in Kansas · Jul 13, 2011

I have to admit that I once had several cameras out in the driveway pointing at different subjects (garage doors, side of car, and other scenes) and got different readings on three different cameras. It makes one wonder "which is correct?" I thought my test last week was a little more accurate, but maybe not.

Sometimes I feel that cameras should not have built in meters. Half of my cameras don't have them and they work perfectly fine without. Then again, the metering on my OM-2 and my Mamiya 645 works very well. The Mamiya seems to be especially good.

I appreciate all the comments and info.

Dave

BrianL · Jul 13, 2011

I check my meters, both hand held and in camera by going to a shaded area with a brick wall of uniform gradation and metering it after posting a target on the wall to center on. This obviates the issue of how large the metering area is and as it is shaded because I did i fairly fast, little or no issue of changing light. I also found the area worked with both incident and reflective metering. While the mortar between the bricks can affect the spot metering care in aiming it can reduce the adverse effects. An issue with meters is some are calibrated to read 18% grey and others 16% gray but many spec sheets do not disclose which the meter is calibrated for. An alternative is a gray card but it needs to be large enough to cover an averaging meter's coverage. I have a 8x10 for this and simply set it up under a set of lights in the house to get away from the issue of changing sunlight. Another alternative is to simply go outside to a thick patch of green grass hold each meter in a hand and meter the grass. As the testing is done at the same time not an issue with changing light. The of course works better with handhoeld meters unless the camera is set up on a tripod. I've been pretty lucky with my meters both hand held and in camera. Usually within a 1/2 f/stop of my Weston Ranger that I use as my standard and in use seems very accurate.

benjiboy · Jul 14, 2011

Dave in Kansas said:
I have to admit that I once had several cameras out in the driveway pointing at different subjects (garage doors, side of car, and other scenes) and got different readings on three different cameras. It makes one wonder "which is correct?" I thought my test last week was a little more accurate, but maybe not.

Sometimes I feel that cameras should not have built in meters. Half of my cameras don't have them and they work perfectly fine without. Then again, the metering on my OM-2 and my Mamiya 645 works very well. The Mamiya seems to be especially good.

I appreciate all the comments and info.

Dave

A man who has one watch knows what time it is, a man with several is never sure :confused:

bobwysiwyg · Jul 14, 2011

benjiboy said:
A man who has one watch knows what time it is, a man with several is never sure :confused:

Bingo!

In-Camera Metering Accuracy

Privacy & Transparency

Privacy & Transparency