Inaccuracy in aperture?

[loccdor]

I have a Samyang 14mm f/2.8 lens that I am testing with the light meter before I go out with it.

When closed down to f/5.6 marking, my camera's light meter and my handheld light meter agree, when pointed at a white TV.

When closed down to f/6.7, the camera gives a reading I would expect from f/8.

When closed down to f/8, the reading is what I would expect from f/11.

Should I just treat f/8 as my f/11 when using the lens in the field? Or is the lens somehow "tricking" the camera and I should follow the lens markings? Thanks.

(Evaluative/matrix metering in the camera, though I don't think it matters, as the subject is a completely white frame).

 

[runswithsizzers]

A person who has two clocks does not know the time.

That is, when two meter readings do not agree, how do you know which one is correct?

 

[loccdor]

I don't doubt the accuracy of either meter under normal circumstances as I've never had a problem with either of them for slide film, I'm more asking about the interaction of the lens' aperture and optical design in the equation.

 

[guangong]

All of my various model Gossen meters agree with reading, but differ from other makes. Usually within latitude of the film. One should use camera with meter reading and change aperture a stop or two up and down to discover which reading is most pleasing to you. Then always use the same camera-meter combination. Runwithsizzors is quite correct.

 

[koraks]

Evaluative/matrix metering in the camera, though I don't think it matters

What does matter, however, is how metering is done. I assume this is with open aperture and the camera calculates exposure based on the aperture set on the camera. As opposed to stop-down metering where the camera actually takes a meter reading with the closed aperture. With stop-down metering, I can see how mechanical inaccuracies will result in different readings. However, since this is virtually certain not stop-down metering, evidently it cannot be a mechanical problem.

 

[loccdor]

Sorry, I should have clarified that. It is stop-down metering as this is a fully manual aperture lens with no communication of any kind to the camera.

 

[loccdor]

The discussion about calibrating 2 light meters seems more to apply when you have 2 meters that differ by some linear offset to each other. What I am reporting is a non-linear phenomenon that changes based on aperture settings. I.e, we can restrict the question to the in-camera light meter, discarding the information about the handheld meter, and theoretically say that if f/5.6 gives X exposure, then f/8 should give 2X exposure, but we observe this not to be the case.

 

[runswithsizzers]

There are too many uncertainties in your testing procedure to conclude the Samyang 14mm f/2.8 lens is the problem.

Some possibilities that may need to be ruled out:
The camera meter may have a non-linear response
The evaluative/matrix metering may be factor -- can you select average metering for the camera?

Can we assume you have repeated the exact same test procedure, but using a different lens on the camera?

 

[runswithsizzers]

The discussion about calibrating 2 light meters seems more to apply when you have 2 meters that differ by some linear offset to each other. What I am reporting is a non-linear phenomenon that changes based on aperture settings. I.e, we can restrict the question to the in-camera light meter, discarding the information about the handheld meter, and theoretically say that if f/5.6 gives X exposure, then f/8 should give X/2 exposure, but we observe this not to be the case.

In that case, lacking any kind of objective measurement to confirm or refute your assumption that metering through the lens is giving an incorrect reading -- then some practical test results are indicated.

What do the negatives show? Or better yet, shoot a roll of slides.

 

[koraks]

Sorry, I should have clarified that. It is stop-down metering as this is a fully manual aperture lens with no communication of any kind to the camera.

Ah, thanks. In this case, it's indeed quite conceivable that the aperture mechanism on the lens is inaccurate. The physical diameter of the aperture on a 14mm lens is tiny. The difference between f/6.7 and f/8 would be 0.334mm. It's easy to see how minor manufacturing tolerances can create significant differences.

 

[Chan Tran]

Yes which camera is it? Most newer Nikon cameras won't do a proper stop down meter in matrix mode. The matrix metering of modern Nikon camera (from the F5 onward) requires that the camera know the scene brightness. With stop down metering the camera doesn't have the scene brightness information.

 

[loccdor]

Thank you all for the feedback. I am about to shoot and develop a roll or two of slides with it. Just wanted a few leads to make sure I wasn't crazy beforehand (or I am, but that's a different story).

It's a Canon Elan 7E. The camera doesn't physically know there's a lens attached, the lens has no chip and you don't have to press the stop-down button on the camera, it is a purely manual-aperture lens of the most basic form.

The Samyang lens is known for fine optics, but poor quality control, so the mechanical issue does seem plausible to me. I don't mind it as long as I'm able to compensate for it. I mainly care about the aperture for diffraction limit purposes. I suppose I can try measuring the area of the circle.

 

[koraks]

I suppose I can try measuring the area of the circle.

Not very feasible at the accuracy required to make much sense of it.

Anyway, I'd not worry about it and just use stop-down metering as you're already doing, and personally with this camera (I'm familiar with it) use partial metering especially when shooting slides. The advantage of stop-down metering is that you'll compensate automatically for any inaccuracies in physical aperture size. I'd trust the camera's light meter in this case over the external one. In fact, in general, I find the Elan 7E's light meter to be a reliable tool (within the limitation that it doesn't have true spot metering capability).

 

[Chan Tran]

If the camera has no communication with the lens does the lens has a stop marking of f/6.7? If not how would you know if it's set for f/6.7??

 

[Maris]

Also check for hysteresis in the aperture mechanism. For example I have an old and worn enlarging lens where the "f8" I get by stopping down from f5.6 is different to the "f8' I get by opening up one stop from f11.
The lens is still consistently useable if I make sure to set apertures always from the same direction.

 

[Chan Tran]

To set f/6.7 you set the aperture ring between f/5.6 and f/8?

 

[wiltw]

I have a Samyang 14mm f/2.8 lens that I am testing with the light meter before I go out with it.

When closed down to f/5.6 marking, my camera's light meter and my handheld light meter agree, when pointed at a white TV.

When closed down to f/6.7, the camera gives a reading I would expect from f/8.

When closed down to f/8, the reading is what I would expect from f/11.

Should I just treat f/8 as my f/11 when using the lens in the field? Or is the lens somehow "tricking" the camera and I should follow the lens markings? Thanks.

(Evaluative/matrix metering in the camera, though I don't think it matters, as the subject is a completely white frame).

Over a decade ago I discovered that lenses originally used with my Olympus OM system film cameras, when mounted on adapters with/without chips to allow the digital camera to focus confirm, measured correctly only at f/4, and that wider or more closed down resulted in metering error.

Here is one of several lenses tested on a Canon 40D with chipped adapter, all frames exposed per the Canon meter's indication. Started wide open and progressing through smaller apertures...all frames should have resulted in the same midtone density if the camera had metered correctly...

Other lenses had deviations from ideal exposure, when metered to either side of f/4 on the lens.

 

[loccdor]

To set f/6.7 you set the aperture ring between f/5.6 and f/8?

Correct.

 

[loccdor]

Not very feasible at the accuracy required to make much sense of it.

Anyway, I'd not worry about it and just use stop-down metering as you're already doing, and personally with this camera (I'm familiar with it) use partial metering especially when shooting slides. The advantage of stop-down metering is that you'll compensate automatically for any inaccuracies in physical aperture size. I'd trust the camera's light meter in this case over the external one. In fact, in general, I find the Elan 7E's light meter to be a reliable tool (within the limitation that it doesn't have true spot metering capability).

Sounds good! The nice thing about this lens is you can set it to 2 meters and f/5.6, and everything from 1 meter to infinity is in focus and sharp. Assuming, of course, the optics are aligned...

That's mainly what I did yesterday and I'll develop some of the shots today.

 

[loccdor]

Over a decade ago I discovered that lenses originally used with my Olympus OM system film cameras, when mounted on adapters with/without chips to allow the digital camera to focus confirm, measured correctly only at f/4, and that wider or more closed down resulted in metering error.

Interesting, I thought the focus confirmation controlled the focus point, I did not know it could have an impact on the aperture of the lens.

Or rather, I guess it's that the camera is expecting the lens not to be used stopped down very far?

 

[MattKing]

It occurs to me that the aperture mechanism in an OM film camera lens is spring loaded, and the the lever that communicates the setting information to the camera is part of the mechanism. I wonder if the mechanism is dependent on the connecting mechanism in the camera providing resistance to the spring?

 

[Chan Tran]

It occurs to me that the aperture mechanism in an OM film camera lens is spring loaded, and the the lever that communicates the setting information to the camera is part of the mechanism. I wonder if the mechanism is dependent on the connecting mechanism in the camera providing resistance to the spring?

I found the Olympus OM lenses are different from the Nikon that is the lens always open to maximum aperture if unmounted. The must be a lever in the lens mount to push the stop down level on the lens to the aperture set on the aperture ring. So if you use the lens on non OM camera the adapter must have a lever that push this.

 

[MattKing]

Typo repair

 

[wiltw]

It occurs to me that the aperture mechanism in an OM film camera lens is spring loaded, and the the lever that communicates the setting information to the camera is part of the mechanism. I wonder if the mechanism is dependent on the connecting mechanism in the camera providing resistance to the spring?

Normally, one lever arm attached to the aperture ring communicates the set aperture to the Olympus SLR camera, so it knows what f/stop is set on the lens to close down an the time of shutter open...that lever has no communication to the digital body (no electrical contacts).
A second lever arm on the lens controls if the set aperture is triggered, or if the lens is wide open...the arm swings at the time of shutter release so the lens closes to the preselected aperture.

When used on an adapter ring on the Canon dSLR, the camera only knows how much light is entering the aperture and striking the meter....it effectively is the same as a TTL camera with a manual aperture lens: meter reading at the closed-down aperture. The ;adapter always puts the automatic diaphram into the 'triggered' position (closed down), but the camera does not know what aperture value was set...it only sees 'less light' and therefore suggests 'slower shutter', which should be proportional to the lesser light intensity hitting the sensor.
Where the Canon metering is failing is when the lens is set to f/4 the meter might say the shutter needs to be 1/1000, and when the lens is set to f/8 it SHOULD meter the resulting-2EV light striking the metering sensor and suggesting +2EV slower shutter or 1/250...but it wrongly suggests an even slower shutter (perhaps 1/125 instead) resulting in the lower density exposure.

 

[MattKing]

Normally, one lever arm attached to the aperture ring communicates the set aperture to the Olympus SLR camera, so it knows what f/stop is set on the lens to close down an the time of shutter open...that lever has no communication to the digital body (no electrical contacts).
A second lever arm on the lens controls if the set aperture is triggered, or if the lens is wide open...the arm swings at the time of shutter release so the lens closes to the preselected aperture.

When used on an adapter ring on the Canon dSLR, the camera only knows how much light is entering the aperture and striking the meter....it effectively is the same as a TTL camera with a manual aperture lens: meter reading at the closed-down aperture. The ;adapter always puts the automatic diaphram into the 'triggered' position (closed down), but the camera does not know what aperture value was set...it only sees 'less light' and therefore suggests 'slower shutter', which should be proportional to the lesser light intensity hitting the sensor.
Where the Canon metering is failing is when the lens is set to f/4 the meter might say the shutter needs to be 1/1000, and when the lens is set to f/8 it SHOULD meter the resulting-2EV light striking the metering sensor and suggesting +2EV slower shutter or 1/250...but it wrongly suggests an even slower shutter (perhaps 1/125 instead) resulting in the lower density exposure.

But does the aperture go to the correct position - or does it depend on the OM camera body lever being in the correct place in order to have the physical aperture match what is set on the ring? Could the problem originate with the adapter mount?
FWIW, when I had a Canon DSLR and used it with an adapter and my OM lenses, I didn't encounter any unexpected exposure inconsistencies.
The same applies to my wife's M43 Olympus OM-D DSLR.
I didn't use the OM lenses often with the cropped sensor Canon DSLRs I had and used for a bit. And I don't often use the OM lenses with the OM-D body now.