1D Mk IV occasional under-exposure (1 Viewer)

[Malcolm Stewart]

I've now had a 1D Mk IV for 6 months, and have shot several thousand images. (Prior to this, I'd used a 1D Mk III for several years.)

In the main my exposures have been fine, but two months ago I had serious under-exposure with my normal settings, shooting similar subjects to my normal. I'd set the ISO to 320, exp comp at -2/3s, and was using my EF 28-105 F/3.5-5.6 USM II in Program mode. I went for a walk under cool conditions, but comfortable for me - so no icing or similar.

During the walk, the problem lasted over around 6 exposures, and exposures were fine either side of the 6 which were similar to the example, which was the worst. I always have the camera up to my eye when shooting, so viewfinder light leakage wasn't a problem. It was cloudy at the time, and I guess the worst was around 3 stops under-exposed - that's after examining it in DPP.

The example image was the worst. Amazingly, it recovers quite well. Others in the six were around 2 stops under-exposed.

Today I decided to check whether exposure was consistent versus a variation in ISO. With my 1D IV tripod mounted, and same lens as above, I shot the same scene from an ISO of 100 to 800 - all of the images were near identical in tone etc., and very similar in jpeg file size.

I've raised this elsewhere, but got just one response, so I'm hoping that there's more experience here. For the record, my 1D IV was powered by a Canon LP-E4 battery which was recognised correctly. In addition to the tripod mounted formal checks mentioned above, other checks have all returned good exposures with the lens in question.

Mystified!
Over the weekend, I'm hoping to visit two WWT Reserves, so I'm hoping that the problem has really gone away.

 

[tdodd]

As far as I can judge, despite a rather odd combination of settings for the exposure of a "landscape" scene the numbers arrived at appear to correspond to what should give the expected exposure for the lighting conditions. i.e. one that is 2/3 stops underexposed since -2/3 EC was dialed in. Unfortunately the image does look to be nearer 2 stops underexposed and, given the metering parameters and exposure mode used, I am at a loss to explain the shortfall.

If exposure had been set manually without metering through the lens (e.g. by using a hand held meter or manual application of the Sunny 16 Rule) I would be tempted to say that somebody forgot to factor in the use of a CPL or ND filter, but that's not the situation here.

I'm as mystified as you.

 

[jalethbridge]

Hi, I had a similar problem and it turned out that every now and again I was hitting the AE lock (*) button without realising it.

 

[GYRob]

it could be sticking apeture blades .

Rob.

 

[tdodd]

it could be sticking apeture blades .

Rob.

That would cause the lens to not stop down fully, which would lead to overexposure.

 

[tdodd]

Hi, I had a similar problem and it turned out that every now and again I was hitting the AE lock (*) button without realising it.

I'm not 100%, but I think the EXIF might indeed be saying that AE Lock was used, but in such a mid toned scene it is hard to imagine what was metered and locked to lead to such underexposure. Maybe the camera was mostly pointing at the sky when (if) lock was set. The grey clouds with +0.7 stops added back on in Lightroom, to compensate for the -2/3 EC, do come out fairly close to middle grey.

 

[GYRob]

That would cause the lens to not stop down fully, which would lead to overexposure.

Not sure about that Tim as the blades could be sticking in the stoped down position and you would not nessasary see it through the view finder as your eye would ajust to some degree for the loss of light .

It has to be shutter/apeture or iso related .

Rob.

 

[tdodd]

When you have f/16 dialed in and you claim that sticking shutter blades would have it smaller than that I'm pretty sure you'd notice through the viewfinder. What would even cause the lens to be stopped down smaller than f/16 unless you'd been using even smaller apertures previously? EF lenses spend 99.9999999% of their life wide open. If they're going to stick at all it's far more likely, IMHO, that the stickiness would prevent them stopping down when called upon to do so, not the other way round.

 

[Malcolm Stewart]

I'm not 100%, but I think the EXIF might indeed be saying that AE Lock was used, but in such a mid toned scene it is hard to imagine what was metered and locked to lead to such underexposure. Maybe the camera was mostly pointing at the sky when (if) lock was set. The grey clouds with +0.7 stops added back on in Lightroom, to compensate for the -2/3 EC, do come out fairly close to middle grey.

Very many thanks. That's it!

I've looked back at the detailed EXIFs for the images which were bothering me, and they did have AE Lock "ON" - the others didn't. It's two months ago since I took the shots when the weather was colder, and I was probably wearing gloves, and didn't realise that I'd pressed one of the two * buttons. All my shots in the last few days have been fine - OK, in some cases, I didn't wait long enough when tracking for the AIServo to properly measure the motion of the target, but mostly they've been fine.

 

[jalethbridge]

Very many thanks. That's it!

I've looked back at the detailed EXIFs for the images which were bothering me, and they did have AE Lock "ON" - the others didn't. It's two months ago since I took the shots when the weather was colder, and I was probably wearing gloves, and didn't realise that I'd pressed one of the two * buttons. All my shots in the last few days have been fine - OK, in some cases, I didn't wait long enough when tracking for the AIServo to properly measure the motion of the target, but mostly they've been fine.

A salutory lesson - always look to blame the user before the equipment! I reacted in exactly the same way as you, could not believe it was anything to do with me. Turned out I was hitting the * button accidentally every now and again.... An excellent way to avoid this, and to improve your photography at the same time, is to move away from AV and use M.
Cheers
Jonathan