How can a camera achieve different exposure bias with identical exposure settings?

dyme · Jan 7, 2012

Hi everyone,

I encountered the following fundamental photography question that I just can't figure out. There are probably folks in this forum that know the answer.

I am taking a set of 3 bracketed pictures of a static scene with constant lighting. I am using a Canon 1D Mark IV which allows me to change the bracketed exposure settings to non-integer amounts, so I took brackets of this fixed scene at 2.7ev and 2ev separation between images, just to experiment.

When I examined the LOW image of each bracket, I saw that the -2.7ev exposure was darker than the -2ev as expected. The difference was not large, but was noticeable. However, when I looked at the image properties of this file (through either Windows or Photoshop), I found to my surprise that the aperture, shutter time, and ISO were all identical for the two images (aperture: f/22, shutter time: 1/60, ISO:100). On the other hand, when I looked under Windows at Properties->Details -> Exposure bias for these files, it said correctly "-2.7 step" or "-2 step." But none of the physical exposure parameters were changed!

How can the camera produce two differently exposed images of the same scene with the exact same exposure parameters? I thought that the only things the camera could change to modify the "brightness" of a photograph was aperture, shutter time, and ISO. So if all three are the same, why is the final picture coming out at a different exposure?

This is really puzzling me. The only thing that I can think of is that the camera IS taking the exact same image for both cases because the physical parameters are identical. Perhaps the -2.7ev exposure is undergoing some post-process in the camera (kind of like the ISO gain) that is darkening the image to match the photographer's desired intention of -2.7ev. But there is really no new "information" in the -2.7ev that isn't in the -2ev. But that explanation doesn't make a whole lot of sense to me.

Does anyone have any ideas? Thanks in advance.

Rephargotohp · Jan 7, 2012

The fact that the image "Looks" darker is the puzzling part to me.

I could come up with two senarios where the same setting could occurs though.

If you were shooting the sequnce of 3 in Shutter priority and set to 1/60. and you max aperture is f/22. The camera may not be able to actually do -2.7ev and stops at f/22 no matter what. But that should have the two images appear the same

The other case would be that both of the bottom exposures are the same But the 0EV exposure has changed so that even though the bottom exposures are the same. The diiference between the bottom exposure and the 0EV exposure has changed to the correct spacing of -2.7 EV and -2EV respectively

dyme · Jan 7, 2012

I was taking the pictures in aperture priority. Since I wanted everything in focus (and I was using a tripod), I set it to f/22. It seems to me that the camera could change the shutter time to adjust between -2.7ev and -2ev. If it couldn't (and I have seen cases where it is physically unable to do something, like you mentioned), then the two pictures should have come out the same. But they didn't. It's very strange.

I also checked the 0ev exposures. Those two are identical. Furthermore, I took the images seconds apart and nothing changed in the scene or the lighting, so I still don't understand why the -2ev and the -2.7ev would exhibit differences when their settings are the same.

ann · Jan 7, 2012

Where you using the exposure compensation dial?

I can' speak to Canon, but a question came up last year in one of my classes about which value changes with EC engaged. I called Nikon and after talking with tech support they had me talk to someone above their level as this was not a normal question for them to discuss.

I was told, neither value would change that the microchip would apply the change. Now before folks start yelling. I was testing this out with my d700 when the shutter went crazy and is now in the shop being replace. :x But when I get it back and I am going to give it another try to see what the metadata indicates with EC engaged.

When I first started the above test, nothing changed, but frankly several shots into the testing process the shutter started acting up so I can't take the first few images as proof of anything.

dyme · Jan 7, 2012

here you using the exposure compensation dial?

To set the bracketed exposures, I went to the AEB menu (by holding the MODE and AF-DRIVE buttons simultaneously) and then turned the dial until I had the desired exposure separation, either 2.7 or 2.

Rephargotohp · Jan 7, 2012

On Canon's anyway, I know Nikon does things differently.

It possible to use both AEB and EC at the same time. But you would see the center point move. In other words if you set -1 EC you would see the bracket now say -3.7EV, -1 and +1.7

dyme · Jan 7, 2012

On Canon's anyway, I know Nikon does things differently.

It possible to use both AEB and EC at the same time. But you would see the center point move. In other words if you set -1 EC you would see the bracket now say -3.7EV, -1 and +1.7

Correct... in mine I keep it at 0, so the three brackets are -2.7, 0, +2.7 in one set and -2, 0, +2 in the other.

OldBaldy2 · Jan 7, 2012

Weird.

Perhaps try it with the aperture opened up a little from f/22 (not sure why that would make any difference, but if 22 is your smallest hole, then it may be doing something weird), and switch it to manual mode rather than aperture mode. It SHOULD then only adjust the shutter, and also report the shutter speed actually used correctly. I'd also switch off all other compensations, to make sure it's not invoking some other trims.

My 5Dc reports the variances in shutter speed as expected.