Why would AE exposure vary this much?

As I explained in another thread, I was taking some shots in a situation where just getting the shots was more important than making them great, so I was using Program AE mode. I don't have a lot of experience using P mode and I'm really puzzled about some of the things it did. These 2 pictures are the perfect example. Taken 4 seconds apart, so the lighting's virtually the same, framing is almost the same, but the exposure is really different. Can anyone explain to me why that is?
These are straight "convert and save" from RAW in DPP with no PP of any kind.

HOSTED PHOTO
please log in to view hosted photos in full size.

HOSTED PHOTO
please log in to view hosted photos in full size.

What metering pattern? Evaluative? Center Weighted? Spot?
Did you focus and / or Lock Exposure and then recompose?

FlyingPhotog wrote in post #12067469
What metering pattern? Evaluative? Center Weighted? Spot?
Did you focus and / or Lock Exposure and then recompose?

Evaluative. Locked exposure and focus both, and I usually focused on the woman on the right, but I can't swear that I did both of these times. That could sure make the difference.

Boy, it can sure get educational when you start wondering why "the camera did that" and decide to find out!

Thanks a lot.

ncjohn wrote in post #12067642
Evaluative. Locked exposure and focus both, and I usually focused on the woman on the right, but I can't swear that I did both of these times. That could sure make the difference.

Boy, it can sure get educational when you start wondering why "the camera did that" and decide to find out!

Thanks a lot.

There's every possibility that you "bit off" more of the woman in one frame than you did in the other and that's what's giving you the variations in exposure.

The way around trying to "out think" your camera is to not do it.

If you specifically select an AF point to use (don't let the camera try and figure out your intended subject) and at least experiment with manual settings, you won't have to try and guess what your camera is doing because you'll already know what you want.

Are you familiar with the "Sunny 16 Rule?" If not, it goes something like this:
- Shutter Speed equal to ISO setting. ISO 100? Shutter Speed 1/100 (or 1/125 if you want to be "old school.")
- f/stop @ f/16

In full, direct sun, this should give you an acceptable exposure.

Now, obviously you probably don't want to be schlepping around with depth of field that will cover from your nose to Kansas, so all you have to do is work the reciprocal exposures down to the aperture you want:

1/125 @ f/16
1/250 @ f/13
1/500 @ f/11
1/1000 @ f/8
1/2000 @ f/5.6
1/4000 @ f/4
1/8000 @ f/2.8

Double the Shutter Speed = Halve the Aperture. Each increase in shutter speed means 1/2 the light hitting your sensor so you have to re-double the amount of light via f/stop to compensate.

In mechanical terms, for every one click of shutter speed in one direction, click your f/stop once in the opposite direction. And again, this is for sunlight between 10am and 2pm. Earlier or later, you'll need to adjust maybe two stops (clicks) of aperture to compensate for the less intense light of earlier in the morning or later in the day.

Easy Peasy!

And I apologize in advance if you are already familiar with any or all of this.

Rather than saying that...take time to learn exposure relationships, using live view and shooting in manual mode,

Mostly I would say it is the framing.
Evaluative metering will try to adjust the exposure to capture detail in all parts of the shot.

With the first shot you have cropped out the dark shadow of the hay bale.

The second shot with the dark shadow means the exposure needed to cover the lightest parts of the scene and the darkest too. Normally this will make the overall shot darker.

Need an exposure crutch?

Why?
Post #47

It's the nature of P mode and evaluate metering. 1st image is f8, 1/125, and 2nd image is f10, 1/200. The scene captured is slightly different causing the camera to change it's exposure calculation. It's noted that you had an exposure compensation set of +2/3. Did you know that? Were you trying to compensate for the white wall?

You should review the recorded histogram for each shot taken to see how the overall light was captured, whether too bright or too dark an exposure.

Evaluative metering is one of those 'camera knows what it is doing, the user never knows!' modes. In principle, it biases the reading to the focus zone used, but it also factors in the brightnesses of zones around the focus zone. But what does that really mean, about how it will treat any single shot?!?!?! Unknown. Here is an example of 18% gray card which is under the focus zone, but the exposure is still not right because of the adjacent sky...

HappySnapper90 wrote in post #12069146
It's noted that you had an exposure compensation set of +2/3. Did you know that? Were you trying to compensate for the white wall?

I have a history of "slightly too dark" exposures and was trying to compensate for that. This was a situation where I really wanted the shots regardless of the exposure so I was trying to give myself an edge without having to think about exposure (which of course was why I was using P mode). I'm used to working by myself, camera on tripod, no distractions; in this case I really wanted the pictures, I had big skittish animals around, and I was trying not to step in horse crap. So I was just trying to not have to think about exposure.

PhotosGuy wrote in post #12068792
Need an exposure crutch?

Why?
Post #47

I have tab in my browser that's been permanently fixed on "Need an exposure crutch?" for about a week.:p Can't wait to try it, but this wasn't the time for experimenting.

FlyingPhotog wrote in post #12067704
And I apologize in advance if you are already familiar with any or all of this.

I knew it all a long time ago but apparently sold it with all my film gear about 15 years ago. Since getting a DSLR I've been relying on either Tv or Av but I get less and less impressed with them as time goes by and I expect I'm going to go back to manual like I used to do. But thanks for the great explanation anyway.

And thanks to everybody for the responses.

Well, that's definitely a good example of why modes other than Manual are "don't care" modes! It's OK when you don't need consistency, but it can be irritating just the same!

Yes, I was trying to do some batch processing, and it requires several batches!

I still think the AE modes are a good thing, but I'm real close to being finished with them. They're supposed to make life easier but in some ways they require more thought and baby-sitting than manual mode. And I never had AE when I was shooting film, so... (Oh, wait, wait, I feel myself getting the urge to say, "...young fella"!)

"M" rules.....

Using manual exposure control lets ME do the thinking. In a situation like the photos in the first post, I would have made no changes to my exposure settings. However, the camera, especially when using "evaluative" changes its goofy mind when you point it at different elements in the scene.

Just to experiment, I had mine set to Tv and spot metering for part of the Renaissance Festival over the weekend. The results were mostly terrible. I didn't even bother trying to fix any of them, since I had enough good shots when it was set to Manual, ISO100, f/8, 1/640 (or sometimes a smaller aperture and slower shutter to get some motion blur).

nathancarter wrote in post #12071064
Just to experiment, I had mine set to Tv and spot metering for part of the Renaissance Festival over the weekend. The results were mostly terrible. I didn't even bother trying to fix any of them, since I had enough good shots when it was set to Manual, ISO100, f/8, 1/640 (or sometimes a smaller aperture and slower shutter to get some motion blur).

The problem with spot metering while in any automated setting mode (Av, Tv, P) is that unless you make sure something with 18% tonality is at the center of the fame at time of exposure, or unless you use Exposure Lock, the result of using Spot will typically be in error. Used properly, even automated setting modes will work (although without the shot-to-shot consistency of using M).