Can someone explain to me why so much emphasis is placed on the exposure based around the focal point when it is one of 63 or so sections that the camera is supposed to average out. With my 7D I find that if the focal point its a dark/black point the whole scene will border on the over exposed and the opposite when it hits a bright spot, say a glass flare on a race car or a piece of chrome the result will be well under exposed. Is there a way around this apart from full manual mode??

Are you sure your camera is set to evaluative metering? What you describe is more likely to occur with spot or partial metering.

Are you using flash?

Originally Posted by Aussieroo No Mark, no flash. I have been shooting for many years so not a rank beginner. This happens with the 50D, 60D and 7D I have and most lenses as well. It seems to me that the cameras give more emphasis to the focal point AE than I thought it should. I find myself making sure I don't focus on either a very bright point or a very dark point. Would be interested to see someone else test this. Take the same sot focussing in different brightness points and see what results you get.

You know, snide comments such as "I'm not a rank beginner" are not helpful. Nor was your "after 12 years" comment to Richard. I was just trying to get more information about what you were doing in order to try to help you.. You provided a bare bones explanation of what is happening, and I (and Richard) are trying to ask questions that might lead us to the right answer. We can't see you, your pictures, or your camera, so we need to ask questions so we can understand what is happening.

If you would prefer people not to ask questions, you should put that in the body of your original post.

I do apologies Mark and Richard, I did not see my comments as snide, hence I have removed them. I was trying to point out that I have tried all the obvious things and would not have asked the question had it been a simple answer, but you are correct, ESP doesn't work well over the internet and I will try to be more comprehensive in my questions in future.

Posted examples, or links to examples, with full EXIF, can help us to help you solve what may be going on.

Probably 80 percent of the time evaluative metering works fine. The times it does not work are when the combination of database lookup and algorithmic rules fail against scene extremes.

This page and article by Canon can help to describe how to handle difficult scenes, light and dark:

http://learn.usa.canon.com/resources..._article.shtml

----------------------

Many photographers choose to "decouple" the AE and AF functions using "AE Lock":

http://learn.usa.canon.com/resources..._article.shtml

----------------------

Using the above techniques and methods, along with the understanding that the metering system is designed for measuring middle gray, goes a long way towards mastering exposure on the Canon 7D. Combine the above techniques and methods with Exposure Compensation (EC) to complete your control of the camera's metering.

Thanks Ziggy, I have been using a version of the zone system since I viewed a video on the subject some time ago. When I want rich blacks where black is a predominant colour I under expose and when white is predominant such as snow shoots I over expose. I mostly shoot on AV settings when subjects are not moving and have time to over ride the cameras average exposure calculations. However when shooting a mixed scene, or car races there are occassions where the focal point hits either a very bright spot or a very dark spot which in action shots it is hard to avoid especially when shooting in AI Servo mode tracking a car around a track, I get 2 shots in that sequence with a noticable difference in exposure. I read that the camera does still emphasise the focal points AE where it was my understanding that in full evaluative mode each small area was treated equally. One of the issues I have narrowed down to the "safety shift" setting in AV & TV where it can cause a shift in exposure. My last shoot at the airshow I was using the back focus button and I found I was still varying my exposure between plus 1 to plus 2 stops to get detail on the underside of the plane shooting into the sky. When I get a chance after work, I will see if I can find a couple of samples to post. While this is no real big deal for me it is still curious that a series of shots taken at 8 frames a second there can be a shot that is so different in exposure to those around it. Thanks for the links I had read the first one already and must admit I rarely use AE lock. I need to look into that and practice that as well.

Here is a quick sample of what I am talking about, The first photo I had the single focal point on the dark doorway and it became very light. the second shot I moved the focal point to the brighter front area and it is a better over all exposure. And you can see from the exif info nothing else changed. To my way of thinking this big a change in exposure is more that I would have expected evaluative metering to produce. Knowing that it does this means I am aware of it and try to avoid it but I am still wondering if there is a way around it? Perhaps a centre group of focal points rather than a single focal point? I haven't found a car sample as yet but you can see from this if the focal point hit a bright reflection on the car whilst tracking the car it would make a big change in the result.

The first photo I had the single focal point on the dark doorway and it became very light. the second shot I moved the focal point to the brighter front area and it is a better over all exposure. And you can see from the exif info nothing else changed.

Shutter speed did indeed change in response to metering from a different point on the image - in the first exif, it's 1/2; in the second, 1/5. The camera is reading the overall scene, but trying to make the "main" area you've focused on more important, if you like. I think that's normal and to be expected... and why in high contrast shots manual is often a better choice, where you can control just how much (or how little) certain extreme areas of the shot will affect the metering by intentionally over or under exposing to compensate. Exposure comp will do the same thing in Av or Tv, of course, although I find that a little trickier to judge.

I meant I never changed any setting and the camera did indeed change the shutter speed. It does seem best to shoot full manual mode especially when one has the time for adjustments as was the case in this shoot where the camera was on a tripod. The difficulty come when shooting action especially when there is a partly overcast sky with the sun in and out sometimes in the middle of a sequence shot. Really no time for adjustments so While I know now the camera give greater emphasis to the exposure on the focal point I would still like a work around apart from full manual mode where quick adjustments cannot always be made during a sequence shoot of say 10 shots. Does anyone shoot car races and do you experience this and is so what settings do you use to overcome this?

Have you considered manual, but with auto ISO? The 7d's auto ISO is REALLY good - I've been impressed. With the new firmware (coming soon) I believe you'll have more control over upper/lower limits as well. Give it a try and see what you think - I've found it very helpful for variable lighting situations, particularly outdoors with changing lighting conditions. I've also noticed that when the camera sets the ISO, even if it goes quite high, noise is minimal because the exposure has been accurately judged (ie not underexposed).

Thank you so much for that suggestion. I never gave it a thought but it makes perfect sense. I will definitely give that a try next outing. Will the iso change on the fly if during a series of shots at 8 FPS and there by change the exposure? Yes I am looking forward to the firmware upgrade. I can't believe that half of the upgrade is already in my wife's 60D. I use 2 24" screens for editing and I find that if I don't pixel peep the noise at high ISO's is very acceptable except as you say in the under exposed areas and when there is a lot of sky that is not correctly exposed in landscapes for example.

I seldom shoot at high fps so can't answer that; best thing is to go out and try it! I'd be surprised if shots running that fast would have significantly changing light (????).

It's not infallible, but I have found it very helpful when shooting in fast, variable lighting. Often the ISO's it chooses are miles away from what I would have picked.... but the shot is usually better than the one I would have got had I been setting the ISO manually and adjusting SS or aperture. It just WORKS. :)

I just tried an experiment similar to your barn-door experiment, but with my Pentax K-5 on Matrix/Evaluative metering. There was no exposure difference regardless of which focal point I chose.

There was a minor exposure difference if I used the focus/recompose method, but it was very minor. So I guess it's a Canon thing?

Good old Canon! I have been checking the Exif on a series of car shots and 5 shots taken in one second shot on a sweeping curve at 1000th of a second in shutter priority have gone from f10 to f7.1 on the last shot and become progressively lighter. Maybe I should try manual mode with auto ISO next race and see what happens.

Do be careful that light is not entering through the viewfinder as that can affect exposure too.

Ralph, try separating exposure metering from your AF point selection.

On the bottom of page 106, of the 7D manual, is says that in Evaluative mode the exposure is applied that was metered at the chosen AF point - whether chosen by the camera or by the user. You can alter this behaviour by using the * Auto Exposure lock button, or switch to other light metering modes as spot, center weighted, or partial metering do not meter at the chosen AF point, but the general center of the image. Each of these threes modes are described further on page 103, and meter at the center of the frame, not at the chosen AF point. This is a good tip to remember, and that I had to review the manual to remember.

You might try an incident meter to get a better average exposure, but even there as the cars move in and out of the ambient lighting, your exposures will need to be varied.

I think your options are to shoot manual, which introduces difficulties when shooting fast action like motor sports, or birds in flight.

Or shoot RAW in Av with center weighted metering, and Auto ISO if you prefer, and reduce the frame to frame exposure variations in Adobe Camera Raw or Lightroom's RAW engine.

Like divamum posted, I never liked Auto ISO in years past either, but with modern cameras it works pretty darn good today.

I would wonder if Pentax handles this AF point selection, exposure point selection in Evaluative mode differently, based on what Mark said.

Originally Posted by pathfinder Ralph, try separating exposure metering from your AF point selection. I would wonder if Pentax handles this AF point selection, exposure point selection in Evaluative mode differently, based on what Mark said.

You can couple AE with AF on my K5, but by default they are decoupled, and that's about as far as I've researched it. I believe my old K20 worked the same way.

A few of us are off to a vintage race meet in a weeks time and I have a list of options to experiment with. I am wondering how the 7D will handle the exposure if instead of a point focus I use a group focal points. Will it work differently or will it grab one of those points and still do the same. Getting into the bowels of the custom settings for AI servo and remembering what they do and be able to change them on the fly is a real challenge. I will use centre weighted as well as the main subject will be the car and the back ground will be secondary. Hopefully now it will be a sunny day, but mid winter I don't like my chances.