5D MkII overexposure/metering Issue

Discussion in 'Canon EOS' started by mike_gendimenico, Oct 6, 2011.

  1. I purchased a 5D MkII last December after upgrading from the original 5D which I owned for about 4 years. Over the past year I've noticed that the MkII seems to regularly overexpose outdoor shots a couple of stops. I primarily shoot in Av mode, sometimes in P. I've even seen this with still subjects where I fire off multiple shots and one shot will be overexposed and the adjacent frame is exposed correctly (no refocus or recomposure between frames). A lot of times I just end up adjusting the exposure down to compensate.
    I never had this problem with my original 5D as the metering was almost always spot on, so you would think the more advanced MkII would have better metering, not worse. I've seen some other posts around the web on this same issue but never saw a solution.
    I've also had the occasion of a flash photo just turning out completely black even though the flash fires and then the next shot is ok.
    I will probably send the camera into Canon next month to have it checked out but figured I would throw up a post here as well.
    Thanks!
     
  2. I read your post but I couldn't tell what your question was. Are you complaining? Searching for happy 5D2 campers? Like minded 5D2 sufferers? Obviously some is wrong with your camera and it needs service (or you have a sticky diaphragm--happened to me with an old Nikkor).
    I've shot extensively with a 5D2 for 2 years and must say the exposure has been consistently good. I can't say it is better than my old 5D--or worse--but I need to use both EC and FEC less than I did with my 10D and 40D. Sure it can be fooled by the usual white sand beach, snow fields, etc., but so have all cameras I have owned and I just dial in -1-2 stops EC and I'm good. Most of the time I don't need to touch EC/FEC. In terms of IQ, it's the best camera I have owned and I'n owned many dozens.
     
  3. Check exposure compensation settings, based on what you said about the second in a series being correctly exposed, you might have the camera set to bracket exposures. As an aside, my 60D did the very same thing about a month after purchase, but that was due to a faulty processor. (as per Canon repair order).
     
  4. I am a 5DII shooter and have exactly the same issue. I noticed that subjects with darker shades than a middle gray are rendered as a middle gray, not a matter of concern. I just adjust Exposure Compensation down and reshoot, often 2/3 to 1 1/3 to the minus.
    It seems, though, that there's more to the issue. On vacation, I shot a lot of land- and city-scapes with a variety of lenses and saw overexposure more often than not. I always reshot, compensating down. In processing my RAW's in CS2, my shots that originally registered as overexposed ARE, in fact, overexposed. Even RAW processing does not fully recover the blow-outs in skies. The adjusted exposures, however, seem fine. I am not seeing noise in the shadows I'd usually blame on underexposure. To be fair, I had exactly the same issues with a Pentax K20D.
    Almost never does a centered exposure compensation screen seem right. I see no difference in rendering with Averaging and Matrix metering. I get better results using Partial, as in my Canon FD days, with Eposure Lock, then recomposing.I am watching this thing carefully. I do remember having just these issues in my slide film days, and painfully regretting not having bracketed exposures in shots I had high hopes for.
     
  5. What meter settings do you use and what subjects are you shooting where you see the issue. Evaluative metering is just that - better than center weighted but far from perfect. It is better than the old Canon FD Center weighted system. Remember shooting digital you have an exposure latitude closer to slide film than print - on FD bodies (I still shoot them) it takes lots of experience to know what exposure compensation you dial into a scene when using center weighted and shooting slide film. The complexity I have found with the evaluative (and Matrix) systems is that they are much less predictable than the old Spot, partial and CW systems. An evaluative system will determine what it thinks it sees and adjust exposure accordingly. I never had the 5D but my 5DII metering is very similar to all my EOS bodies back to the 1V and 3 - the body I find that differs is my 7D (and indeed Canon claims a different metering system that reacts to colour). The 5DII would benefit from multi spot metering (a system that you only get on the 1 series these days but one that I have used and loved since it was introduced on the T90).
    What I do generally (with my 7D 1DIIN and 5DII) is use sport or partial metering quite a lot - I also focus using the center AF point and re-compose (the Spot and partial will thus be set for the focus point). With complex subjects or where the point of focus is not a good place to meter I either use AF on the AF on button or more usually I just bracket. In general in difficult light just use spot or partial metering nine times out of ten this solves the problem. If you think you have a meter problem just use a tripod and a grey target and compare different bodies (I have done this with bodies right back to the original F1 and I found that the DSLRs tend to expose about 1/2 a stop brighter than the older film bodies). If you do not have another body see if the four metering patterns produce the same result - if the grey card is evenly lit and fills the frame they should!
     
  6. >> I noticed that subjects with darker shades than a middle gray are rendered as a middle gray.

    Which is the expected result. Reflective light meters are tuned to render scenes with an average tone of middle gray. That's why snow look gray
    in an uncorrected exposure.

    To the OP: it would help if you could post an example. For more consistent results, meter in P or Av, but shoot in M (no auto ISO please). Unless the light changes, each frame will have the same exposure. If it doesn't, there might be a problem with your camera. The histogram and the blinking highlights display can help you evaluate exposure accuracy.
     
  7. The large wheel on the back can move the exposure compensation amount accidentally if you're not careful and if you
    have the camera switched to the second ON mode.
     
  8. Mike,
    It could be that what we are seeing here calls for a scientific trial on our part, where we control the variables (same lens, scene, light conditions, metering setting, and ISO, using a tripod) and see if our results are consistent (per Dan's advice above).
    IF our results, after opening a RAW in post processing, show highlights that can be dealt with and clean shadow areas (a reason for buying the 5D2), it could be that our 5D2's are OK, that they are acting like some digital SLR's. When using my 5D2, I keep a running dialogue in my head for dealing with the scene. I shoot once and look for blinking highlights and/or the highlight train wreck to the right. I then adjust Exposure Compensation to the minus side. I have gotten used to leaving Compensation at a minus 2/3 for grab shots. If the scene presents difficulty, I resort to Partial Metering, placing the partial area half in shadow and half in highlight (the split-the-difference technique), OR place it on evenly lit grass or medium-looking stone and try that. Sounds a bit involved, but What The H, maybe the mental activity will ward off Altzheimer's.
     
  9. What metering mode are you using, ie: spot, centre weighted, matrix, ummh: I think there's another, can't think. Spot would obviously be most touchy, prone to mis-read.
     
  10. I am mainly looking to see if anyone else has this issue and trying to determine if it's an operator or camera issue. AEB is off and yes, I'm aware of the EC wheel setting since that's how I stop the exposure down to compensate for this problem. I mostly use Evaluative metering and only change that for something like very bright backgrounds. I really don't like to use Manual for events, concerts and the like. Call me lazy but if my old 5D metered consistently with the same settings, I don't understand why the newer model can't do the same.
    Will see if I can dig up an example this afternoon. Since the camera is still under warranty I will most likely send it back either way and have it checked out.
     
  11. Mike try compering the old and new bodies in a controlled environment with a grey card target and see if they read the same. All my FD and EOS bodies were within 1/2 a stop when I did this test (about 20 bodies). Then try with different modes on the 5DII (making sure the target fills the frame - use CW, Evaluative, Spot and Partial). If you see significant differences between the 5DII and other bodies or between the 5DII modes than you have a camera problem. If not then it may still be the way you meter.
     
  12. I would like to try a controlled test, but crazy travel schedule through November so that will have to wait. In the mean time here are a two recent photos that were shot sequentially but exposure ended up a few stops different.
    These are both are shot in Av, evaluative metering, No AEB, f/5, ISO 500. This photo shutter speed was 1/200.
    00ZRJ2-404911584.jpg
     
  13. Second frame, Same settings but shutter speed dropped to 1/125 even though all other settings were the same with no refocus or recomposure.
    00ZRJ7-404913584.jpg
     
  14. Something to keep in mind, did your surrounding change at all. For instance sometimes the slightest surroundings change can cause the very sensitive metering system in the EOS brand cameras to make unfavorable adjustments yielding such changes in one photograph to the next. This can happen even if the wind blows a few branches in a tree and changes the light source just a little bit, or if a cloud would partially cover the sun which would be your light source and the camera would adjust accordingly to that moment. Also, the meter may have looked at that tree behind the subject and adjusted to that instead of your subject causing the over exposure of the photograph.
    You also may want to send your camera to Canon service and have them do an evaluation test on your camera's metering system to see if they can reproduce this problem and adjust the camera or even replace the metering system circuitry if that is what needs to be done.
    The Canon service center which I have worked with before does an excellent job at repairing and calibrating the camera so it is working perfectly after repair with them.
     
  15. Mike - for that type of shot i would use partial metering but I am old fashioned. The difference in exposure settings 1/200 ISO 500 F5 to 1/125 ISO 500 F5 is quite small - 2/3 of a stop which may well be due to a difference in lighting between the two shots. That said the difference in images looks to be a stop (although this is hard to judge on a Mac screen). So as Duane says there may not be a camera problem.
    This link may help
    http://www.learn.usa.canon.com/resources/articles/2011/qt_photograph_snow_article.shtml#
     
  16. Thanks for the input guys. It's about 2/3 of a stop difference if I compare in Lightroom but I have had more severe overexposure at other times, just don't have those photos handy. Many times I have to adjust the EC ~-2/3 to get a properly exposed photo where I never had to do that with my old 5D. Again, I will probably send the camera in for service within a month or so to have it checked out.
     
  17. I just quess, that there might be two possible reasons for that kind of behaviour. If the first metering is very near of just choosing the shorter time, and the variation of light then chooses the other one of longer time value seems to give the 2/3 difference of the exposure.
    The other reason might be the difference of the metering system and especially if using the matrix metering, which sometimes gives in my opinion more random exposure. In matrix metering I have noticed, that the exposure is quite often overexposed. With other metering modes the shot will be better controlled and does not vary that much.
     
  18. I too own a 5D2 and can report that: a) Occasionally I get underexposed images but usually it is my fault because I have fiddled with the EC dial b) Light conditions fooled the camera meter. When none of these occur, exposure is normal. One thing you can do to check camera meter is use an external spot light meter (borrow if you don't own one) and with camera in spot metering mode compare readings. Caution: The display on the camera back has auto brightness adjustment, based on ambient light; this may fool your eye as regards exposure, if you are judging exposure from the camera display brightness.
     
  19. Finally sent my camera in for service this past month and got it back late last week. In addition to replacing the right top cover (banged up from a drop), they did report that the AE assembly adjustment was incorrect which was causing the uneven exposure. Not sure if that mis-adjustment was the result of the drop but at least I know there was something wrong with the camera and not operator error. :) I do have to say the camera came back spanking clean and looks almost new. (It was pretty filthy after a few multiple day music festivals this past year...)
     
  20. Mike,
    I have the same exact problem when I shoot dance photos on stage. I have three shots in a row where the dancers are still in a pose, and the middle shot is over exposed, the frames on each side are not. It happens quite frequently. I shoot Av, maxed at f/2.8, and 3200iso, partial metering or center weighted. On the over exposed shot, the black curtain in the background is close to neutral grey. So it is obvious it picked up on the curtain. This happens usualy when there are two dancers aligned to the 1/3rds of the frame. I think that these modes are unstable under my conditions since they are weighted towards the center, and the dancers are just enough off center, that it may get confused.
    Shooting manual exposure obviously is a work around, when the lighting doesn't change rapidly. But in scenes where the lighting is changing fast, the chances of getting a properly exposed shots are even less.
    I am going to try using evaluative metering, and then use exposure compensation to get the correct exposure. I suspect that this method will be much more stable to any spatial light variations across the frame.
    Do you have some sample images you can post? What metering mode do you use? Maybe there is a problem in their system or your camera that causes the metering to get "lost" like the autofocusing can sometimes do. Feedback control systems aren't always perfect.
     

Share This Page