Do some Nikon DSLRs really have a slightly greater tendency to "overexpose in bright contrasty conditions" than do those of other brands?

Discussion in 'Nikon' started by landrum_kelly, Nov 3, 2012.

  1. I am certainly making no such claim. I just wondered why a DPReview article would say something very similar to that:
    Focussing and metering [on the D3200] are generally very reliable, although like many previous Nikon models the camera has a tendency to slightly overexpose in bright contrasty conditions. In those situations it's worth dialing in 0.3 or 0.7EV negative exposure compensation to protect the highlights. [citation link]
    Yes, I had seen this tendency on the D3200, but the high pixel density of the D3200 could perhaps be a factor there. No, the DPReview article did not actually say that the problem might be greater for Nikon than with other brands, but the question immediately arose in my mind.
    Is there such a tendency, and, if so, is very high pixel density the primary cause of the problem?
    I have addressed this at length in a photo forum (about one of my own photos) where I have talked about this only real disappointment that I ever found (apart from ergonomics) in the D3200, which otherwise turned out to be a real honey of a camera, especially for the price. So far, I have been having the conversation solely with myself:
    Even on that issue, of bright and contrasty situations, the remedy did indeed seem to lie with nothing more complicated than dialing back the exposure compensation either during the shoot or later in processing NEF files. Still, are Nikon files more finicky to deal with?
    I was just wondering how any or all of this might affect landscape photography, and, if so, with which particular cameras? I happen to like "bright contrasty situations," and I need to know how to deal with them if these kinds of problems routinely arise.
    I'm not trying to be a troll on any of this. In fact, I just sold my last piece of serious Canon equipment today, my beloved EF 24-70mm f/2.8L. If processing Nikon files has different challenges from those of Canon or other brands, I need to know about them. The results I see with Nikon cameras are incredible. I just want mine to be among them.
  2. "I have been having the conversation solely with myself" - this is the best way to converse...:) nobody argues, and you are always right.
    "I'm not trying to be a troll" - ...? ... it did cross your mind, didn't it ?
  3. Pixel density has nothing to do with overexposure - metering does. Some Nikon cameras seems to meter slightly differently than others. I found my D200 tended to underexpose. While my d80 tended to get it wrong a lot. My D3 and D800 are generally right on the money in the metering department.
    "are Nikon files more finicky to deal with?" "If processing Nikon files has different challenges from those of Canon or other brands, I need to know about them" They are no different from each other.

    "The results I see with Nikon cameras are incredible. I just want mine to be among them" All things being equal, the results will be the same. Incredible results are the result of incredible technique and subject matter.
    I shot dual platform for a couple of years, and sold off almost all my Nikon gear and then went back to Nikon exclusively because Nikon bodies had a couple of feature I needed that Canon did not. I know others have gone in the other direction for the same reasons. Each manufacturer has features that attract photographers to their system because another system may not have what they need. But it had nothing to do with overall IQ or file attributes. Shooting Canon side-by-side with Nikon will give you pretty much the same results with variances mainly due to perhaps slight metering differences as well as picture control settings.

    "just sold my last piece of serious Canon equipment today, my beloved EF 24-70mm f/2.8L" I found this to be one of the best lenses I had ever owned and was the hardest for me to give up. In my opinion, it is slightly superior to Nikon'es 24-70mm lens.
  4. "Do some Nikon DSLRs really have a slightly greater tendency to "overexpose in bright contrasty conditions" than do those of other brands?"
    An unqualified YES! IMO.
    In fact you can omit the word "slightly" from that previous question. It's time Nikon sorted out its flakey and unpredictable 3D matrix metering. It seems to behave like average metering when you have a bright subject against a dark background, and an AF-point-weighted spot meter when you focus on a dark area of a scene. This is precisely the sort of mis-programmed behaviour that leads to overexposure in many common situations. And never expect it to retain any detail in your skies!
  5. Hey Rodeo Joe: I remember your "Where's my Sky, Dude?" thread a few months ago- D800 I think. I believe that that long thread (mostly) concluded that matrix metering was a tool and once you learned how it reacted you could adjust accordingly. I see you were not convinced! :>)
    I believe the D700 has about 2 stops of recoverable highligh detail where as the D800 is more like 1 stop. Don't know about the D3200. I think that including the raw recovery information in the metering of the camera is reasonable. I guess jpg shooters might feel differently.
  6. Hi Robert.
    "I believe that that long thread (mostly) concluded that matrix metering was a tool and once you learned how it reacted you could adjust accordingly."​
    That was the "conclusion" forced on the thread by a few (2 or 3) vociferous contributers, who refused to acknowledge the argument that matrix metering should return a fairly high hit rate of useable exposures without intervention. They also failed to explain exactly how it reacted - apart from apparently randomly!
    Sticking a wet finger in the air could also be regarded as a metering method that requires some user adjustment.
    If you're going to have to watch every exposure and treat the first ones as metering shots, then matrix metering is plainly unfit for purpose. You might as well use average metering and work from there, or from a histogram. In fact a live histogram, such as is offered by many cheap compacts, would be much more useful.
  7. "I'm not trying to be a troll" - ...? ... it did cross your mind, didn't it ?​
    Frank, the only thing that crossed my mind was that people might think that I was a troll.
  8. Pixel density has nothing to do with overexposure - metering does.​
    I understand that, Elliot, but is overexposure the only thing coming into play when you have 24 mp spread across a 1.5x crop sensor? That is, given that overexposure per se is what it is, is it not possible that something else in addition might be happening in the case of the D3200? First, is it not possible that the pixels are so tightly packed that they are interfering with each others' signals more? Second, is it not possible that the appearance of "bleeding" or "smearing" is likewise intensified due to the density of the pixels? If so, the effect should get worse in a hurry as the ISO goes up--and disproportionately so, compared to cameras with lower pixel density. (I did not shoot the D3200 at higher ISOs.)
    I am just speculating here, of course. I am just wondering if there might be more than one factor affecting image quality in contrasty situations with the D3200. On the other hand, given Joe's claim to have seen the same thing (or something like it) with the D800, perhaps you are both right that the only relevant issue here is metering. I really have no idea, and the issue with the D3200 is strictly academic for me at this point, since I have sold mine, whereas what happens with the D7000 and the D800 is not.

    is Rodeo Joe'ss thread, in any case.
  9. Subject: Do some Nikon DSLRs really have a slightly greater tendency to "overexpose in bright contrasty conditions" than do those of other brands?

    My D3100 has a tendency to compress highlights when shooting in raw. In fact, I thought that there was something wrong with the first one I got and returned it for a replacement. After doing additional tests, it seems that there is less smooth variation of tones available in the highlights as compared to the rest of the tones. I found that lamps have glow rings around them that are a solid color, almost like a JPeg effect. It might due to the fact that the raw files are only 12 bit and don't have the range necessary to show smooth transitions in the highlights. To protect the highlights, I always have a negative compensation dialed in. Shadow detail is often recoverable, but blown highlights have no detail and just turn grey if I try to recover them.
  10. Landscape Photography? I don't use Matrix metering for landscape, instead using spot and I also bracket.
  11. My D200 exposes fine. I use the center weighted meter myself and adjust my exposure if I need to prior to the shot.
  12. From my own subjective experience I should say that the D200 and D700 have no such tendency. As an ancedote however, I recall how sorely disappointed I was -coming from an F4 which handled this well with matrix metering- to find that the next-generation F100 I bought frequently caused severely underexposed slides in backlit scenes. I believe this was discussed in this forum sometime during the late nineties. So there´s really nothing new under the sun, though it is overexposure this time round.
  13. Thanks to everyone for redirecting me back to exposure pure and simple--and to the importance of getting it right.
  14. From my own usage and experience, my D80, D90 and D700 do overexpose in such contrasty lighting conditions. My D50, D3200 and D300 do not.
    It's fairly predictable, so live with the Exp. Comp. I need dialed in for the appropriate lighting.
    NB. I'm saying my copy of these cameras have these tendencies and don't wish to imply a general trend, although it does add to the theme of this thread!
  15. My D80 tends towards over exposure in certain conditions. My D1h is slightly under exposed I believe to preserve highlight detail. My D300 seems to be quite good.
  16. One additional factor in the matrix metering "decision" is the placement of the focus point in the frame. Thom Hogan had a discussion on this when it first appeared (maybe D3/d300 era). If you set the camera on the tripod with a contrasty scene, the exposure varies by whick focus point you use. IE my dog is next to a deck door inside the house. It is dark inside and bright outside. Without moving the camera if I shift the focus point from the dog (dark) to a flower pot on the deck outside (bright) the exposure darkens by one stop even though the scene in the camera has not changed. The camera assumes that the focus point you choose is of more interest, so weights it higher in the algorith. I guess a pure matrix metering would not do this.
    In the legendary "Where's my sky, Dude" picture (no offense Rodeo Joe. I'm really just joking with you) if the focus point is on the darker foreground the sky will more likely blow. I think if the focus point was moved to the sky(without changing acual focus) more sky detail would be preserved and the foreground would be darkened. Alternatively, active D lighting will sort of do the same thing. I must admit in practice I forget this and blow some highlights here and there but my landscapes don't run fast so it doesn't usually bother me.
  17. In high-dynamic-range (harsh-lit) scenes, default processing does not have enough DR. Therefore there is no correct exposure - you either blow highlights or shadows or a bit of both. Nikon cameras mostly do the last of these, go somewhere for the midtones (generally; then there are further things such as the active-focus-point bias). Notice the midtones strategy is just fine if you use Auto Exposure Bracketing for HDR. I would prefer a configuration option to make the meter more protective of the highlights, or an additional "ETTR" metering mode.
    Mind the "Picture Control" has a strong effect here. AFAIK the PC does not affect the meter, but setting the Vivid PC (generally, pushed contrast and saturation) you are much more likely to blow highlights than with Standard or Neutral.
    ADL is an additional story here, and a highly disputed one due to the lack of documentation. It messes with exposure in two opposing ways. The software (instant processing) part of it administers a fair dose of fill-light, which mainly substantially lifts the shadows but also moves the highlights up by about a stop. The other thing is the underexposure of the raw data, which appears to just about cancel the effect of the first one. Both together, ADL doesn't seem to have much effect on blowing highlights (but it does transform the histogram in the highlights in specific ways) when shooting JPG. If you shoot raw and tweak exposure watching the RGB histogram, the instant processing is likely to fool you into underexposing raw data. But if you shoot raw in varying light and don't have time to adjust EC for each shot, the ADL underexposure counters the tendency of the meter to blow highlights. Notice the ADL implementations also appear to change with camera generations.
  18. Metering on some more recent Nikon bodies (D7000) do IMO tend to expose more overall v some of the older ones. Even the newer D800/D600/D3200 bodies do too.
    The metering is also biased to the AF point in matrix mode so your average scenic shot 2/3 land 1/3 sky will likely be a bit too hot and blow out in some cases. Nikon are not alone in high contrast scenes most cameras will blow the sky, but overall I would say Nikon do give more exposure v other makers.
    I never rated the D7k's metering much it was heavily biased to the AF point and if you took shot of someone in a black suit for example would over exposure almost 2 stops in some cases.

Share This Page