D700 highlight clipping

Discussion in 'Nikon' started by photo5, Nov 19, 2008.

  1. I shot with my new D700 over the weekend, and used my beloved Nikkor 28mm f2.8 AIS lens, hoping for some great
    results. Inside was OK but outside in the bright sun I got a very strange result. The highlights are clipping
    severely. I don't think my D300 would have done this, but the D80 would have. I'm pretty shocked to say the least
    that a pro camera from Nikon would do this! I did not shoot a RAW of this scene, I was just out shooting JPGs.

    Has anyone else noticed this behavior with the D700, and is this a bug? It's not the fault of the lens, to my
    knowledge.<P>

    <img src="http://hull534.smugmug.com/photos/420724234_ZdEJh-L.jpg"><P>
     
  2. Could we have some more information on the file?
    Metering? Aperture and shutter? Any exposure comp?
     
  3. Hello Dave,

    I am new to the D700 too and my first impression in contrasty outdoor situations is the same, although I shot raws. They can be
    adequately corrected afterwards on the computer, so all is not lost. But I am convinced, there is an in-camera setting as well hidden
    somewhere in the menus. Eventually I'll discover it after spending time with the manual.

    But I agree with you, it's a strange out-of-the-box setting for a professional camera which I have never experienced with my trusty D200.
     
  4. I have experienced this issues on all my cameras, D300 and D200 included (I feel I`m continuously checking and adjusting exposure), don`t know if more or less than with the D700.

    Don`t know how the D700 metering system & contrast control works but looks like to have this highlights under control the rest of the image will probably be pushed to the dark, resulting in a too dark scene. I suppose you were using the matrix meter mode. Also, don`t know how your picture
    settings are. Did you check the hystogram?
     
  5. Use the Active D-Lighting mode if you have like these. The ADL mode basically exposes for highlights and applies a tone curve to the image.

    That, or use exposure compensation.
     
  6. I've had 9 DSLRs over the years and not one of them (including the D200) would have left the highlights unclipped shooting outdoors without some kind of exposure compensation intervention.
     
  7. You really have a perfectly exposed shot here. The bright areas on the engine nacelles and along the fuselodge are what is called specular highlights.

    A pola filter would have helped maybe.

    Turn on the Active D-lighting in the retouch menu. That will help some. If that does not work, then then use the best exposure you can get and use the highlight/shadow controls in photoshop or Nikon NX2.

    Lowering the exposure and bringing up shadows in post processing is another option.

    One more option for what you have right now, is to make a blended exposure. Go to thelightsrightstudio.com, digital darkroom tab, ten look at the quicktime move on blended exposures.
    The technique is to merge a processed underexposed version with a normal and use a luminescent mask to do auto blending.

    Last option is HDR.

    Nice pic of a B47. Where did you find it? It is amazing when you realize this is made from late to mid 1940`s technology.
     
  8. For a camera that is said to have a 12 stop dynamic range I find extraordinary that there would be blown
    highlights in what seems to me to be no more than a 5 stop latitude scene. OK, 6 stops max.

    One could of course suggest that incorrect exposure is what is at issue here, but I could argue that the scene is
    correctly exposed. The shadows look right given that there's plenty of sun and nothing in the sky look washed out
    which is what would have happened if there was an overexposure.

    Dave, you know as well as I that if you had set the camera to ADL there's almost no chance of getting blown
    highlights
    in this image.

    I don't have a D700 but do have a D300 and this "problem" is consistent on the D300. That is, I sometimes get
    blown highlights when ADL is off but when it is on for a shot like this I don't have any issues.
     
  9. ShunCheung

    ShunCheung Administrator

    I checked Dave's image in PhotoShop, and I got the histogram attached below. The exposure actually looks pretty good for this image. The body of the plane is sliver/white and highly reflective while the sun is shining directly on it. There is some specular highlight in the front part below the cockpit area so that some clipping is expected. If you want to avoid that, reshoot when the sky is more overcast.
    00RY7n-90275584.jpg
     
  10. When I first bought my D3 several months ago, I ran some side-by-side tests against my 5D. With hesitation, I sold my 5D as the IQ seemed pretty much identical. Yet, months later, until this weekend, I was unhappy with the overall IQ of my camera only when shooting outdoors in bright environments like your situation (for me, it is often the beach).

    While at the beach this past weekend, I used bracketing on a number of different shots and came to realize that my D3 tends to overexpose by about 3/4 of a stop. My 5D seemed to nail the exposure virtually all of the time which is one of the reasons why its pictures look so good. I have now set my exposure compensation to -.7 [for outdoor shooting] and am finally delighted with the stunning color and high overall detail I am now getting. There is no substitute for the perfect exposure.

    I suggest the next time you are shooting a similar subject that you bracket your shot and evaluate the results. You may be pleasantly surprised and what your camera can do for you if you get the exposure right. D-Lighting or equivalent will 'fix' the shadow areas if needed.
     
  11. I don't think you should be using matrix metering on a stationary subject like this - take an incident reading or
    spot meter the highlights of the plane and place it at +2 or thereabouts.

    RAW helps a lot with dynamic range and highlight recovery. Use Nikon Capture NX2 to get the best out of the
    camera in this respect.

    Matrix metering may be less accurate on lenses that don't have a CPU since the without lens identification, the
    camera doesn't know the wide open falloff pattern.
     
  12. There are limits to what can be done with automated modes of reflective metering, I usually shoot in manual mode when the light is constant. That way you only need to nail the exposure once.
     
  13. Here is an example of what can be done to salvage a difficult shot:

    The top shot was -1 EV, the middle shot was as metered.

    The bottom shot was 'fixed' in CS3. These were JPG files. You can expect even better results with RAW files.
     
  14. image here:
    00RY9w-90289584.jpg
     
  15. "For a camera that is said to have a 12 stop dynamic range I find extraordinary that there would be blown highlights in what seems to me to be no more than a 5 stop latitude scene... "
    It must be taken into account several parameters: RAW or compressed .jpg, bit depth, and in camera .jpg settings. Probably, shooting 14bit RAW and adjusting exposure we could have an almost perfect shot. Don`t expect 12 stop latitude shooting with different settings. Shooting .jpg, and with whatever the contrast setting selected that range could be reduced to your 5-6 stop range (lower contrast=higher range, higher contrast=shorter range). That`s absolutely normal, don`t burn your D700s till now... ! :)
     
  16. Looking at the hystogram provided by Shun, I`d underexpose a bit, and adjust contrast during PP if needed.<p>
    What Elliot says.
     
  17. I found this too and wonder why Nikon chose such a contrasty default tone curve. I shoot RAW, though, and use Neutral
    Picture Control with slightly boosted saturation. I adjust in post to achieve whatever tone curve is appropriate for the shot.

    If you want maximum dynamic range, shoot RAW.
     
  18. Thanks for all your responses. I was shooting only JPG as I go to the Museum of Flight near Seattle often and these are just informal test photos. I had the camera Active D-Lighting set to 'Auto', no EV adjustment, ISO 200, aperture priority, Matrix metering.<P>

    Here's another photo, equally troubling. I shot for a year with my D300 and never got these kinds of results that I can remember. Same camera settings.<P>

    <img src="http://hull534.smugmug.com/photos/420723811_FFYkv-L.jpg"><P>
     
  19. It could just be that for whatever reason the camera overexposes by one full stop with 28mm f2.8 AIS, though my Nikon D300 made perfect exposures with it. I shot the same airplane six minutes later, granted from a different angle and the sun may have been a little more diffused, with the Nikon 50mm f1.8 AF-D and got this result:<P>

    <img src="http://hull534.smugmug.com/photos/420726674_JsV2z-L.jpg"><P>
     
  20. Overexposed, period. I mean, to have so many details in the shadow under the plane (second photo) you have to overexpose
    the highlights in such a contrasted scene. The camera meter is not magic. In some extreme conditions, you have to interpret and
    compensate, using CW or Spot. Many people have settled to a permanent -1/3 to -1 full stop with the D700. That's the nature of the beast.
    Try
    this same photo with slide film, you'll probably get even worse results. But, who said the D700 has a 12 stops dynamic range? I've never
    seen that. Isn't that the equivalent of being able to correcty meter a scene which would require, say, f:2.8 at 1/60 for the shadows all the
    way to f:22 at 1/250 for the highlights, or thereabout?
     
  21. I found this too and wonder why Nikon chose such a contrasty default tone curve.
    The reason is that many people love punchy, high-contrast, saturated photos. It sells cameras. Nikon is able to use a contrasty curve because of the low noise of the camera. If you want a lower contrast result you can always adjust the picture control settings (either in camera or in post) and the lowest contrast settings available are really low contrast. In addition you can use the active D-lighting feature. I usually don't - I go for a high contrast result a lot of the time, but that's just my personal preference.
    Dave, this is a theory that I led to believe in after I realized that only the central spot meter is available when lenses without CPU are used. The reason Nikon doesn't allow the peripheral spot meter points to be used (the indicator moves but measurement is always through the center point) is that there is a lens- and aperture-dependent "coupling coefficient" which describes how the camera has to compensate the peripheral spot meter readings for each lens and aperture. With each Nikon lens that has the correct CPU in it, the lens is identified and the compensation applied. With the CPU data, the matrix meter is more accurate and off-center spot meter points can be used. This is one reason I am not so hot about "chipping" lenses as the correct information might not be given to the camera (when the wrong chip from another lens is used). I believe it's one of the main reasons why Nikon has been reluctant to allow matrix metering on non-CPU lenses. Many users insisted on matrix metering with MF lenses and Nikon finally gave in, but the consistency of the matrix meter is not the same as with CPU lenses.
    I don't think this is a big deal really. When I use the matrix meter I understand that the result is not completely predictable and I shoot RAW so I have some headroom for post-processing adjustments. Whenever I am shooting still subjects I use the spot meter though - it measures through the center of the lens which is the least affected by possibly less than nominal transmission at wide open aperture.
     
  22. You are shooting a highly reflective surface in direct sun. A specular highlight is a point/spot in a scene that directly reflects a light source (sun) into the camera. In this case, entire lengths of curved surfaces of polished aluminum along the planes are functioning as specular highlights.

    The general rule is to ignore speculars when metering as they are basically pure white and contain little to no detail. Your camera seems to be doing a pretty good job of this on matrix (and I'm sure you can get more detail out of raw). If you actually spot metered for the specular highlights, the rest of your image would be very very dark. Point your camera directly at the sun and spot meter. Then do the same for a dark object in shade. You will indeed see well more than 12 stops of range. As Shun said, if you want no portion of a shiny reflective surface in your shot to hit pure white, photograph on an overcast day. Another option is to do HDR bracketing.
     
  23. ShunCheung

    ShunCheung Administrator

    Dave, your 2nd example (posted at 11:05) looks fine to me. Only a couple of areas with specular highlights are clipping. You can easily verify that with thresholds in PhotoShop (I am attaching it below). See it shows the clipping areas. Your 3rd sample posted at 11:13am is simply underexposed, as the silver/white plane looks gray. In these situations, I would spot meter the bright areas of the plane and make sure that it is 2 to 3 stops over medium. In the old days when I shot slide film, it would have been 2 stops over.
    00RYIS-90333584.jpg
     
  24. Thanks for your replies. The airplane is grey, not silver or white, so the photo taken with the 50mm looks correct to me. Here is a photo taken in 2007 with my D80:<P>

    <img src="http://hull534.smugmug.com/photos/166099940_eFz2J-L.jpg"><P>
     
  25. ShunCheung

    ShunCheung Administrator

    Dave, since only you were there, you should know best. But those images (including the one from the D80) look underexposed to me.
    Not only does the plane look unnaturally gray, the sky, the lawn, the wheels, etc. all look underexposed.
     
  26. "Don`t expect 12 stop latitude shooting with different settings"

    12 stops goes something like this:

    1/8000, 1/4000, 1/2000, 1/1000, 1/500, 1/250, 1/125, 1/60, 1/30, 1/15, 1/8, 1/4, 1/2 sec (that's right there are
    13 stops here but the latitude is 12)

    If someone can tell me such a scene I would be very much interested in seeing how a D700 - or
    any other digital camera - would capture the entire range with NEF 14-bit uncompressed, or any other combination.

    I doubt that even a black cat sitting inside a cave with a midday sun or something similar would give 12 stops of
    latitude.
     
  27. Shun,

    It's about a half-stop under, you're right. But it's a grey airplane and the camera meter makes anything grey. Next time I'm down there I'll shoot RAW and see how it goes. I need to lower the default contrast settings on the D700 when I'm out shooting JPG's. At 12mb a shot, shooting RAW files is taking a toll on my available hard drive space!!
     
  28. "At 12mb a shot, shooting RAW files is taking a toll on my available hard drive space!!"

    You should try scanning medium format film. Each 67 scan at 16 bit depth is over 500 MB.
     
  29. ShunCheung

    ShunCheung Administrator

    Come on, I have been using DSLRs since 2002 and have been shooting Nikon NEF since day 1.
    So far all of my digital images can fit onto a 1T hard drive with some room to spare. (Yes, I have multiple copies, probably too many copies, store at different locations.) Today, you can get a 1T external hard drive for under $150. Disk space really should not be an issue.
     
  30. "If someone can tell me such a scene I would be very much interested in seeing how a D700 - or any other digital camera - would capture the entire range with NEF 14-bit uncompressed, or any other combination... "
    I don`t bother so much about this latitude issues and tests, as you say is a really wide range... (I don`t know if it is true, I cannot affirm this!). I suppose this are results of sensor testing laboratory procedures with calibrated grey scales. This entire range could not be used on the real life, but could be really useful: Think on the image of this thread; even worst, think that a stronger overexposure is produced; then, shooting with this extended dynamic range will easily save your pic from that clipped highlights!
     
  31. Shun, of course you're right. I've two externals next to my Mac, a 250gb and a 500gb (the 500 is my Time Machine backup drive). I suppose I could retire the 250gb (which is about 9gb from full) and keep it in the closet and get a new 1tb drive. But my real issue is, how many RAWs do I need of this darn DC-2 parked outside the museum of flight? :) It's so beautiful that every time I walk by it, I have to take a photo of it.
     
  32. Whew...and I thought it was just my Minolta 7D that had these problems. From all the responses, and plethora of work arounds, I see it is a common DSLR issue. The thought of having to massage all my outdoor shots taken on sunny days gives me a headache...its times like this I am glad we still have film;) .
     
  33. ...or perhaps I should say...the thought of having to fix all my sunny day shots sounds like about as much fun as scanning film, well, maybe a little better than that, but not by much.
     
  34. It's a common JPG issue.

    I did some more research from images I've shot in the past, and found an identical one shot with my D300, and the JPG looked the same. So it's not the D700. And certainly, a RAW does capture way more information. I guess I just need to get off the JPG horse and shoot RAWs with this camera. Thanks again for all your responses.
     
  35. "Expose for the shadows and develop for the highlights" is still a valid dictum using today's DSLRs. You have to shoot raw in order to do this, of course. You "develop" the image with your raw converter instead of a developer solution as with film. Several have already pointed out the in raw conversion you can recover much of the highlight detail. In my experience it would be preferrable to capture the shadow detail you want and recover the highlight detail in the conversion rather than underexposing and then trying to "pull up" the shadow detail, which typically results in increased noise due to underexposure. You'd get the same problem with underexposed film. Some things never change, so it seems.
     
  36. According to dpreview, the D700 has 12 stops in RAW (14bits), and 5 less in JPEG. Less dynamic range means a steeper S-shaped curves (=sigmoidal exposure curve). It clips your highlights (value 255 for each channel), and/or makes your shadows black value 0). You get better, snappier contrasts, though.

    When setting your D700 to "NEUTRAL" and reducing contrast you will get the maximum DR in your JPEGs, but probably they will look more boring.

    With Nikon Capture NX2, the only chance to get the highlights back is to decrease exposure, and turn up D-lighting. But any further processing step in NX2 only operates only on a 8 bit range.
     
  37. I've just downloaded the 60 day trial for NX2, and will check out my recent RAW captures from the D700 with it. So far Adobe Camera RAW has many limitations that NX2 will be able to get over in tricky lighting conditions. I was amazed with the accuracy of the D700's meter when shooting over the weekend. It knows how to expose for a bright white surface without making it grey, as most other cameras will. Must be finding a similar scene in its memory banks. I found I didn't need to dial in any compensation at all.
     

Share This Page