New D7100 underexposes

[owen_farmer]

<p>Sorry for the confusion in the two prior posts. This is the first time I am attaching photos to a post.<br>

Owen<br>

<br>

<img src="http://d6d2h4gfvy8t8.cloudfront.net/18090925-lg.jpg" alt="DSC_0160.JPG with Nikon 16-85 VR lens" width="500" height="334" border="0" /></p>

<h1>DSC_0160.JPG with Nikon 16-85 VR lens</h1>

<p> <br>

<img src="http://d6d2h4gfvy8t8.cloudfront.net/18090926-lg.jpg" alt="DSC_0226" width="333" height="500" border="0" /></p>

<h1>DSC_0226.JPG with Nikon 16-85 VR lens</h1>

 

 

[Matthew Currie]

<p>Have you tried this with a different lens? </p>

<p>I notice from the EXIF info that the dark shot was done at shutter priority, and the lens was wide open. If you were using manual ISO there was no way to increase exposure. If you were using auto ISO, is it possible that the upper limit of ISO was capped by mistake? </p>

[elliot1]

<p>There are numerous factors which can come into play when it comes to exposure. You may find this previous thread of interest:</p>

<p>http://www.photo.net/nikon-camera-forum/00aKLZ</p>

<p>Ultimately, if you shoot RAW and get an underexposed image, you can easily balance the image without blowing out the highlights. It is possible with JPG images, but is not as easy and not always as successful. Ultimately you want to get the exposure right or as close to being correct as possible in camera.</p>

<p>It is also important to understand that under many shooting circumstances, especially when the scene has strong bright and dark areas, that the camera's meter can be not as accurate as you want it to be, which is where spot metering becomes quite advantageous.</p><div></div>

[ShunCheung]

<p>EXIT data show that Owen's first image was captured at f3.5, 1/250 sec, and ISO 250. That is about 4 stops under sunny 16, and I can see that maybe underexposed by a stop to two.<br>

<br />I would double check the meter under "sunny 16" conditions and see whether it is giving the right readings.</p>

<p>In one of the other two threads Owen started (and now deleted), he mentioned that exposure is sometimes right but sometimes not. I would double check that exposure bracketing is not unintentionally switched on. If so, it may explain the situation.</p>

[bebu_lamar]

<p>I wonder what's the metering mode? Matrix, CW or Spot?</p>

[bebu_lamar]

<p>I see that's in matrix mode. But does the exif indicate underexposure if in shutter priority and the lens is wide open and still not sufficient exposure?</p>

[chuck]

Sunny 16 would say using ISO 250, the exposure should be 1/250 at F/16, would it not? That would say a image

exposed at f/3.5 ought to be 4 stops over exposed, not underexposed?

[Matthew Currie]

<p>Sunny 16 dictates the exposure in full sun. This is a shady shot, so one would have to compensate accordingly. </p>

<p>What Owen has not specified is whether Auto ISO was on or off. If it was off, the camera simply ran out of options when the lens went to 3.5. It could do nothing but underexpose. If it was on, then it could be some other problem. </p>

<p> </p>

[ShunCheung]

<p>Sunny 16 is merely a known exposure: under sunny conditions around noon time, aperture setting should be f16 when shutter speed and ISO settings are the same, e.g. 1/100 sec for ISO 100, 1/250 sec for ISO 250, etc. If Owen's D7100 is giving a different reading, especially if it is off by more than a stop, something is definitely wrong.</p>

<p>It is something that is very easy to check, on film cameras or digital cameras alike, and you don't even need to capture one frame. That is why people have been using this method for decades to check meter readings.</p>

<p>However, I would double check other settings such as bracketing, exposure compensation, etc.</p>

[bebu_lamar]

<p>I wonder in the case that when you set the camera on shutter priority and it simply run out of aperture would the EXIF indicate this condition? </p>

[ShunCheung]

<blockquote>

<p>I wonder in the case that when you set the camera on shutter priority and it simply run out of aperture would the EXIF indicate this condition?</p>

</blockquote>

<p>But at that time the viewfinder should have shown a clear error, indicating underexposure. Of course, one might overlook that.</p>

[pete_s.]

<p>Those images are <strong>perfectly</strong> exposed.<br /> In picture 1, the bright building in the back hits RGB 255.<br /> In picture 2, the white pants and the white foldable chair hits RGB 255.<br /> If the camera had exposed more than it did, it would have overexposed and blown out those areas.</p>

<p>This is just a case of the photographer wanting something else than a perfect exposure.<br /> This is why spot metering was invented. This is why exposure compensation was invented. This is why cameras comes with a manual mode. Someday when the camera is smart enough to figure out <strong>exactly</strong> what we want then we only need one button on the camera and nothing else.</p>

<p> </p>

[ShunCheung]

<blockquote>

<p>Those images are <strong>perfectly</strong> exposed.<br /> In picture 1, the bright building in the back hits RGB 255.<br /> In picture 2, the white pants and the white foldable chair hits RGB 255.<br /> If the camera had exposed more than it did, it would have overexposed and blown out those areas.</p>

</blockquote>

<p>That is complete nonsense concerning the first image.</p>

<p>99% of the people would find it totally acceptable that the white building in the corner of the background is blown out. That is an insignificant area in the image. The problem with the first image is that the main subjects are underexposed by one to two stops. The lighting was not challenging such that either matrix, center weighted, or spot should have done a good job.</p>

[pete_s.]

<p>If you want to check if your camera overexposes you set the camera to center weighted metering and fill the entire frame with an evenly lit matte gray or white card, paper or wall. Then check the histogram on the camera. If it has a spike a little to the left of the center then the meter is fine.</p>

 

[pete_s.]

<blockquote>

<p>99% of the people would find it totally acceptable that the white building in the corner of the background is blown out.</p>

</blockquote>

<p>That is completely irrelevant. This is about what the camera thinks.</p>

<p>And technically speaking the exposure is perfect. It's as hot as it can be without blowing out anything significant.</p>

<p>If this would have been a landscape image it would not have been acceptable to blow out those parts of the image. The camera can't know when it's acceptable and when it's not. It doesn't know what you want or what you are shooting. In matrix metering it's guessing what you probably want. That doesn't make it right or wrong.</p>

<p>PS. If the metering mode would have been centerweighted metering then we could reevaluate if the camera is underexposing or not.</p>

<p> </p>

[pete_s.]

<p>The exposure in image 1 is EV 10.3. Sunny 16 is EV 15 so it's exposed 4.77 stop under sunny 16.<br>

The exposure in image 2 is EV 13.6. Sunny 16 is EV 15 so it's exposed 1.3 stop under sunny 16.</p>

<p>As said above it quite possible that exposure compensation and other things where in effect as well. Best thing for the OP is to post the original images. Tools like Photoshop Elements doesn't keep the EXIF info intact when used for resizing.</p>

<p> </p>

[scott_murphy5]

<blockquote>

<p><a name="00dV2I"></a><a href="/photodb/user?user_id=24372">Shun Cheung</a><br /> It is something that is very easy to check, on film cameras or digital cameras alike, and you don't even need to capture one frame. That is why people have been using this method for decades to check meter readings.</p>

</blockquote>

<p>Shun your comment is overly generalized. The only way can truly <em>accurately</em> use this rule of thumb to check your meter's accuracy is to meter off of an <em>18% Gray Card</em>. Outdoor scenes can very <em>wildly</em> in EV range and what the meter calculates as exposure for an overall 18% gray (Zone V) can vary by just as much, regardless of whether it is spot, center weighted average or matrix. The only way to be sure it is right is if you use a commonly accepted <em>standard</em>, and that is an 18% gray card.</p>

<p>A rule of thumb we used to use is to meter off or your palm and open up a stop. But Caucasian skin can vary significantly from "Zone VI" and so it is not really all that reliable.</p>

[Palouse]

<blockquote>

<p>The exposure in image 1 is EV 10.3. Sunny 16 is EV 15 so it's exposed 4.77 stop under sunny 16.</p>

</blockquote>

<p>But yet you say it is perfectly exposed? I'm confused. Looking at a histogram of picture 1, I see a left-skewed distribution tight to the left edge of the frame. Clipping?</p>

[Matthew Currie]

<p>I looked at the histogram too, but what I saw was that, although indeed the majority of the information is at the left, there is a little almost all the way to the right, especially in the blue. </p>

<p>If the exposure were increased, then not only the wall, but the fence and the patch of driveway to the right, as well as the blue posts, would be overexposed, I think. That would be more overall real estate than I'd expect a matrix metering pattern to sacrifice. </p>

<p> </p>

[pete_s.]

<p>The histogram looks like it does because the scene has different light levels.</p>

<p>The background which is outside and lit by the sky and the subjects which are in shade and lit by reflected light. There isn't one exposure setting that would expose them both individually correct.</p>

<p>Let's say that you wanted to shoot a similar scene professionally and you wanted to include the background. Your only option would be to raise the exposure on the subjects under the roof by either using strobe lighting or using natural light and reflectors. Then you could find one exposure setting that would expose everything as one would have expected.</p>

<p>Quick fix is to shoot as it is in raw, have it exposed as it is and then raise the exposure a couple of stops while retaining the highlights. Since we didn't blow the highlights the only thing we suffer is added noise, which would probably be acceptable since we could shoot at low ISO and even a 2 or 3 stop push would only add minor amount of noise on modern Nikon.</p>

<p>Other choices would be to pick another background (in shade) by shooting from another angle, possibly going to a longer focal length and increasing subject distance (original was shot at 16mm).</p>

<p>Or just blow the background out and live with it. Or bring everyone outside.</p>

<p> </p>

[chuck]

Does this only happen with one lens?

 

Is it possible the lens is faulty and reporting the wrong maximum aperture to the camera?

[owen_farmer]

<p>Thanks for all the information.<br>

I am still getting used to this camera (previously I had a D90). I made 2 errors. I had the camera set for shutter priority, and I intended to use Program mode. Also, I intended to use Auto ISO, but did not. I don't believe that the EXIF data sent to photo.net includes whether Auto ISO is used. However, I still have this photo on the camera, where the ISO of 250 is displayed in BW, not red. I believe that red would indicate that Auto ISO is on.<br>

This is an argument for not deleting images from the camera until after the images have been reviewed. I use a Mac and iPhoto. iPhoto does not store Auto ISO on/off.<br>

Thanks for the help.<br>

Owen</p>

[ShunCheung]

<p>Another setting to check is Active D Lighting, which can sometimes underexpose a scene to protect the highlight area. However, I kind of doubt that it would underexpose your main subjects by as much as two stops.</p>

<p>I, for one, wouldn't hesitate to have some small overexposed areas inside the frame if that means my main subjects are better exposed. And I just switch Active D Lighting off.</p>

[Matthew Currie]

<p>The red ISO display tells whether auto ISO has changed your setting, but not whether it is on or off. If it leaves your setting alone, it will be in black.</p>

[rodeo_joe1]

<p>Ah! Pete, if only Nikon's metering really was intelligent enough to do ETTR exposure. It would quieten a lot of criticism of flakey Matrix metering.</p>

<blockquote>

<p>"The only way to be sure it is right is if you use a commonly accepted <em>standard</em>, and that is an 18% gray card."</p>

</blockquote>

<p>- Not so. Average, centre-weighted or spot metering is designed to render anything it "sees" as 18% grey. It follows that it really doesn't matter what shade of white you point the meter at. It can be white, black or any shade in between as long as it's a fairly neutral, uniform and evenly lit surface. The important thing is that the camera renders it as 18% grey. And an 18% card knows nothing of the light that's incident on it, so you would have to meter the incident light to know what 18% reflected light was likely to be. So why not use an incident meter in the first place?</p>

<p>It's also plain that Zone V and 18% grey aren't the same thing. That's according to St. Ansel's own writings on the subject, and the many film sensitometry tests that he, or his assistants carried out. Zone V is clearly 3 stops (or Zones) down from a surface that approximates to 100% Lambertian reflectance. 18% reflectance is only 2.5 stops down from such a reflector. Even if you say that Zone IX should be allocated to 100% reflectance, there's still a half-stop discrepancy one way or the other.</p>