Consistently Underexposing Images, Please Help!

[ant_nio_gomes]

<p>Hi Tom,<br /> Are you considering how you camera light meter "reads" the spot measure? <br /> Sorry for saying this as you indicated you already read about the procedure, but some digital cameras may "read" in a way to avoid highlight burning, i.e., when you read a highlight area it will consider it as zone V instead of zone VI and if you apply a compensation of 2 stops you will end in zone VII and this may be one of the reasons for the 1 stop underexposure you refer.<br />Does it give the same result when you use a matrix ( or a correspondent one, depending on the camera maker) measure?<br>

If this way you still get a consistent underexposure, you may compensate it but 1 or 1.5 stops seems a bit too much for that and you should contact the assistance.<br /> Usually I've my LCD turned off as a default and only check the images when I need, to avoid the temptation and look just at the image there and not the histograms.<br /> For sure you can adjust the luminosity of the LCD but I doubt it will be the solution as this one doesn't give you a constant image when the outside light changes - you probably had occasions you moved to a shadow area just to see something on it, and noticed how bright it becomes at night if your room light it is not a strong one.<br /> Regards,</p>

[simon_crofts]

<p>It sounds like the meter is correctly calibrated, so it should just be a matter of choosing the right mid-toned object to meter off. Or, if you meter off an object that is not mid-toned, learning to guage the correct amount of exposure compensation. Which is all a bit of a skill and requires practise and experience.#</p>

[tom_collins3]

<p>First of all, thank you so much for everyone's help in solving this problem, and/or educating me. Just to clarify, <strong>in my above tests I have taken the flash completely out of the equation while trying to make sure I understand my camera's light meter and the histogram produced. </strong></p>

<p>In regards to the comments to not use spot metering - help me understand this: when I took the test shots of my daughter inside on the couch using spot metering on the eye closest to me, the meter indicated a perfect "0" exposure, but the histogram was at least a full stop underexposed. <strong>I also tried this same shot using the evaluative metering mode</strong> and the meter showed that the same exact manual settings would produce a "correct" exposure, but the resulting histogram was the same amount underexposed (at least a full stop). So, this doesn't seem to be a case of metering off of the wrong spot unless I'm missing something. </p>

<p>While we're on the topic, when shooting available light only (no flash) what metering mode do most of you use (most often) for portraits and wedding photography, and how do you determine where to take your meter reading? I realize this will change based on extreme situations (harsh back lighting, etc). For available light photography (no flash) I was always taught to spot meter off of the eye. Is this incorrect?</p>

[ant_nio_gomes]

<p>Hi Tom,<br /> Your camera covers a 2.3º angle spot measurement, so the distance to the eye and if it fills the full measuring area or not may influence your readings. On the other hand, I guess you will get different measures if you target a light blue eye or a dark brown one, because the meter will see both as neutral gray and indicate you two measurements (even if the differences are not that big - I must confess I have no idea what it will be). <br /> As other comments point out, choosing the area for the spot measure is important but this doesn't mean you can not use spot metering for your intended purposes. You have just to adapt your exposure according to the areas you meter - for instance, the palm of a caucasian hand returns a reading you can consider as close to a neutral one, so if you take it under the same lightning conditions of your shooting situation you need no compensations. If this is not your case, you have to find the differences or just as plain game, use a 18% gray card instead.<br /> I tend to adapt to the situation and in some cases even to use a handheld meter, for 1º spot or incident light readings. But you can even disregard the meters and use Sunny 16 rule - if you know how to use it properly you will get a very high percentage of keepers (for natural light, obviously).<br /> Regards</p>

[kris-bochenek]

<p>I had the same thing and asked simillar question, I had my D300 JPEG set to Vivid with custom settings, my RAW files were pretty off from what in camera JPEG looked like. I changed to Neutral and the problem is gone.</p>

[jussihirvonen]

<p>I think this is not so much a case of lack of skill as it is a technical problem. I think there is something wrong with your camera. Go have it fixed, and you'll start getting perfectly exposed pictures. Your camera seems to be so consistent in underexposing the images that it must be a malfunction of some sort.</p>

[phule]

<p>A sample photo would really go a long way to helping diagnose this problem.</p>

[george_prescott]

<p>Does exposure compensation affect the meter reading in manual mode? I'm at work so I don't have access to my camera to check for myself. If so, it will cause the problem you're seeing.</p>

<p>George</p>

[tom_collins3]

<p>@Rob, I will try to post a sample photo where my meter says that the exposure is correct, but the histogram is underexposed. It will be later tonight. <br>

@George, I haven't found a way to dial in exposure compensation on the 7D while in manual mode. I see where to set exposure bracketing, but it appears that exposure compensation is greyed out in Manual and only available in modes like TV, AV etc. If anyone else knows where to check this on the 7D in manual mode, please let me know. </p>

[phule]

<p>Just to be clear: there is no such thing as exposure compensation in Manual mode. You alter the exposure by directly changing either the shutter speed, the aperture, or the ISO.</p>

[tom_collins3]

<p>@Rob, that's what I thought, but I wasn't 100% sure. I stated that I was shooting in manual, so I was confused when everyone kept asking about exposure compensation. Thanks for the clarification.</p>

[ant_nio_gomes]

<p>Hi Tom,<br /> I assume that when you look at the viewfinder you can see a bar going from -3 to +3 with a 0 (zero) at the center and you use this 0 to judge your exposure to be spot on.<br /> This bar is divided by small dots corresponding to 1/3 of a stop and going along it both sides is the way you can use to compensate your exposure in manual mode.<br /> Regards</p>

[tom_collins3]

<p>@Antonio, yes, that is the scale that I'm using to determine proper exposure, and the issue that I'm having is that when I set my manual settings to set the meter to "0", my histograms are at least a full stop underexposed. I'll post some pics tonight. </p>

[ant_nio_gomes]

<p>Hi Tom,<br>

Yes I got it from the beginning. My last post was meant to answer your question about exposure compensation in manual mode - when you get to 0 and want to compensate the exposure just change speed or aperture in a way the meter goes, lets say, to the second little point before +1 to compensate by 2/3s of a stop. Obviously you do this while the camera still pointing exactly to the same point.<br>

At least this is the way I compensate for manual spot measure with the D700, but I think it will be also the way for your Canon.<br>

Regards</p>

<p> </p>

[cjk]

<p>Tom: <br>

Spot metering will give you the correct exposure for a small area of the photo (your daughter's eye or face in the example you gave).<br>

I might be wrong but the histogram describes the exposure of the <em>whole</em> picture so it's not surprising that they are different.<br>

2 things you could try:<br>

- Fill out the frame with an 18% grey card and check the results<br>

- In the case of your daughter's photo, crop the photo to the area you spot-metered on and check the histogram then.<br>

-CJK</p>

[phule]

<p>[[exposure compensation in manual mode]]</p>

<p>Exposure Compensation is a specific function on Canon (and many other) cameras. It is a method to alter the exposure and override what the camera has selected automatically in Av, Tv, or P. </p>

<p>If you are metering in manual mode and adjusting shutter speed, aperture, and/or ISO, you're no longer using Exposure Compensation. You are simply altering the exposure.</p>

[mwtphoto]

<p>you are right Cesar, the histogram evaluates the whole scene so no wonder that the histogram is "off" compared to the light meter. Tom took a shot of a tan wall and the meter and the histogram where the same. Throw some dark items in the shot and spot meter off the tan wall and you will see that the histogram will appear underexposed compared to the meter.</p>

[picturesque]

<p>Your answers are scattered among the responses you received. The following are of note:</p>

<p>1. The histogram shows the values for the entire scene/frame, not just from within the metered spot. Histograms can vary and still be 'correct'. Judging whether a histogram is one stop off still involves judgement.</p>

<p>2. Reflective meters want to make the world middle gray. If you metered your daughter's eye and 'zeroed' the meter out, your result was correct to make the values within the spot meter circle middle gray. The result is not necessarily correct for the entire scene. If your daughter has fair skin and light eyes, eyelashes and brow, definitely, you would have had to manually/mentally compensate, or you definitely would get an underexposed scene. Caucasian skin in the metering circle by itself, would skew your results.</p>

<p>The patio chair you metered--what color/value was it? If it was a light color, you will get underexposure for the whole scene. If you are not compensating in your head as you spot meter various subjects, you aren't metering 'correctly'.</p>

<p>3. Nothing about metering is precise, except within the parameters you set. In the scene with your daughter, maybe you want her face to be correctly exposed, and the surrounding scene is a lot darker or lighter than her face. What is a 'correct' exposure setting and a 'correct' histogram? Maybe the image you want--with her face correctly exposed--produces a histogram that doesn't fit the classic 'correct' histogram.</p>

<p>4. I don't use evaluative metering at all because you are accepting the camera's automatic response to the scene and at least, I--don't have have any idea what the camera is using for parameters. When I use in camera metering I use averaging or center weighted metering. At least the concept is simple to understand and I can 'kind of' predict the results based on the values of what is in the scene.</p>

<p>For weddings, events, and all things, actually, I use a handheld, incident meter. As above, when I use in camera metering, it is center weighted. For incident reading you put the meter at the subject's position and meter the light falling on the subject.</p>

<p>Spot metering can be very useful for weddings, since the bride and groom are usually wearing white and black. However, you need to know how to apply the results you get from the spot metering. Deliberately pick your targets, knowing 'kinda' what value you are dealing with. Know how the camera will see that target value, and know how to compensate for it. If I didn't have my incident meter, I might use spot metering, and meter targets close to middle gray in value to get my general exposures.</p>

<p>I've never heard of spot metering off the eye before. I've heard of spot metering skin. You can even meter your palm, because palm skin stays relatively the same value, and you can precisely determine exactly how much off middle gray your palm skin is. Mine is 1 and 1/3 stop lighter than middle gray.</p>

[tom_collins3]

<p>Thanks Nadine! This definitely clears some things up!<br>

The patio chair is a darker brown color and the background was lighter grey concrete. I may be getting "spot metering for the eye" and "focusing on the eye" confused. I've seen the "metering of a palm trick" and that's a good one. Also, sounds like I need to use my gray card more!</p>

[simon_crofts]

<p>With the wall test I described it doesn't matter at all whether the wall is pure white or 18% grey, or for that matter, black - either way if the meter is working correctly, it should give you the exposure needed to represent the wall as the same density as 18% grey, and the histogram should bunch up in the middle. It did in this case, so that is pretty much absolute proof that the meter is working correctly. The only wrinkle there is that you might want to check that it does the same in lighter and darker lighting conditions, that an error isn't creeping in for some reason in a darkly lit room or in bright sunlight. Just find a wall in a darkly lit room, and one in bright sunlight, and double check - if it bunches in the middle, then all is OK with the meter.</p>

<p>Spot metering off the eye is not a good approach. The eye has a range of tones, from black pupil to white eyeball, and a range in between. They eye may be shaded by eyebrows or it might not be. The brightness of all these tones might just if you are lucky happen to cancel each other out and you might end up with a correct result, but this kind of random result isn't what spot meters are about.</p>

<p>Light caucasian skin is typically around a stop lighter than mid-grey, so if you're metering off skin then you need to add around a stop of exposure. If you are getting regular underexposure that might go to explaining those results, <em>if</em> you're emtering off skin. But you will also get radically different results whether you meter off skin on a sunlit side of the face compared to the side of the face that is in shadow, the results could easily be two or three stops different. So you really need to understand how the spot meter works and know how to interpret the results. It isn't a matter of pointing it at a certain spot on the subject and dialling in what is recommended by the meter. The information is just a starting point that needs interpreting.</p>

<p>Personally, I think the best thing that a photographer can do is read up on and understand the zone system, as expounded by Ansel Adams, it may seem a bit intimidating at first sight, but really it's not that bad, and quite clear and logical. To use a spot meter effectively, the kind of concepts it's talking about ought to be nice and clear to you.</p>

[simon_crofts]

<blockquote>

<p>and the background was lighter grey concrete</p>

</blockquote>

<p>If for example you had spot metered off this concrete, then that would explain the underexposure. And spot metering off skin, especially skin lit by sunlight, would also cause one or (sunlit skin) two stops underexposure.</p>

[john_deerfield]

<p>In terms of spot metering <em>techniques </em>it is helpful to know that highlights are about 2-2/3rds stops above middle gray. And black (with detail) is 4 stops below middle gray. You can test <em>your</em> cameras exact dynamic range/parameters using a target card. In the shot below, I used a gray target. The top is "correct" 18% gray. For bright white (protected highlights), I simply made my EV 2 2/3rds stops brighter, the 2nd shot (same gray target). For black, with detail, I simply changed my EV by -4 stops, the last shot, same gray target. If there is something white I need to protect in the frame, I simply spot meter on this and add 1-2/3rds to 2-1/3 stops depending on how "white" I want to render the white. In large this depends on how my highlights are looking. I could do the same in reverse and spot meter on black, but for me personally, I usually use white.</p>

 

[john_deerfield]

<p>In terms of spot metering <em>techniques </em>it is helpful to know that highlights are about 2-2/3rds stops above middle gray. And black (with detail) is 4 stops below middle gray. You can test <em>your</em> cameras exact dynamic range/parameters using a target card. In the shot below, I used a gray target. The top is "correct" 18% gray. For bright white (protected highlights), I simply made my EV 2 2/3rds stops brighter, the 2nd shot (same gray target). For black, with detail, I simply changed my EV by -4 stops, the last shot, same gray target. If there is something white I need to protect in the frame, I simply spot meter on this and add 1-2/3rds to 2-1/3 stops depending on how "white" I want to render the white. In large this depends on how my highlights are looking. I could do the same in reverse and spot meter on black, but for me personally, I usually use white.</p>

[john_deerfield]

<p>Duplicate post?</p>

[john_deerfield]

<p>And the image never attached...</p>

<div></div>