How is "correct exposure" defined? Scene brightness vs image brightness.

<p>Circa 9 months ago I have started with 35mm film photography, I use a Canon EOS 620, mostly ISO 400 color negative film and ISO 3200 b/w film.</p>

<p>My question is, what exactly is meant by correct exposure, or how is correct exposure defined in the context of auto exposure?</p>

<p>Also, for a given, static scene, do all film cameras determine exactly the same shutter/aperture values (assuming that the lenses are the same/equivalent)? Or is this a decision influenced by aesthetic preferences of the engineers who programmed the auto exposure?</p>

<p>To put this question into context, I started thinking about auto exposure when I was doing shots at night in "Program" mode (using a tripod). This leads to very long exposure times. The resulting images are a bit as if I had shot them in daylight, i.e. they are much brighter than the scene actually looked. It is almost as if I had shot them through night vision goggles. Apparently, at least for the EOS620 correct exposure means to achieve a certain overall brightness in the resulting <em>image</em>; while what I would have expected, is an attempted approximation of the overall brightness of the <em>scene</em> itself!</p>

<p>Then I thought, actually I would need to shoot always with a constant shutter/aperture to reproduce brightness of the scene… because "correct exposure" apparenly <em>compensates</em> for light conditions while I want to reproduce the light as it actually was. (Sorry if this sounds confusing, don’t know how to explain it better).</p>

<p>"Correct exposure" in B&W typically is defined as photographing a 13-18% grey card at an exposure to reproduce a negative density sufficient to produce 18% grey print using predetermined chemical composition, temperature, development times, etc. (sorry I don't have the densitometric factor readily available). Auto exposure curves are slightly different for each manufacturer, but the latitude of negative films typically encompasses the differences so they are undetectable. Further, "correct" exposure metering, is highly dependent on the nature of metering...be it matrix, center weighted (in all its variants), or spot metering. Color negative film typically has much more latitude than positive (slide film). B&W film typically has lots of latitude. I personally think you are overthinking the issues...unless you do all your own development and printing according to standards you have established, you will probably have to rely on reasonably "proper" exposure rather than "correct" exposure...because there isn't a simple answer.</p>

<p>It is very simple: the correct exposure is the exposure that leads to the final image that you envisioned in your mind. As cameras cannot yet read minds, it is up to you to determine what the proper exposure parameters should be.</p>

<p>[[The resulting images are a bit as if I had shot them in daylight, i.e. they are much brighter than the scene actually looked.]]</p>

<p>How are you evaluating the images? In the consumer world, any machine prints coming from color negative film will likely be printed very different than what was captured. The machine will see what it thinks is a very underexposed negative (when what you simply have is a lot of black in the scene, as it is night) and will overexpose the print. This is done automatically.</p>

<p>As Stephen says.<br>

Meters in general try to find an "average" exposure. Some more complex cameras, even including ones like your EOS 620 have built-in programs to vary that for specific defined circumstances (picture modes, etc.)</p>

<p>That's why a snow scene, shot as metered, will usually produce grey/blue snow rather than white snow.</p>

<p>C/N film typically has 2+ stops latitude either side of "correct" so "sunny"-16 will almost always produce usable results. B&W about the same.</p>

<p>Slide films, may they rest in peace, were very much more limited in exposure latitude.</p>

<p>In film days, many photographers purposely underexposed a little to get more color saturation.</p>

<p>The science of measuring the light and setting exposure is as old as photography (first permanent image 1827). The first exposure calculator dates 1888. The first electric light meter, Western Electric, 1932. In the mid 1930's, Messrs Jones and Condit at the Kodak Laboratory published, statistically, a typical sunlit scene integrated to a reflectance value of about 18%. Kodak Labs publish a recommendation; place a Kodak film box in the scene. The box reflects18% of the ambient light. Measure the box top, set your exposure. 1941, Ansel Adams, a prominent landscape photographer and his friend, Fred Archer, a photo magazine editor, jointly published the Zone System replaces the box top with an 18% battleship gray placard.</p>

<p>The gray card.replaces the Kodak box top. This 18% gray target became the de facto standard. Today film and digital sensors are calibrated and film and ISO values are established using the 18% gray card.</p>

<p>Because of the pitfalls associated with reflected metering, a second measuring method evolved called the incident-light reading method. This method places a transparent sphere placed over the entrance of the light meter. The meter is positioned close to the subject and pointed backwards towards the camera. Thus, the meter measures the light just prior to striking the subject (incident old French word for about to happen).</p>

<p>As the science of light metering evolved, light measuring circuits migrated to become built-in. Add to these circuit, chip logic and camera acquired brains. Now we have a hodgepodge of methods to meter. Data can be spot, integrated, averaged, buffed, and manipulated. It’s an engineer’s delight. The stakes are high, camera makers place their reputation on getting an optimum image. In the end, it’s the skill of the photographer and his knowledge of light and shadow that triumphs. </p>

<p>Take 1 case to answer you as the full answer is very long. The case that you took the picture at night and the print looks like day. The answer is that all camera exposure system try to do it that way. Take a very controlled subject like a big gray board. The exposure system on all camera would try to reproduce it as a big gray board in any kind of light. Day or night.</p>

<p>Agree with above postings, but would add that you need to decide whether this is a camera exposure problem or a printing problem. If the shadows on your negatives of your night shots are dense instead of being clear film, and the highlights are so dense that you can’t see detail, then the negatives are overexposed. If not, but the shadows are a mushy gray instead of black, this is a printing problem – the automatic printer doesn’t “know” that the shadows are meant to be black. Get the lab to reprint.<br>

In general, a night scene with floodlit buildings and deep shadows is well beyond the brightness range of any film – the only type of metering that can deal with this is spot metering of highlights, with a camera exposure setting of 2 stops more than this. This will give good highlight detail and no shadow detail at all (which is probably what you want).</p>

<p> </p>

I would go further by saying that "correct exposure" is a misnomer. So is "correct focus, "correct DOF", etc. "Desired exposure", "desired focus" and "desired DOF" may be better terms.

 

Consider this image in your viewfinder. Left half is a bright yellow barn door under the mid day sun. A white goose lays in front of it. Right half is the inside of the barn in deep shadows. A black cat lays in the shadow. The black cat is 15 feet behind the white goose.

 

The meter reading tells a photographer that his camera is incapable of capturing both the highlight and shadow details in one shot. He will have to make a decision.

 

- Expose and focus to capture the details of the goose, and lose the shadow details (and sharpness) of the cat.

 

- Expose and focus to capture the details of the cat, and blow out the highlight (and lose the sharpness) of the goose and door.

 

- Take two shots and merge them in Photoshop.

 

- Etc.

 

- Come back in the evening or on a cloudy day for a softer light.

 

For each of the above shots, the exposure and focus setting would be different. One is not more "correct" than the other. Each works to produce a "desired" image, in exposure, focus and DOF. To achieve this, a photographer needs to first envision what kind of image he wants to capture. Then he has to know how to control the camera for that objective, and not let the camera "think" for him.

 

By picking an image similar as described above and practicing to shoot different "desired" shots, a beginner should be able to learn how to control a camera.

The case I keep thinking of is dusk or near-dark.

 

If you let the camera auto-meter such a scene normally, it's going to give you results dismayingly over-exposed, comparing to your perception of the scene.

 

In such a case if you're auto-exposing you'll need to aggressively set under-exposure compensation. Or expose similarly in Manual.

 

And the inverse applies for something like sunlit snowy scenes, with lots of whites.

Ultimately the correct exposure is the one in your minds eye! If the image you take reflects what you wanted then exposure is correct. About 85% of the time you can estimate that the cameras automatics will give you what you want. For the exceptions you either widely bracket, ie shift the exposure 1/3rd to 2 stops over and under what the camera meter says, and/or use your experience. After taking photographs for over 50 years I can tell you that I don't yet have enough experience!

 

Keep trying, Oh and on a good day a wild guess can often be the best answer.............................

The "correct" exposure is one that shows the scene as it would be outdoors on a sunny day - At least that is what the gray card and light meters try to do.<P>

 

Outdoors on a sunny day with 100 ISO, a gray card reading (or meter reading) would be:<BR>

1/125 sec@f/16<P>

 

Slightly overcast would be:<BR>

1/60 sec@f/16<P>

 

Heavy overcast would be:<BR>

1/30 sec @f/16<P>

 

With those setting everything would be properly lit as if it were a sunny day. The photos would be almost identical except for shadows that would be in a sunny photo but not in an overcast photo. The skin values of people in the photos would all fall on Zone 6.

You may think that a heavy overcast scene doesn't look two stops or four times darker than a sunny scene. To you it doesn't but to the camera it does. The dynamic range of the human eye/brain is far greater than that of film/digital sensor. If you were to record that heavy overcast scene at 1/125 sec@f/16, the resultant photo would look far darker than you remember. Skin values would fall to Zone 4 even though you saw them as zone 6. <P>

 

Here is a person standing at my front door under the porch ceiling:<P>

 

<center><img src="http://jdainis.com/shadow.jpg"></center><P>

 

I never had to carry a flashlight in the daytime to see who was at my front door. My eye/brain has the dynamic range to see the person clearly. The camera exposure was correct for the background but the camera dynamic range could not also put the person, who was four stops darker, into the right exposure.<P>

 

Taking a gray card reading at night would set the exposure to have the gray card appear as it would in the daylight (and everything else around it). Take the gray card reading at night and then close down two stops to place the gray card from Zone 5 to Zone 3 would be a good starting point.

It depends to some degree on whether post processing (or darkroom) work will be applied or whether the images will be

used right out of the camera. In the latter case, you'll need to expose to get the colors and tones that your want,

especially on or near the subject,

 

With post processing, you might be able to expose differently in order to maximize dynamic range and contrast. But this

approach requires skill and experience.

<p>Thanks for your responses, especially to James Dainis - your replies made it very clear!<br /> Now I understand that the thing with auto exposure is that it compensates for actual light conditions, it tries to make a picture which has always more or less the same brightness irregardless of the actual scene brightness.<br /> Which I find strange, I think it would have been possible to program auto exposure such that it somehow tries to approximate the scene brightness in the image, which I think would make much more sense... anyway what I am experimenting with now is deliberately under-expose for scenes with low scene brightness in the hope of getting pictures that are dark as well and look more like the scene actually was. But I feel it is difficult to hit the right spot.</p>

"...I think it would have been possible to program auto exposure such that it somehow tries to approximate the scene brightness in the image..."

 

 

That would cause many photos to simply look underexposed. If you were on my front lawn taking photos of person you would realize that there were hard shadows on the person's face due to the direct sunlight. You would move to the shade of the porch to eliminate those shadows. The camera meter would see that the scene is now two or three stops darker and would indicate an exposure setting to open up for the darker scene, make it look like open daylight. Your eye/brain would not see any darker scene, the person's flesh values still seem to fall on Zone 6 to you. You would want a person's face to fall on Zone 6 for a good portrait. If the camera exposed to have the person's face fall on Zone 5 (one stop darker) to reflect the fact that they are in a darker area than the bright sunlight, then the photo would just look underexposed. You might get away with telling people that the photo was taken at night so that is why it looks so dark. Or, you were looking for a moodier feel. In the meantime, get off my lawn!

<p>I think the term "correct exposure" is subjective. As noted by some of the posters, it really depends on your vision of what the final image will look like. In my shooting style, I don't use auto modes, but do use aperture and shutter priority as well as totally manual. I'd guess that if I let the camera do an auto exposure, that would be technically correct, but likely won't be what I want.<br>

Probably 95% of the shots I take would be deemed less than correct exposure, but they are what I want. </p>

<p>Except for slide film which you intend to project as is and not scan nor make print from it it makes more sense to set the exposure so that you capture the most amount of details. With both color negative film and digital you can render the final image the way you saw it.</p>