A LOOK AT EXPOSURE

 

 

An Example

Here is an example that shows how a camera maps its tonal scale, given in half f-stops around 18% gray, to its histogram.  In addition, the grayscale version of the image below is shown with selective tones mapped to that cameras 1/2 f-stop step-scale.

 

 

 

 

 

 

 

 

 

 

 In the center of the graphic below is a 13 bar step chart that was created by exposing an 18% gray card in 1/2 f-stop intervals.  This step chart represents the camera's response to exposure variation over -3 to +3 f-stops around the central value, i.e. correct exposure of the 18% tone.  These 13 tones are, in turn, mapped to their associated histogram slots on the standard 0-255 scale (see blue arrows).  The 13 exposures were not taken in an experimentally perfect way, and at least two of the data points appear to be a bit off the mark.  Nevertheless, the data still shows the non-linear relationship between exposure and histogram position.  As set, this camera produces a file with noticeable highlight separation, with no apparent roll-off as pure white is approached.  As a result, overexposure latitude is limited.  On the dark end of the spectrum, there is noticeable compression of the tones and apparently considerable room for exposure error.  This underexposure latitude is, no doubt, tempered by increasing image noise in the darkest tones.

The tones of the grayscale version of the sample image are shown mapped to the camera's step scale and the histogram for that sample image is shown at the top of the graphic.  This sample image spans most of the usable tonal range of the camera and fits neatly into the 6 + f-stop dynamic range.  We have tones from textured black (-3 f-stops) to a pure white specular highlight on the earring (see orange arrows).  All other tones fall pleasantly in between. 

 

 

Exposing for RAW and  JPEG Capture

How you set your exposure should depend on how you are capturing.  In general, you want to expose for jpegs such that the rendered file closely matches the scene, just as film shooters do when using transparency (slide, reversal) film.  The goal with jpeg capture is to produce the finished product in camera.  When capturing in raw, most experts suggest using the greatest possible exposure, such that the brightest printable tones are moved as close as possible to the right side of the histogram.  Specular highlights, such as those from jewelry, may or may not fall within the printable range.  By doing this, you retain the greatest amount of tonal information in your raw file.  Photoshop and digital-capture expert Bruce Fraser has written an excellent white paper explaining the underlying reason for this approach.  Click the highlighted link to the right to access it: Bruce's article at Adobe.com

Dynamic Range and Exposure

Dynamic range is a measure of the difference between the faintest luminance and the brightest luminance recordable by a  medium.  In digital cameras, this would range from the faintest tones recordable above the inherent electronic signal noise, to the point where the pixels reach saturation and blooming.  For film, it would be the minimum and maximum useable transmission densities, and for printing paper, the minimum and maximum reflectance   The number of the f-stops of light that can be packed between these extremes is the dynamic range. 

Dynamic range must be assessed at several levels, from the luminance range of the original scene, to the maximum recordable range of capture, to the maximum displayable range of the output medium.  Usually, it is the output stage that limits the total dynamic range that can be utilized.  When the luminance range of the scene matches the dynamic range of capture, and that in turn matches the dynamic range of the output medium, life is good.  To get there, is not so simple.

Scene Luminance Range

If you are working in a studio, you can usually adjust your lighting to keep the total dynamic range of the scene to within the dynamic range of both the capture and output media.  This is usually accomplished through an adjustment of the fill lighting.  If you are photographing under available light, you may not be so lucky.  Outdoor scenes can easily exceed the luminance recording range of most capture methods and photographers must decide which parts of the scene must fall outside of the recordable zone, i.e. which areas will be relegated to pure white and black.  And, even if you can capture those tones, you may not be able to print all of them.   Scene luminance can range from a few f-stops to more than 15. 

Capture Dynamic Range

The luminance range you'll be able to capture with your camera will depend on the specific medium that you've chosen.  In general, black and white negative film is the king of dynamic range.  With appropriate development and exposure, up 14 f-stops can be recorded.  This is followed by color negative film, which can record approximately 10 f-stops, though  one can expect some color distortions and tonal compression at the extremes.  Digital capture comes next, with premium SLR cameras capable of 6-7 f-stops in jpeg mode, and 8 or so f-stops in raw.  Fuji's S3 and S5 cameras can capture about another 1.5 f-stops in raw, and medium format digital backs somewhat more.  At the bottom of the capture heap is color reversal film.  You can expect up to 6 f-stops of range with color reversal film.  Even if a medium is capable of capturing a very wide range of tones, there may not be an output medium that can display all of them. 

Output Dynamic Range

The range of viewable tones or output dynamic range is also strongly dependent on the chosen  medium.  Newspaper images have the lowest viewable dynamic range, topping out at about 4 f-stops.  Fancy art magazines approach photographic print paper in total range, nearing 6 f-stops.  Photographic print paper can reproduce, at most, between a 6 and 7 f-stop range.  Computer displays and, optical or digital projection reproduce a noticeably larger range. 

Putting It All Together

The total viewable dynamic range is a function of the capture dynamic range, the output dynamic range, and any adjustments that may be applied to either via chemical, optical, or digital means.  In general, the stage in the the process that has the least dynamic range, dictates the total viewable range.  When adjusting lighting, this stage will dictate the maximum scene luminance appropriate for the task.  For instance, if your image is destined for a magazine page which can only hold a dynamic range of 5 f-stops, your lighting will have to be somewhat flat for best results, and the total scene luminance should fall within 5 f-stops .   Of course this is a bit oversimplified, and through digital manipulation, for instance, one can expand or contract tonal range significantly.

Certain media have a relatively fixed dynamic range and characteristic gamma.  Color negative and positive film fall into this category.  Likewise, optical printing to color paper is generally limited to a known dynamic range and characteristic gamma.  As a result, portrait photographers who use color negative film and RA-4 c-prints work with known quantities and can plan accordingly.  They know that their printing paper limits their total dynamic range to 1:100 or somewhat over 6 f-stops.  As all their media are adjusted to work together and deliver a rather average total gamma, for best results, the dynamic range of the scene should come close to that of the printing paper.  If the scene dynamic range is noticeably less, the prints will look flat.  If it is greater, prints will be snappy, but deep shadows and bright highlights may be rendered as  black and white respectively. 

Other media offer the photographer significant flexibility in the adjustment of gamma, and to a lesser extent dynamic range.  Photographers who use black and white film can make significant adjustments via adjustments to development and exposure.  Similarly, digital photographers, especially those who capture in raw formats, can do so through image-editing software.  For these media, the link between the scene luminance range and the dynamic range of the range-limiting medium, is not so clear. 

Underexposure is demonstrated in much the same way.  The same target was photographed, but  at one f-stop less exposure than normal.  Notice how the histogram tones have been pushed down and compressed toward the black end of the histogram.  Clearly, as configured for these captures, this camera delivers  more latitude at the dark end of the spectrum, but with  considerable compression of tones.  

Target Underexposed One F-stop

 

Underexposed Histogram

 

 

 

 

 

 

 

One lesson learned from this experiment is that the tonal steps of a  histogram are not necessarily linear with respect to exposure.

To demonstrate how overexposure is represented on a histogram, the same target used above was re-photographed, but with the lens aperture opened  by one f-stop.  Notice how black has been pushed up to dark gray and most of the tones associated with the white portion of the target  pushed right off the end of the histogram.  Tonal separation in the lighter tones appears steep and, clearly, there is not a lot of overexposure latitude.

Overexposed Histogram

Target Overexposed One F-stop

 

Using a Camera Histogram to Establish Correct Exposure

A camera histogram is a very useful tool, especially for photographers who shoot in RAW mode.  It can be used as a rough check of exposure accuracy and to ensure that important tones remain within the range of the sensor.

What is a Camera Histogram

The histogram is a representation of tonal frequency against tonal position.  The vertical axis of the histogram represents the number of times a tone occurs in the image (frequency) and the horizontal axis represents the tonal value, from pure black at the left to pure white at the right.  Histograms can generally display a composite of all color channels or a breakout into the red, green, and blue channels.  The camera histogram is usually based on the tonal rendering for a jpeg file, with contrast and white and black points already adjusted.  As such, it doesn't really show the entire range of capture, but rather a typical rendition of it.  If you are shooting RAW, you may have a bit more exposure latitude than appears in your histogram.

Setting Exposure with a Histogram

Several manufacturers now make targets consisting of a white, a gray, and a black area that can be used to set both exposure and white balance.  For the examples below, a homemade version was assembled and photographed in soft, but unfortunately uneven  light.  A digital capture of the target is shown below along with its histogram taken from Adobe Photoshop.  Notice how the white, gray and black areas can be identified as three peaks and are indicated by the letters W, G, & B.  Unfortunately, those peaks are a bit broad and jagged due to the uneven surface and illumination.  For the tones selected for this target, a centering of the peaks within the histogram generally results in near-perfect exposure.  By the way, the gray tone on this target is slightly lighter than an 18% gray.  Some suggest placing 18% gray as middle gray on the histogram.  When shooting in jpeg mode, I have found my best exposure occurs with 18% gray a little to the left of center on the histogram.

Target at Normal Exposure

Normal Exposure Histogram

 

The difference between an exposure for a continuous source and a flash source is shown in the graph below.  The continuous source is graphed as a red line and the strobe as black.  The areas under these curves represent the total energy of the exposure or effective exposure.  Curves that enclose equal areas provide equal effective exposure, whether they result from a flash pulse or a continuous source whose duration is constrained by a camera shutter.   If both of these exposures were applied to the same capture, as might happen when adding flash to an ambient light exposure (e.g. strong fill flash), the total exposure would be the sum of the areas or about twice the exposure (~ +1 f-stop).  

Flash and Continuous Exposure Graph

Metering Strategies for a Reflective Meter

Now that it is pretty clear that a reflected-light meter will calculate shutter and aperture settings that will render any metered object as an 18% percent midtone, what do you do if there aren't any 18% midtones in the scene, or at least any you can get to with your meter.  One approach is to take a reading or several readings of an area that contains a balanced mix of dark and light areas.  Chances are the average reflectance is not that far from 18%.  Another approach is to meter a very dark area and a very light area and split the difference.  If you have a neutral gray 18% card (Kodak, Gretag Macbeth, etc.), place it in the scene and meter off of it when practical. 

The more you work with a reflectance meter under a variety of lighting conditions, the better will be your sense of where  the various tones fall on the exposure scale. Under normal contrast lighting for instance,  you may find that tones in your captured image render as pure white when they are 3 f-stops brighter than something that reflects at 18%.  Similarly, tones 3 f-stops darker than 18% may yield a deep black with only the faintest detail.   Using this knowledge, you might meter off of object that should be rendered as a textured white and open up the lens aperture 2.5 f-stops more than recommended by the meter.  To the same end, you could meter off of a deep, textured shadow and close down the lens aperture approximately  2.5 f-stops. 

Flash and Continuous Light Meters

We've covered the difference between incident and reflected meters.  Meters can also be differentiated by the type of illumination they measure.   Some measure only continuous lighting, others only flash, and quite a few measure both.  The best of the combined flash/continuous meters can discriminate between the flash and continuous portions and accurately indicate the strength of both.  The cheaper combined meters are less reliable in this regard and are best used for one illumination type or the other.

The circuitry needed to measure flash illumination is significantly more complex than that used for continuous illumination. A simple meter to measure continuous light can be cobbled together from a selenium photovoltaic cell, a few resistors, and a meter movement--that's all.  However, flash meters must have circuitry that both detects the sudden rise of a flash pulse and then calculates the total energy of that pulse.

Reflective reading on center strip

Reflective reading on left strip

Most reflective meters are calibrated to return perfect exposure when the object or scene reflects, on average, 18 percent of the light that strikes it.  Some manufacturers calibrate their meters for a slightly different reflectance.  As an example, if we have an evenly lighted scene and we take a reflective reading of an object with an 18 percent reflectance and set our camera accordingly, all of the objects in the scene should fall into their proper place and good exposure should result.  On the other hand, if we point our meter at a white object in the scene and use the resulting exposure recommendations to set our camera, we will find the result too dark.  In fact, the white object should look gray, 18 percent gray.  Similarly, if we point our meter at a black object in the scene and use that reading, the result will be too light, and the black object will appear gray, 18 percent gray.  This is shown in the three photographs below.  A board consisting of a white, a gray, and a black strip was photographed under flat lighting.  In the first image (left), a reflective reading was taken of the center (gray) strip and camera set accordingly.  In the second image (right), a reflective reading was taken of the white strip and the camera set accordingly.  In the third image (bottom), a reflective reading was taken of the black strip and, again, the camera set accordingly.  As expected, things fell pretty much into place in the first image, but in the second image, everything was greatly underexposed and the white strip is gray.  In the third image everything is greatly overexposed and the black strip is now gray. 

Reflective reading on right strip

 

In this section we'll tackle exposure basics and  take a look at how exposure,  the camera histogram, and the dynamic range of the scene are related.  

Camera Exposure Controls and Studio Flash

 When using studio strobe, there is generally no means of interfacing with a camera's through-the-lens exposure and flash controls.  You will have to determine and set exposure manually.   

Manual Exposure Control and Measurement

 

Exposure Control with Continuous Lighting Sources

Whether you are shooting film or capturing digitally, the basic principles of exposure control are the same.  You have to get the right amount of light energy to the capture medium.  With continuous sources of light (daylight, light bulbs, etc.), the photographer has basically two means for doing this: adjusting the intensity of the exposure (lens aperture), and adjusting the duration of the exposure (shutter speed).  With studio strobes, the duration of the exposure is set by the strobe's pulse width, limiting adjustment to intensity only.

Adjusting  Exposure Intensity via Lens Aperture

Light intensity is adjusted via the camera lens aperture.  Lenses are produced with standard aperture settings known as f-stops.  Typical whole f-stops are:  f1.4, f2.0, f2.8, f4, f5.6, f8, f11, f16, f22, f32.  Each succeeding f-stop lets in half of the light of the preceding f-stop and halves the effective exposure.  For instance, the light intensity at f4 is half that at f2.8.  Of course, you could look at it the other way round and say that the intensity at f2.8 is twice that at f4.  Applying the successive halving of intensity, you'll find the intensity at f8 to be 1/8 that at f2.8. You may have noticed that this strange sequence of f-numbers is actually the powers of the square root of 2, albeit rounded at the odd powers. 

Adjusting Exposure Duration via Shutter Speed

When working with continuous lighting, the time that the capture medium is exposed to light is adjusted via the camera's shutter speed.  The camera's shutter system is calibrated for standardized durations.  Typical whole shutter durations are: 1 sec., 1/2 sec., 1/4 sec., 1/8 sec., 1/15 sec, 1/30 sec., 1/60 sec., 1/125 sec., 1/250 sec., 1/500 sec., 1/1000 sec.  From this point forward we'll refer to the duration of the exposure  as the shutter speed.  The faster the shutter speed, the shorter is the duration.  Similar to the intensity halving of f-stops, each succeeding shutter speed is 1/2 the duration of it's predecessor, and cuts the effective exposure value in half. 

How Shutter Speed and Lens Aperture Interact

The intensity and duration of the exposure act in equal measure.  Halving either, halves the total exposure.  When doubling one and halving the other, exposure remains the same.  Except at the exposure extremes, several combinations of duration and intensity can be chosen to yield identical effective exposure.  As an example, the following combinations yield identical exposure: f5.6 & 1/125 sec., f8 & 1/60 sec., f2.8 & 1/500 sec., and f11 & 1/30 sec. 

The ability to trade aperture for shutter speed and visa versa is great for the savvy photographer.  For instance, to minimize motion blur, a sports photographer might choose the widest lens aperture paired with the highest shutter speed.   At the other end of the spectrum, a landscape photographer might pair a slower shutter speed with the small aperture necessary to keep the image foreground and background in apparent focus.

Exposure Control with Studio Flash Systems

Adjusting Exposure via Lens Aperture

The lens aperture is used to change the intensity of illumination for strobe sources just as it is for continuous light. See the explanation above for continuous sources.

Shutter Speed and Strobe Illumination

For strobe-only illumination, the exposure duration is almost always set by the strobe's pulse width.  Strobe pulse durations are typically less than 1/500 sec. and most shutters are either unable to synchronize with flash above this speed or, in the case of leaf shutters, are unable to exceed this speed.  As a result, the strobe pulse sets the duration by default.

Synchronization

There are two types of shutters found in most high-quality cameras today: leaf shutters, and focal-plane shutters. 

Leaf shutters, which are found primarily in compact,  medium-format, and large-format cameras, synchronize with studio strobe systems at any shutter speed.  Leaf shutters operate in the same manner at all speed settings, triggering the flash just as all the blades of the shutter swing fully open. With an exception or two, leaf shutters attain a maximum shutter speed of 1/500 sec.  As leaf shutters cannot open and close faster than the duration of a flash pulse, they cannot reduce the duration of illumination. 

Focal-plane shutters present another issue; most can be set to durations shorter than strobe pulses, but not while synchronizing properly with the flash.  Most SLR digital and 35 mm film cameras are equipped with focal-plane shutters.  Unlike leaf shutters which operate in the same manner at all speeds, focal-plane shutters operate in one manner up to and including the synchronization speed, and somewhat differently for higher speeds.  Focal plane shutters consist of two curtains, so called as earlier versions consisted of two cloth curtains, one which opened to start the exposure and other which closed to complete it.  Today's shutters are more likely to be made  from aluminum alloy or carbon composite.  Nevertheless, they operate in much the same way.  For speeds through the synchronization speed, the front curtain opens fully to expose the film or sensor.  If the camera is operating in front-curtain synchronization, the flash is triggered just as the front curtain completely clears the frame.  The rear curtain then closes at the end of the exposure duration.  For rear-curtain synchronization, the front curtain opens and the flash triggering is delayed until just before the close of the rear curtain.  In either case, the image frame is fully cleared when the flash is triggered.   Above the synchronization speed, there is no time during which the whole image frame is clear of the shutter curtains.  The front and rear shutter curtains move together for most of the exposure, traveling as a slit that traverses the image frame.  If a shutter speed above the maximum synchronization speed is used, only a fraction of the full frame will be exposed to the flash. 

 

Measuring Light

Most photographers who use a variety of strobe lighting in their work rely on a handheld flash meter.  For film users, a good flash meter is nearly a necessity.  Digital users can get by using the histogram of their camera, but a meter is far more convenient.

Exposure Meters

Exposure meters can be classified by how they measure light (incident, reflective) and what type of light they measure (continuous, flash).  The simplest and least expensive meters measure continuous reflected light.  The most expensive are capable of measuring it all, and then some.  Most meters can display exposure in terms of shutter speed and aperture, holding either the aperture or shutter speed constant (priority metering) while varying the other parameter. 

Reflective Measurements

The illustration to the right shows a reflective meter reading being taken.  Reflective meters measure light reflecting back from an area.  As with a camera lens, reflective meters have an angle of view, often referred to as the angle of acceptance.  Some reflective meters gather light over a broad field and others, such as spot meters, from solid angles down to 1 degree.  If you take reflective readings of  various objects in an evenly lighted scene, the reflective meter will return a variety of different results depending on the reflectance of the individual objects.  An incident meter moved about the scene would return the same exposure result for all positions.  This begs the question; if different objects give different reflective readings, which reading do I choose to get a good exposure?

 

Incident Measurements

Incident Light Measurement

The illustration to the right shows an incident meter reading being taken at the subject position.  Incident meters are usually equipped with a hemispheric diffuser that averages the incident light over an 180 degree solid angle.  Incident meters measure the light falling on an area, not the light reflected from it.  For objects that reflect a large portion of the light incident upon them in a diffuse manner, incident meters normally provide excellent results. Both skin and clothing are predominantly diffuse reflectors, so incident measurements work well for portrait photographers.  Incident meters will not give ideal exposure for objects that either transmit, absorb, or directly reflect much of the light that hits them

 

 

 

 

 

 

Reflected Light Measurement