Civil War, The Digital Sensor Vs. Film

This month we're beginning a new department here in "Shutterbug" we call Point of View. Photography is a wide-ranging field that engenders passion in its practitioners, and like all great forms of expression creates opinions formed through experience and reflection. In its early days one of the great debates was: Is Photography Art? This was the subject of many essays and heated discussions among players and spectators. Today, issues such as film vs. digital, format choices, the validity of computer generated images, photography as exploitation or revealer, and even the merits of ink jet vs. silver prints cause similar debate. We are opening this department up to readers, manufacturers, and retailers--in short, everyone who lives and breathes photography and who has an opinion about anything affecting imaging today.

Here's how to get involved: write us an e-mail at or send us a letter with a proposed topic and a synopsis of your idea. Once approved, we'll ask you to send us about 500-1000 words on the subject chosen. The idea here is not to push any product or wave any flag, but to create discussion about photo and imaging topics of the day. We reserve the right to edit whatever you send in, although we will never edit intention or opinion but only for length and, hopefully, for clarity. We reserve the right to publish your work on our web site as well, so you can join the archives and be a resource for opinion for years to come.

We'll kick-start the department with the following discussion and hope that it gets your juices flowing. So, get thinking and writing and share your Point of View.

Long brewing, the debate between advocates of digital and film has grown livelier, and often testier of late, with advocates of the two mediums going head to head in forums, studios, and camera clubs around the country. What once was thought to be peaceful coexistence might be turning into an either/or situation, with votes being cast in the dollars spent. Although digital's rise has been meteoric, there's no question that film cameras and film practitioners still prevail. But with the recent wave of high-megapixel digital SLRs, and the coming affordable products in this class, more photographers will be undergoing their own internal struggles as to which way to go.

With that in mind I thought it might be a good time to set down some of my own thoughts about the two forms of imaging. Whether you believe that it's an either/or situation of course depends on how you approach photography and what you want to accomplish in your work. But what began as a local skirmish fought in the fields of commercial photography has now spread to just about every aspect of our craft and trade. The opinions expressed are anecdotal and experiential, rather than the result of any sort of scientific test. They are intended to further debate rather than come to some final conclusion. We'll begin with some background on the different ways in which these systems capture and hold light.

Capturing Light
Both film and CCD or CMOS sensors are receptors that aim to replicate the blink of a human eye. They gather light's energy and coalesce the stimulus into a coherent image, one that can be read again and again as a moment frozen in time. They are aided in this task by the lens, which focuses the light at or near a point on the plane of the receptor, and the shutter and lens diaphragm, which control the volume of light that is the exposure. When light strikes film it causes a change in the metallic silver salts that reside (in color film) within the various color-recording layers. This change is in direct proportion to the amount of light that strikes the film. The latent image is then developed in chemical baths that convert the change to various densities of silver deposit, which (again, in color film) is replaced with dyes, again in proportion to the amount of light that initially struck the film. The color of the recorded light is trapped within these various recording layers, a sandwich that when combined forms a color image that represents in some fashion the color and brightness of the original scene.

While grossly oversimplified, this scenario is typical of how film records an image, although each type of film will differ in how it responds to light depending on its dyes (the various color saturation and contrast personality of each film) and its speed, or overall sensitivity to light. One of the chief advances in film photography over the last decade has been the lessening of the tradeoff between light sensitivity and characteristics usually associated with a good quality image, such as grain, sharpness, and fidelity of color rendition. In creating various film "personalities," filmmakers have also refined the art to be able to produce emulsions that respond to color as a recorder of neutral, vivid, or even very high color saturation. In short, we are in the Golden Age of film technology.

The digital sensor is a complex microchip that is composed of photo sites, electrical receptors that when exposed to light create a charge. After exposure this charge is transferred to a microprocessor that integrates the resultant signal and converts it to binary codes. The microprocessor may be on the chip itself or a separate device within the workings of the system. Each photo site becomes the source of an address made up of these codes. This address identifies color, brightness, and tone and, with integration, creates a relationship between the various sites in terms of contrast, color balance, and tonal spread to create an image.

Grain And Pixels: Does The Analogy Work?
While it might seem that the pattern of silver-halide grains spread through a film emulsion are analogous to these photo sites, the grains actually behave in quite different ways. Although work has been done to make the grains in silver halide uniform, they still have a fairly random shape and tend to spread among each other during and after development. Chips are purely industrial creatures, thus more uniform in their pattern. The pixels--as these photo sites are known--are not random in shape, but are square, hexagonal, or other proprietary shape, but are nonetheless uniform throughout. Color is created by two principle methods. The most prevalent is the overlaying of a checkerboard of red, green, and blue filters that serve as light passes and traps to give the resultant signal a certain color character, all of which is later integrated in processing. The other, the Foveon chip in particular, is a layering of red, green, and blue filters that are more akin to the architecture of film, although, again, an integration process forms color. This difference in the image forming process, and the random vs. near architectural structure, in part defines and differentiates the look of a silver (or dye) formed image and a digital image.

Advantage: Tie. If we accept that every medium has its own "look" we just accept that look and move on. If we expect digital to emulate the photographic look and use that as criteria then photo quality ink jet printers make the argument that indeed it can. While film is at its highest quality ever, digital images can be startling in their sharpness, smoothness (continuous tone), and color rendition. Indeed, some say that digital can and will rival the resolution of film.

Resistance To Deterioration: Expiration And Fog
Film, like milk, has an expiration date after which images made on it will not be as good as when the film is "fresh," or more properly, "mature." Undeveloped film, even when fresh, is subject to change by exposure to heat, radiation, and other forces of nature before exposure and can fade, crack, wrinkle, or become scratched after exposure and development. Film will also become damaged when subject to intense airport screening or even when allowed to sit in high-altitude locations. In short, it's fairly fragile.

While the digital chip is not subject to "fogging" or other detrimental factors there is some problem with what is euphemistically called "pixel death." In essence, the pixels can "go dark" and will not record any image information. Generally this can be counteracted by something called "pixel mapping," which in essence uses algorithms to make assumptions about what "should" reside there in a given subject or scene. Pixels may go dark on their own, due to manufacturing defects, or more likely go "blind" by repeated exposure to intense amounts of light. And sensors, being charged electrical devices, can attract dust and like anything else will not suffer the insult of injury without being damaged. But they are not subject to fogging radiation like film and the digital image resides in code, not in an emulsion spread on a sometimes too-fragile base.

Advantage: Digital. This may be said out of ignorance, as we are not sure (or we are not told) about a CCD or CMOS sensor's eventual rate of deterioration. But having had too many film rolls go out of date, a few damaged by airport scanners, and more than a few shot years back just lose color and density, the fragility of film makes this a point in the digital column. A digital file can be copied over and over to save it if necessary without any information loss; we simply can't say the same about film.

Exposure Tolerance: Overexposure
When subject to gross overexposure the metallic grains upon development will exhibit a growth beyond their intended borders, yielding a halo effect around intense highlights and a "blocking up" that can cause image information to go awry. While films do have an anti-halation backing to prevent light scattering, this does not prevent the intense rendition of highlight areas due to gross overexposure. In addition, if subject to too much development the silver salts in the latent image will go amok and cause blocking as well, resulting in gross densities that mask the image or highlight areas of the image in a deep shroud, making printing and even viewing difficult. In color film this overexposure or overdevelopment may also cause color crossovers, a condition that even the most radical filtration during printing might not be able
to solve.

There is no development per se of the digital image, but there is a similar and even greater danger with overexposure. Think of the photo site as a well with a limited depth into which you pour too much water, or light. The well begins to spill over and there is no place for the water (light) to run except onto other sites. Although advanced digital systems have a drain, if you will, that will begin to handle the electrical spillover, many do not or handle it with less sophistication. The result is known as "blooming" and may create a halo around highlights, throw off color (artifacts), and even create comets of light within the image itself. These overexposed areas become "blanks" in the image file. There is even some talk that consistent and gross overexposure of a digital sensor can cause pixel blindness, discussed earlier.

Advantage: Film. Even taking into account that slide films are equally intolerant of overexposure, the majority of film shot today is negative film, thus photographers are spared the indignity of not being able to handle highlights. Though too much density is certainly not good, all but the most blown out highlights can be printed through to yield detail and texture.

Exposure Tolerance: Underexposure And Low Light
When a film is underexposed the shadow areas lose separation and darker shades blend to black. If underexposure is gross colors begin to shift and there is an overall loss of image fidelity.

Digital sensors will also suffer in their image quality output when underexposed. Interestingly, an image that might seem underexposed on preview in the camera's monitor may have enough information to be useful when later "processed" in an image-editing program. And digital sensors seem to outdo film in the amount of information they can gather in low light. This has been noticed in many situations and leads to the conclusion that digital sensors are much more light efficient in low light and/or handle underexposure in a better fashion than film.

Advantage: Digital. Add to this the fact that you can boost contrast on every frame, rather than have to push an entire roll, and digital is just more versatile in situations where there are a variety of lighting levels in a shooting session.

Processing Variations
Film exhibits major problems if not properly developed. This is the most consistent reason why film can fail to fulfill its potential, even if the initial exposure was correct. While most labs do a good job, some have bad days or are consistently poor in what they deliver. This can be discouraging to photographers, as they often blame themselves for what is really a third-party screw-up. Digital images are processed inside the camera or after with image-editing programs. If the camera delivers poor processing it's a bad camera or if the image-editing program is poorly used it will yield bad images. But at least it's in the hands of the photographer.

Advantage: Digital. It's the photographer's responsibility, and at least they can take the credit or the blame.

Exposure Latitude
Exposure latitude is defined as the degree of over- and underexposure in which a useable image can be recorded. This does not mean that images of equal quality will be delivered throughout this range; it indicates the limits of tolerance of the media.

Negative films are generally thought to be able to handle a range of five stops of light, with the distribution being three stops overexposed and two stops of underexposure. This is one reason why single-use cameras work. Direct positive films are much narrower in their tolerances, due to the reversal process to which they are subjected. That tolerance is about one stop overexposure and one to two stops of underexposure. In practice, digital sensors are more akin to slide films in the way they behave. This is due to the "well" analogy described earlier. What occurs is that when grossly overexposed there seems to be a wiping out, or interference of the signal, and once subject to such conditions the pixel seems incapable of recovering and delivering any useable image information. The latitude on the underexposure side is a bit better.

The implication of this is that photographers should expose digital by biasing that exposure toward the highlights, common practice among slide shooters. Those who have worked with negative film and practice the tried-and-true maxim of exposing for the shadows and developing (or printing) for the highlights will find it simply does not work well with digital. Negative films that are overexposed may have too much density in the highlights, but unless this is a gross overexposure these highlights can be "fought through" for details and tone. Slide shooters know that too much exposure on the highlights yields blank film, or at the least no texture in highlight areas.

So, one truism for exposure with digital is to expose or bias exposure for the highlights. Digital does offer an out to this dilemma. If exposing for the highlights results in underexposure of the shadow areas (which will occur in high-contrast situations) another exposure can be made of the same scene (assuming the camera is fixed on a tripod) for the shadows and the images can be combined later. This is like making separation masks in the darkroom with film, a tricky proposition known to veteran darkroom workers but not easily accomplished in the typical home darkroom. But, as mentioned, the tolerance for underexposure in digital is quite wide.

Advantage: Film. The five-stop range is just too great for digital to compete with, although digital techniques can overcome a high-contrast situation better than film, given some digital darkroom expertise.

Lighting Conditions: Flat Light
Flat light is defined as ambient light with little or no contrast differential throughout the scene. With film it can result in rather dull images. The classic response to this is to increase developing time to increase image contrast, known as "pushing." Of course, if roll film is being used the entire roll must be subject to the push, meaning that for effective results photographers would have to either shoot the entire roll under the same overall lighting condition or swap rolls in and out to work under different lighting conditions. The oft-proposed Zone System for rollfilm photographers works with this premise of swapping rolls.

For digital, dull light is no problem, as it is quite simple to alter image contrast for each frame at the time of exposure or after when the image is processed further in the computer. Contrast enhancement is one of the real values of working in digital, as each pixel area can be changed with ease or the entire image can be boosted to higher contrast through the use of in camera or later in computer controls. Anyone who has worked with the Levels controls in Photoshop has a revelation about contrast control.

Advantage: Digital, just because it's so much easier and more convenient than when working with roll film.

All film is subject to an effect known as the failure of reciprocity, which occurs with long exposure times. Reciprocity in exposure means that if you add or subtract a stop of exposure (in effect, doubling or halving the exposure time) you will get a corresponding increase or decrease in density of one stop on the film, given proper development. When you exceed a certain exposure time (which varies according to the film, and can be anywhere from beyond 1 sec to beyond 10 sec) reciprocity fails and you have to add exposure time to compensate to get the desired density. The factor (or addition to the required time) gets greater as exposure further exceeds the limit of reciprocity effect.

There is no equivalent effect with digital sensors, but long exposures do result in a low-energy recording that will yield increased noise in the image. Noise is akin to grain in film, but can be better described as the flecks of visual static that appear like snow in the image. These flecks are known as artifacts, and they can have color tinges around their edges as well. Many cameras have what is known as a noise reduction function that can be turned on or automatically kicks in (if activated) whenever the system detects excess noise in an image, or when a long exposure time is used. This is a microprocessor function that detects and removes noise by comparing the artifact to the surrounding pixels and changes the noise to match those pixels, rather like a smoothing effect. This improves the image but the processing time can be quite long, and as the image is processing no other images can usually be recorded.

Advantage: Tie. Pick your poison--longer exposure times for film or noise reduction functions in digital. At least film doesn't get those pesky flecks, although the color cast can get weird.

Filters And Contrast Control In Black And White
If working with black and white film, altering the values recorded within the scene through the use of color filters can control contrast. These filters block the complementary color light and pass the same color light. The classic use is to mount a yellow filter to deepen the blue in the sky, a green filter to enhance the forms and patterns of foliage, and a red filter to differentiate between red and green forms (or to enhance the red subject in any scene, which might not be differentiated without the filter). In addition, polarizing filters can be used to control contrast created by non-conducting material reflections.

All images recorded by digital sensors are color images. They are composed of three "channels" of Red, Green, and Blue (RGB). If black and white is the desired result, conversions allow the enhancement of one or another of these channels (or a combination of them all) or the outright elimination of one or another channel. The ease of conversion and the ability to manipulate these channels in an image-editing program eliminates the need for the use of filters for contrast control when shooting for monochrome images.

Advantage: Digital. There's little need for contrast control filters because you're always shooting in RGB and monochrome conversions can be customized with ease.

Color Balance
Films are manufactured with a certain color balance, with "daylight" (5500Þ Kelvin) being the norm. Color balance refers to how the film will behave in ambient light and how true color (without overall cast or bias) will record when the film is exposed under various lighting conditions. For example, if a daylight film is exposed inside a room lit by incandescent bulbs the image will record with an amber cast. This occurs because the light source is deficient in blue.

Our eyes do not see this deficiency, as we adapt to the light and order colors accordingly. Film cannot do this as it is made to respond to light with color rendition as if all light is daylight (or "white" light with the full spectrum included). There are classes of films with tungsten balance (either 3400 or 3200ÞK) that can be used with artificial continuous light sources. But these light sources must be of a specific nature for the color to be dead on. The only true method of getting true color under a variety of light sources is to use a color temperature meter that recommends color-correcting filters over the lens. Of course, a certain color bias in a scene can add mood and flavor, and color correcting, for example, the dazzling warmth of the setting sun's affects on subjects is not desirable.

Digital sensors have the ability to correct for any color temperature light source through a processing step known as white balance. The sensor itself does not do this, but it is added during post-exposure processing in the camera. Setting a certain white balance (of which there are usually five or six available) will place the processing system within a range of white balance (thus Kelvin temperature) settings. Custom white balance allows the photographer to be more precise, if desired, and to set the color balance for an exact color balance, acting like a color temperature meter without the need for adding color-correcting filters over the lens. In addition, color moods can be enhanced by using white balance as a color enhancing filter to add a touch of warmth (using "cloudy" white balance) or cool blue (using tungsten white balance) to the image. In essence, digital cameras literally eliminate the need for filters for color correction and enhancement.

Advantage: Digital. Many cameras allow you to work as if you have a color temperature meter on board, with built-in filtration for mood effects to boot.

Framing Rates
Film allows for high framing rates, with some film cameras shooting as fast as 10 fps (frames per second). The highly sophisticated motors in these cameras react to pressure on the shutter release when the camera is set for a continuous shooting mode. This is a great advantage for sports and action photographers. These framing rates are accompanied by extremely fast autofocusing and autoexposure functions and make film cameras totally responsive to a photographer's reflexes and eye. Of course, shooting a roll of film in 3-4 sec means that you better be good at changing rolls or have a few back-up cameras on hand if the action keeps on happening.

Digital cameras can only make an exposure when the sensor has cleared the previous image and after image processing has sent the image to the converter and memory card. The processing time will vary according to the speed of the processor and the size of the image file (resolution). If functions such as noise reduction are used this processing time will increase. Some cameras feature an on-board buffer, where image information is stored before processing, allowing higher framing rates than would otherwise be available. The capacity of the buffer determines how many exposures can be stored (modified by the size of the image) and how fast the image information can be shunted to it. The storage area does not hold images, per se, but image information that must be processed for it to become an actual image file. This means that the images have to be processed at one point, and when the buffer is full it will halt photography until the images have been processed and moved to the memory card. This means that cameras that offer higher frames per second recording do so only in a set period of time. The specs will read, for example, 3 fps for 10 sec, or a limit of 30 frames at a high burst rate. Again, these specs will be modified according to the resolution of the recorded image.

Advantage: Film cameras, but only if you need high burst rates that can be accomplished within 36 frames. But the processing time of buffered images gives film cameras an advantage as well, as most seasoned photographers can change rolls faster than all those images can be processed. True, pro sports photographers have made the switch to digital, but for all the rest of us film cameras still have an edge.

Image Capacity Per Roll Or Card
Film is spooled in a set number of exposures per roll--24 or 36 exposures for 35mm film and perhaps 12 or 15 exposures with 120 film, depending on the frame's aspect ratio in the medium format camera used.

The capacity of the digital "film," memory cards, is dependent on two factors--the resolution and compression of the recorded image and the capacity of the memory card itself. The resolution determines the size of the file. If, for example, an image is made at the full capacity the 3-megapixel chip can deliver the file size will be about 9MB (megabytes). If the image is uncompressed (TIFF mode) and a 32MB card is loaded then three images can be fit on the card. If the resolution is changed, or if the resolution and compression ratio are changed, then the card will hold proportionately more images. The numbers can be confusing, but following the frame countdown on the camera's LED panel will reveal all.

Advantage: Tie. At least with film you know how many images you have left, but if you're shooting low resolution or high compression images with digital and have a fairly high-capacity card you can get hundreds of images per card. On the other hand, if you have a 16MB card with a 5-megapixel camera and want uncompressed images you better go out and get yourself a high-capacity card. Using that 16 is like going out with one sheet of film in a holder with a 4x5 camera. So even though the card is reusable you better have a few as back-ups when you travel.

Economics: Cost Of Media And Processing
This is a bit of a no-brainer. Each time you want to photograph with a film camera you have to buy a roll of film and pay for processing. A digital memory card is reusable, although we have yet to hear just how many cycles of exposure and download it will handle. But with card prices dropping every six months and capacities going up there's no question that economics come down on the side of digital. It's like using throwaway batteries or rechargeables--the numbers just add up.

Advantage: Digital, hands down.

Instant Feedback
To see the results of your work with film requires that the film be developed, an often-hushed interval where more fingers are crossed than more photographers care to admit. Of course, one of the key advantages of digital photography is the immediate feedback it provides on aesthetics (pose, point of view, and composition) and exposure (contrast and tonal spread). The former is a bit overplayed as an advantage, as the LCD monitor is often a poor indicator of focus and even exposure due to its small size and low resolution. In playback mode, however, many digital cameras allow for two methods of being able to make a better judgment call. One is a zoom function that can bring up cropped areas of the image full frame. The other, and perhaps more important function, is the histogram read-out. A histogram is a visual map of tonal spread and is perhaps the best way to judge how well you have exposed the image. Some cameras also have an verexposure warning function, where areas that have received too much exposure will flash with a user-selectable color to indicate an exposure problem. However, any information you get from the LCD is often blocked by its poor readability in bright daylight. Using an accessory hood or viewing the LCD in shade is always best.

Advantage: Digital. Seeing results right away might be what sells more digital cameras than any other factor, given that you're not trying to see what you got in bright light.

Image Quality
You knew we had to get to this one eventually, and here is where the most debate occurs. Many factors influence image quality in both film and digital, including the quality of the lens, the exposure, the steadiness of the camera when exposure is made, the quality of light, and so forth. Perhaps some sanity can be brought to the discussion by bringing in the issue of resolution and compression on digital vs. the fixed resolution, if you will, of film.

In digital, resolution and compression are key elements in obtaining quality images for a particular end use. Notice the emphasis on end use. Unlike film, where a set frame size records the image, thus determines enlargeability within fairly loose bounds of quality, a digital image recording can be altered to, in effect, be applied to very specific end uses. A smaller image file that might fall apart when enlarged to an 8x10 print will be more than sufficient for a smaller print or as an image that will not see paper but be used for monitor viewing only. And a larger image file that will do well in an 8x10 print will be overkill for a monitor image.

By manipulating file size with resolution and compression choices, efficient use of memory card capacity and ultimately hard drive or other storage capacity can be made. If in doubt as to the intended use the image should be photographed at the best potential quality setting, using either low compression JPEG or TIFF or RAW file format. The size can always be made smaller later. But if the intended use is for small prints or e-mail or web images, then a smaller file size can be used for the initial recording.

So, like film of a certain size, the sensor can also only deliver an image of a certain size and quality end use. The megapixel count of the sensor and how you preprogram the settings determine the high end of what the image can be used for. Even 1- and 2-megapixel cameras yield perfectly fine results for the web and e-mailing images.

Advantage: ??
In the end, image quality is based on a very personal feeling about what an image should look like and how you think and feel that tonal values and colors should spread throughout the frame. Given that both types of media are used in optimum fashion and that the prints or output are created by competent craftspeople, there could be a tossup. But in the end it's up to you as to what will stir your visual instincts and passion. Digital has certain distinct advantages in capture and processing potential, but film still holds the imagination of many photographers who are not just being old-fashioned or stubborn about adopting new technology. Digital has added a lot of creative fun and new challenges to the art and craft of photography. But is it time to say that it's ready to replace film? Only you can decide that.