Ekserb
You really hate me.
- Joined
- Jun 19, 2002
- Posts
- 4,226
NOTICE: As of January 2021, I've closed the hosting account where most of the images on this thread were being shared from. As a result, those links no longer work and the photos won't be viewable. I apologize for the inconvenience.
Got a question about taking or sharing pictures? Ask here and get an answer!
This thread isn't about sharing your work or the work of others, unless doing so illustrates a technique of photography. It's not about critique of individual photos, unless the photographer has requested honest and relevant feedback on how to improve an image.
What this thread is is a place for anyone to learn a little more about photography to improve the quality of the photos they share here or with friends and family.
Camera types:
Mobile phone cameras. Some people don't even think of these when they think of a camera, but the fact is that almost everyone carries a camera with them all the time in the form of a cell phone. While the majority of cheaply-made phones have equally crappy cameras, there are some that rival a "real" camera in quality and features. There are even web sites dedicated to photographs taken with nothing but a cell phone camera. Speaking from experience I can say that my own iPhone 4 consistently amazes me with the quality of its pictures. And the (sometimes freely) available editing apps can transform a photo from snapshot to artwork with a few finger taps.
EDIT: Since writing this, I have upgraded my phone to the iPhone 6s and the camera is amazing. It really is a suitable replacement for a DSLR in a lot of instances. Certainly for photos that will be shared only online, the images this thing produces are nearly indistinguishable from those produced by a DSLR, albeit one with a fixed lens. And modern image editing apps on the phone can recreate just about every feature of a desktop editing application. Only the most astute viewer would be able to tell this came from a phone:
EDIT 2: Have you seen the photos that come from an iPhone 11 Pro? Come on! There is very little reason for anyone but the highest level professional to buy anything but a great phone camera anymore.
Point-and-shoot (P&S) cameras are the least expensive, lightest, smallest, easiest-to-use, and relatively poorest performing types of cameras you can buy. Modern P&S cameras will take very nice pictures with little to no user meddling, but they are usually limited in lens quality and selection, image quality, and advanced manual control.
Compact zooms typically have a better lens than a P&S, at the expense of being larger and heavier. Their lens dictates the larger size of the camera so they won't usually slide easily into a pocket, though a purse can easily accommodate one of these cameras.
Compact interchangeable lens cameras are the newest sensation in digital cameras. The camera bodies are very small, but they can accept interchangeable lenses in a very wide array of focal lengths and light gathering ability. In some ways, these are better than a DSLR (see next paragraph), but they also come with some of the drawbacks of a DSLR-type camera. Having a removable lens allows you to use the very best focal length for a given scene, but the act of replacing the lens can introduce dust into the sensor cavity which may or may not add ugly spots to your final images. While most all cameras now have some kind of dust-removal system built-in to shake off these dust particles, these systems can only do so much and care is always recommended when changing lenses. Also, since no single lens is perfectly appropriate for every scene, you'll need to carry with you several lenses to fit any possible occasion. (The lens you leave at home is the one lens you will really need that day!) Feature sets of these camera are quite good, with some bodies having nearly all the bells and whistles of a full-size DSLR.
Single Lens Reflex (SLR) cameras, their modern digital heirs, DSLRs, and what Sony calls SLTs (Single-Lens Translucent) cameras*, are the gold standard for typical professional field use. Their image quality (when paired with a top-of-the-line lens) are second only to extremely expensive digital scanning-back view cameras and large-format film cameras, and they are also the heaviest and most expensive cameras used by most people. These camera bodies are also sometimes equipped with so-called "full frame" image sensors that are the same physical dimensions and aspect ratio as a 35mm film frame (36mm x 24mm, 3:2 aspect). The cameras are large to accommodate a "mirror box" that reflects incoming light toward a viewfinder (typically an optical viewfinder, but more recently they are becoming electronic) mounted on top of the camera body. Due to their design, these cameras also tend to have the fastest, most precise autofocus. They can cost anywhere from several hundred dollars to several thousand, and the price of a complete system (camera body, lenses, vertical grip, flash, etc, though few enthusiasts would ever say their system was "complete") usually puts them out of reach for an average user, though many people are perfectly happy with a body and single "kit" lens for under six or seven hundred bucks.
*Sony SLT cameras use a fixed mirror that is partially reflective, so most of the light passes through the mirror to the sensor while a smaller percentage is reflected up into what would normally be a pentaprism or focusing screen. The advantage here is the mirror doesn't move when the shutter is activated, so there are fewer moving parts that need to reset after every exposure. This also means the frame rate for bursts can be relatively higher than for a DSLR mechanism and because the mirror is always reflecting an image to the shooter's eye, there is no need for what's known as "mirror blackout" during the exposure; the shooter can see continuously what he/she is imaging as the camera makes one or more exposures. This camera type appears to have been a temporary fix for Sony while they pursued the mirrorless cameras discussed in the next section. As it stands now (2019), Sony makes only one SLT camera body (A99 Mark 2) and no SLRs, so it looks like the Alpha lens mount common to both the SLR and SLT are going the way of the dinosaur.
Mirrorless cameras are the latest trend in high-end digital cameras. In overall design, they look a lot like DSLRs, but they lack the mirror box of those cameras and are smaller and more compact because of it. Native-mount lenses for these cameras are smaller and lighter than their DSLR counterparts, so the entire system is much easier to deal with. The flange distance (measured from the lens mount to the sensor) is also shorter, so these camera have the ability to mount virtually every lens ever made by any manufacturer for which someone has made an adapter. This is a huge win for users who have special lenses from an older camera and don't want to change their entire system to accommodate a new camera body. Sony has been the leader in this area, although Canon and Nikon have recently released their own take on the design. [Updated this info October 2018]
View cameras are similar in appearance to what you might see in an old movie, where the photographer would compose the shot with his head under a black fabric and then insert a sheet of film into the back of the camera before exposing it with a shutter in the lens itself. While these cameras look complex, they are actually very simple in design. A lens board is mounted parallel to the film and focus is adjusted by moving the lens board closer to or further from the film plane (or vice versa). A fabric bellows keeps light from infiltrating the camera and allows the entire unit to be collapsed into a small form for travel or storage. Modern view cameras can be film or digital, with film sheet sizes all the way up to 8" x 10" or larger! Digital scanning backs work much like smaller digital cameras, although the scanning process is relatively slow and can result in strange distortions when shooting moving scenes like water or trees blowing in the wind. Film cameras can be found for a few hundred dollars, but digital scanning backs can cost over $50,000. Ouch!
A note on resolution: Camera manufacturers put a lot of emphasis on the pixel resolution of their cameras' sensors. 10-megapixels (MP), 14-MP, and higher have become the standard for even the lowliest point-and-shoots, with many high-end DSLRs now attaining 24-MP (and recently even 50+MP). All this does is give you some leeway when it comes to cropping the final image. The images will not be significantly sharper or better than photos taken with a 5- or 6-MP camera unless you intend to print them at sizes larger than 12" x 18". The biggest contributors to clean, sharp images are a good lens and careful attention to exposure. I have a print hanging on my wall that was made with a 6-MP Konica Minolta DSLR and it looks fine printed at 20" x 30" (and that was with an old, crappy lens).
So, what kind of camera takes the best pictures? None of them. The best pictures are made by the best photographers, regardless of the equipment used. I've seen images made with a cell phone that rival in emotion and impact those made with a $5000 Canon DSLR. That doesn't mean that a professional can get the same quality from a P&S that he can from a view camera, but the best images can only be wrung from an expensive tool by someone who knows how to tweak every setting to its optimal performance. (By the way, telling a photographer that his photos are beautiful and then commenting that he must have a really nice camera is like assuming Michelangelo used only really expensive brushes—it's an insult.)
Picture-making is all about light control:
Instead of thinking of photography in terms of camera control, you really need to think of it as light control. The camera is just a tool that allows you to control the amount of light that falls on the focal plane. Whether you allow more or less light onto the sensor helps determine the brightness, contrast, and emotion of the final image (composition also determines the impact of the final image—more on that later). While some of these things can be massaged in post-processing software, there are limits to what a computer can do with a too-dark, too-bright, or blurry original image.
Before setting your exposure, take a look at your scene through squinted eyes. Look around for the very brightest part of the scene and pick out highlights and bright areas like reflections, puffy white clouds, sea foam, bright sidewalks, etc.—anything that immediately gets your attention. These bright spots will probably be over-exposed (also referred to as "blown out") in your shot unless you correct for their brightness. Your subject may appear to be correctly exposed, but white clouds will appear as big, pure-white blotches instead of finely detailed wisps. The quickest way to spot an amateur photo is to look for over-exposed clouds.
Once you've found the trouble spots, set your exposure based on the brightness of the brightest thing in the scene. This may be hard to do unless your camera has some way to spot meter a scene, but you can shoot several frames with varying settings until you get one that does not over-expose those areas. How do you know if they're over-exposed? Some of the more advanced cameras have a system that flashes the pixels in the over- and under-exposed areas of the photo on the preview LCD screen after the image is recorded. Some may have a histogram on the preview screen. A histogram is basically a bar graph that shows the number of pixels exposed at every level over the entire image and it's the best way to look for exposure problems after the shot is made. A good histogram will look like a bell curve with tapered ends and a hump in the middle. This shows that most of the image is exposed around the middle of the camera's ability to record light and the highlights and shadows (to the right and left ends of the graph, respectively) are not clipped at their extremes. Once a bright or dark pixel is clipped, no more image information can be stored at the pixel—it is either pure white or pure black, and neither of these is a desired result.
(I might add here that most DSLRs have different means of measuring the exposure coming through the lens. Average meters exactly what it says by taking an average of the entire scene to set the exposure. Center-weighted uses the middle-ish part of the scene. There is also a system called Matrix metering that uses a more intelligent means of figuring out what parts of the scene deserve to be over- or under-exposed and does a pretty darn good job of giving a well-balanced image in varied circumstances. Spot metering uses a very small area at the center of the image while ignoring the areas around that center spot. Most people who really know what they are doing will use the Spot pattern to meter several parts of the scene and then adjust the camera settings based on those different readings. When I'm shooting on a location I will typically meter the brightest thing in the area and adjust the manual exposure based on that reading. After that, I leave the settings alone unless the lighting changes or I move to another location. After a lot of experience I can tell you that 1/1250 second, ƒ5.6, ISO 100 will expose perfectly a surfer at midday without blowing out the breaking waves.)
Preview of recorded image with and without clipping areas (red for over-exposure and blue for under-exposure):
Histogram showing over exposure:
Histogram correctly exposed:
The mid-exposure hump can be smooth or peaked, as long as the peaks don't reach all the way to the ends of the graph. A spike near, but not at either end of the histogram is okay and will actually make for high contrast in the image without loss of detail.
Why not look for the darkest areas of the scene and set the camera for those? People tend to look at shadows and expect to see darkness. You expect to see nothing when you look into a dark cave. When you look at a very bright object, however, you expect to see extreme detail, even if you have to squint your eyes to see clearly, you still expect to see something other than just a big, white blob. Exposing for the highlights allows us to control those bright spots and let the shadows work themselves out. If you have to force a dark exposure to maintain the detail in the highlights, so be it—your audience won't know the difference.
Specular highlights are a different story. These are reflections—usually of the sun or a direct light source—that cannot be subdued with any amount of exposure control (at least, not without sacrificing the rest of the image to extreme under-exposure). If you're shooting in the direction of the sun across a pool of water, for example, there will be glints of light all over the surface of the water. These are going to be washed out in the final image and that's fine. Again, this is something that people expect to see when they look at a very bright source of light.
Over-exposed specular highlights:
The difference between JPEG and RAW. All cameras will capture images in the JPEG file format. This method of storing picture information is known as "lossy" compression because the algorithm used to make the file size as small as possible throws out redundant pieces of information. The problem is in how it decides what is redundant. It samples large chunks of visually similar pixels and stores those chunks as blocks of identical pixels. What may have looked like a gradual gradient, or ramp of colors and shades, becomes in the JPEG file a solid block of one particular color. Where two of these blocks meet, there can be an abrupt change known as a compression artifact.
Most people may not even see these artifacts, but they are there and they degrade the image quality just slightly. Every time the image is edited or re-compressed, these artifacts become slightly easier to see until they are ugly and irreversible.
Another drawback to JPEG's lossy code is that it immediately throws out any useable overhead captured by the camera's sensor. Digital sensors are able to "see" a lot more than the camera records and they don't recognize any white balance settings until after they are processed by the camera's onboard software. Recording in RAW format does exactly what its name implies: The recorded file is exactly what was seen by the sensor at the time of exposure, minus any camera processing. There is no sharpening applied, no white balance adjustment, and no noise reduction (though some cameras do have built-in noise reduction at the sensor-level). The RAW file has all kinds of detail that can be drawn out in post-processing software, but most of it is lost in the JPEG compression.
As an example, a key feature of RAW shooting is the ability to recover details in over-exposed areas of an image. In the last section, I mentioned blown out clouds as a measure of a photographer's skill level. A lot of shots with this defect can be salvaged by shooting in RAW format as the highlight details are still there, just hidden in the file's additional bits. Any software capable of editing RAW formats can extract some (but not necessarily all) of this detail, restoring those clouds to wispy or billowy goodness.
Let's say you're shooting indoors with lights near an open window during the day. The white balance of the scene is tricky and shooting with your camera set to Daylight white balance (WB) will result in the interior of the scene looking very yellow from the tungsten light sources. If you set for Indoor WB, the exterior light through the window will be very blue. JPEG photos will have this WB setting applied and there is only so much you can do to change this setting without loss of details and some strange color shifts. RAW files, on the other hand, are easily adjusted to Daylight or Indoor WB, or anything in between. You can adjust the color temperature to your own tastes until the scene more closely matches what you experienced on the set.
Aperture, or how to control your depth of field:
(The next few paragraphs pertain to all cameras, though the photographer may have limited control over these settings depending on the level of automation built into the camera.)
If you have a lens marked ƒ5.6, that is the widest opening of the aperture (iris) for that lens. (Each full ƒ-stop of a lens doubles the amount of light that passes through to the sensor or film. Unfortunately, they're not numbered logically, so you kind of have to know these things from experience, but ƒ4 is twice as much light as ƒ5.6, ƒ2.8 is twice as much light as ƒ4, and ƒ2 is twice as much light as ƒ2.8, and so on. An ƒ1.4 lens lets in sixteen times as much light as a ƒ5.6 lens at their widest respective settings.) The aperture numbers are actually a ratio of the iris diameter to the focal length of the lens. They represent the amount of light that enters the lens and they are the same for all lenses, no matter the focal length. In other words, ƒ4 on a 17mm wide angle lens lets in the same amount of light as ƒ4 on a 200mm telephoto lens. If lens manufacturers expressed the opening as an absolute diameter, the aperture setting would be different for every focal length and would require complex math to figure the correct exposure. ƒ-numbers aren't the easiest to understand, but they are much better than the alternative.
Most photographers use aperture to control depth of field (DOF), which is the amount of your scene that appears in acceptable focus (from near to far) while using the shutter speed to control the overall exposure. (Obviously, there are times when you need to use shutter speed creatively and aperture and/or sensitivity settings can be used for exposure control.) For wide landscapes, you typically want a very deep focus to get everything from the foreground to the background in sharp focus. An aperture setting of ƒ11 or smaller (down to ƒ32 or ƒ64 on some lenses!) is good for these times. For portraiture, you probably want to use a wider aperture and thus a narrower depth of field to exclude the background by blurring it out. ƒ4 or wider results in only the subject being in focus while anything in front of or behind the subject is heavily blurred. (Apertures wider than ƒ2.8 can yield a focus so narrow that only the tip of the nose is in focus while the eyes are slightly blurry.) The focal length of the lens is another factor that affects depth of field, but most "portrait" lenses are between 75mm and 150mm. Wider angles have wider depth of field and longer lenses have narrower depth of field for any given aperture setting.
Also note that the distance from the subject affects the actual DOF with objects further away having a greater DOF and objects close to the lens having a relatively narrow DOF. In other words, photographing a person standing next to a car results in a wider focus than shooting an insect on a flower.
Wide depth of field (narrow aperture of ƒ14):
Narrow depth of field (wide aperture of ƒ5.6):
Most lenses will perform best around the middle of their aperture range. Even the most expensive lenses will have some slight degradation of quality at either end of the aperture range, but the better the lens quality, the less noticeable this degradation is. The point is, when you use an inexpensive "kit" lens, you might want to shoot at ƒ8 or ƒ11 to get the sharpest possible image at the expense of a wide depth of field, whereas you can shoot a better lens at ƒ4 and still retain sharpness while keeping a relatively shallow depth of field. If you buy a zoom lens with a variable aperture (the lens is marked 28-80mm ƒ4.5-5.6, for example) the lens lets in less light when you zoom to the longer focal length, so the widest opening available at 80mm is actually 5.6 instead of 4.5. Since the lens is probably not at its sharpest at ƒ5.6, you should stop it down to ƒ8 or smaller to get the best sharpness at the full telephoto setting. This may or may not be acceptable for the given scene. A better lens will have a constant aperture throughout its zoom range and a prime lens (fixed focal length) is typically sharper and faster than a zoom lens. (As usual, there are exceptions. Some really nice zooms are sharper than a cheap prime lens.) Shooting with a lens with a maximum aperture of ƒ1.4 gives you the freedom to use a wider aperture and still stay in the middle of the aperture range. ƒ4 on a ƒ1.4 lens is typically sharper than ƒ4 on a ƒ4 lens.
Photographers have a word for the blur effect created by a wide aperture: Bokeh (pronounced BO-kah). The smoother and silkier the blur, the better the bokeh. As it turns out, a perfect lens would produce blurred elements that were evenly out of focus from their center to their edges with a clean edge. While in theory this would seem optimal, in practice this would not look good. So, lens manufacturers go to great lengths to design slightly imperfect lens designs that yield blurring that tapers off to smoothly graduated edges. The best lenses are often prized for their creamy soft bokeh effects. Things like the number and curvature of the aperture blades also affect the blur quality. You can often count the number of blades in a lens by looking at the out-of-focus lights in the background of a night scene—the pinpoints of light will adopt the shape of the aperture (usually five to nine blades). Star-shaped flares will also indicate the number of blades, but the number of star points will be doubled on an odd-numbered aperture (five blades will create a ten-pointed star pattern). This is due to refractions inside the lens and the effect is greater when the aperture is smaller.
I was going to add a few technical paragraphs about acceptable focus and hyperfocal distance, but that stuff is just way too complicated and not really something everyone needs to know. If you must learn it, Google the aforementioned terms and be prepared for a lot of dry reading. And really, if that kind of thing is interesting to you, you're probably not reading this primer anyway.
Got a question about taking or sharing pictures? Ask here and get an answer!
This thread isn't about sharing your work or the work of others, unless doing so illustrates a technique of photography. It's not about critique of individual photos, unless the photographer has requested honest and relevant feedback on how to improve an image.
What this thread is is a place for anyone to learn a little more about photography to improve the quality of the photos they share here or with friends and family.
Camera types:
Mobile phone cameras. Some people don't even think of these when they think of a camera, but the fact is that almost everyone carries a camera with them all the time in the form of a cell phone. While the majority of cheaply-made phones have equally crappy cameras, there are some that rival a "real" camera in quality and features. There are even web sites dedicated to photographs taken with nothing but a cell phone camera. Speaking from experience I can say that my own iPhone 4 consistently amazes me with the quality of its pictures. And the (sometimes freely) available editing apps can transform a photo from snapshot to artwork with a few finger taps.
EDIT: Since writing this, I have upgraded my phone to the iPhone 6s and the camera is amazing. It really is a suitable replacement for a DSLR in a lot of instances. Certainly for photos that will be shared only online, the images this thing produces are nearly indistinguishable from those produced by a DSLR, albeit one with a fixed lens. And modern image editing apps on the phone can recreate just about every feature of a desktop editing application. Only the most astute viewer would be able to tell this came from a phone:
EDIT 2: Have you seen the photos that come from an iPhone 11 Pro? Come on! There is very little reason for anyone but the highest level professional to buy anything but a great phone camera anymore.
Point-and-shoot (P&S) cameras are the least expensive, lightest, smallest, easiest-to-use, and relatively poorest performing types of cameras you can buy. Modern P&S cameras will take very nice pictures with little to no user meddling, but they are usually limited in lens quality and selection, image quality, and advanced manual control.
Compact zooms typically have a better lens than a P&S, at the expense of being larger and heavier. Their lens dictates the larger size of the camera so they won't usually slide easily into a pocket, though a purse can easily accommodate one of these cameras.
Compact interchangeable lens cameras are the newest sensation in digital cameras. The camera bodies are very small, but they can accept interchangeable lenses in a very wide array of focal lengths and light gathering ability. In some ways, these are better than a DSLR (see next paragraph), but they also come with some of the drawbacks of a DSLR-type camera. Having a removable lens allows you to use the very best focal length for a given scene, but the act of replacing the lens can introduce dust into the sensor cavity which may or may not add ugly spots to your final images. While most all cameras now have some kind of dust-removal system built-in to shake off these dust particles, these systems can only do so much and care is always recommended when changing lenses. Also, since no single lens is perfectly appropriate for every scene, you'll need to carry with you several lenses to fit any possible occasion. (The lens you leave at home is the one lens you will really need that day!) Feature sets of these camera are quite good, with some bodies having nearly all the bells and whistles of a full-size DSLR.
Single Lens Reflex (SLR) cameras, their modern digital heirs, DSLRs, and what Sony calls SLTs (Single-Lens Translucent) cameras*, are the gold standard for typical professional field use. Their image quality (when paired with a top-of-the-line lens) are second only to extremely expensive digital scanning-back view cameras and large-format film cameras, and they are also the heaviest and most expensive cameras used by most people. These camera bodies are also sometimes equipped with so-called "full frame" image sensors that are the same physical dimensions and aspect ratio as a 35mm film frame (36mm x 24mm, 3:2 aspect). The cameras are large to accommodate a "mirror box" that reflects incoming light toward a viewfinder (typically an optical viewfinder, but more recently they are becoming electronic) mounted on top of the camera body. Due to their design, these cameras also tend to have the fastest, most precise autofocus. They can cost anywhere from several hundred dollars to several thousand, and the price of a complete system (camera body, lenses, vertical grip, flash, etc, though few enthusiasts would ever say their system was "complete") usually puts them out of reach for an average user, though many people are perfectly happy with a body and single "kit" lens for under six or seven hundred bucks.
*Sony SLT cameras use a fixed mirror that is partially reflective, so most of the light passes through the mirror to the sensor while a smaller percentage is reflected up into what would normally be a pentaprism or focusing screen. The advantage here is the mirror doesn't move when the shutter is activated, so there are fewer moving parts that need to reset after every exposure. This also means the frame rate for bursts can be relatively higher than for a DSLR mechanism and because the mirror is always reflecting an image to the shooter's eye, there is no need for what's known as "mirror blackout" during the exposure; the shooter can see continuously what he/she is imaging as the camera makes one or more exposures. This camera type appears to have been a temporary fix for Sony while they pursued the mirrorless cameras discussed in the next section. As it stands now (2019), Sony makes only one SLT camera body (A99 Mark 2) and no SLRs, so it looks like the Alpha lens mount common to both the SLR and SLT are going the way of the dinosaur.
Mirrorless cameras are the latest trend in high-end digital cameras. In overall design, they look a lot like DSLRs, but they lack the mirror box of those cameras and are smaller and more compact because of it. Native-mount lenses for these cameras are smaller and lighter than their DSLR counterparts, so the entire system is much easier to deal with. The flange distance (measured from the lens mount to the sensor) is also shorter, so these camera have the ability to mount virtually every lens ever made by any manufacturer for which someone has made an adapter. This is a huge win for users who have special lenses from an older camera and don't want to change their entire system to accommodate a new camera body. Sony has been the leader in this area, although Canon and Nikon have recently released their own take on the design. [Updated this info October 2018]
View cameras are similar in appearance to what you might see in an old movie, where the photographer would compose the shot with his head under a black fabric and then insert a sheet of film into the back of the camera before exposing it with a shutter in the lens itself. While these cameras look complex, they are actually very simple in design. A lens board is mounted parallel to the film and focus is adjusted by moving the lens board closer to or further from the film plane (or vice versa). A fabric bellows keeps light from infiltrating the camera and allows the entire unit to be collapsed into a small form for travel or storage. Modern view cameras can be film or digital, with film sheet sizes all the way up to 8" x 10" or larger! Digital scanning backs work much like smaller digital cameras, although the scanning process is relatively slow and can result in strange distortions when shooting moving scenes like water or trees blowing in the wind. Film cameras can be found for a few hundred dollars, but digital scanning backs can cost over $50,000. Ouch!
A note on resolution: Camera manufacturers put a lot of emphasis on the pixel resolution of their cameras' sensors. 10-megapixels (MP), 14-MP, and higher have become the standard for even the lowliest point-and-shoots, with many high-end DSLRs now attaining 24-MP (and recently even 50+MP). All this does is give you some leeway when it comes to cropping the final image. The images will not be significantly sharper or better than photos taken with a 5- or 6-MP camera unless you intend to print them at sizes larger than 12" x 18". The biggest contributors to clean, sharp images are a good lens and careful attention to exposure. I have a print hanging on my wall that was made with a 6-MP Konica Minolta DSLR and it looks fine printed at 20" x 30" (and that was with an old, crappy lens).
So, what kind of camera takes the best pictures? None of them. The best pictures are made by the best photographers, regardless of the equipment used. I've seen images made with a cell phone that rival in emotion and impact those made with a $5000 Canon DSLR. That doesn't mean that a professional can get the same quality from a P&S that he can from a view camera, but the best images can only be wrung from an expensive tool by someone who knows how to tweak every setting to its optimal performance. (By the way, telling a photographer that his photos are beautiful and then commenting that he must have a really nice camera is like assuming Michelangelo used only really expensive brushes—it's an insult.)
Picture-making is all about light control:
Instead of thinking of photography in terms of camera control, you really need to think of it as light control. The camera is just a tool that allows you to control the amount of light that falls on the focal plane. Whether you allow more or less light onto the sensor helps determine the brightness, contrast, and emotion of the final image (composition also determines the impact of the final image—more on that later). While some of these things can be massaged in post-processing software, there are limits to what a computer can do with a too-dark, too-bright, or blurry original image.
Before setting your exposure, take a look at your scene through squinted eyes. Look around for the very brightest part of the scene and pick out highlights and bright areas like reflections, puffy white clouds, sea foam, bright sidewalks, etc.—anything that immediately gets your attention. These bright spots will probably be over-exposed (also referred to as "blown out") in your shot unless you correct for their brightness. Your subject may appear to be correctly exposed, but white clouds will appear as big, pure-white blotches instead of finely detailed wisps. The quickest way to spot an amateur photo is to look for over-exposed clouds.
Once you've found the trouble spots, set your exposure based on the brightness of the brightest thing in the scene. This may be hard to do unless your camera has some way to spot meter a scene, but you can shoot several frames with varying settings until you get one that does not over-expose those areas. How do you know if they're over-exposed? Some of the more advanced cameras have a system that flashes the pixels in the over- and under-exposed areas of the photo on the preview LCD screen after the image is recorded. Some may have a histogram on the preview screen. A histogram is basically a bar graph that shows the number of pixels exposed at every level over the entire image and it's the best way to look for exposure problems after the shot is made. A good histogram will look like a bell curve with tapered ends and a hump in the middle. This shows that most of the image is exposed around the middle of the camera's ability to record light and the highlights and shadows (to the right and left ends of the graph, respectively) are not clipped at their extremes. Once a bright or dark pixel is clipped, no more image information can be stored at the pixel—it is either pure white or pure black, and neither of these is a desired result.
(I might add here that most DSLRs have different means of measuring the exposure coming through the lens. Average meters exactly what it says by taking an average of the entire scene to set the exposure. Center-weighted uses the middle-ish part of the scene. There is also a system called Matrix metering that uses a more intelligent means of figuring out what parts of the scene deserve to be over- or under-exposed and does a pretty darn good job of giving a well-balanced image in varied circumstances. Spot metering uses a very small area at the center of the image while ignoring the areas around that center spot. Most people who really know what they are doing will use the Spot pattern to meter several parts of the scene and then adjust the camera settings based on those different readings. When I'm shooting on a location I will typically meter the brightest thing in the area and adjust the manual exposure based on that reading. After that, I leave the settings alone unless the lighting changes or I move to another location. After a lot of experience I can tell you that 1/1250 second, ƒ5.6, ISO 100 will expose perfectly a surfer at midday without blowing out the breaking waves.)
Preview of recorded image with and without clipping areas (red for over-exposure and blue for under-exposure):
Histogram showing over exposure:
Histogram correctly exposed:
The mid-exposure hump can be smooth or peaked, as long as the peaks don't reach all the way to the ends of the graph. A spike near, but not at either end of the histogram is okay and will actually make for high contrast in the image without loss of detail.
Why not look for the darkest areas of the scene and set the camera for those? People tend to look at shadows and expect to see darkness. You expect to see nothing when you look into a dark cave. When you look at a very bright object, however, you expect to see extreme detail, even if you have to squint your eyes to see clearly, you still expect to see something other than just a big, white blob. Exposing for the highlights allows us to control those bright spots and let the shadows work themselves out. If you have to force a dark exposure to maintain the detail in the highlights, so be it—your audience won't know the difference.
Specular highlights are a different story. These are reflections—usually of the sun or a direct light source—that cannot be subdued with any amount of exposure control (at least, not without sacrificing the rest of the image to extreme under-exposure). If you're shooting in the direction of the sun across a pool of water, for example, there will be glints of light all over the surface of the water. These are going to be washed out in the final image and that's fine. Again, this is something that people expect to see when they look at a very bright source of light.
Over-exposed specular highlights:
The difference between JPEG and RAW. All cameras will capture images in the JPEG file format. This method of storing picture information is known as "lossy" compression because the algorithm used to make the file size as small as possible throws out redundant pieces of information. The problem is in how it decides what is redundant. It samples large chunks of visually similar pixels and stores those chunks as blocks of identical pixels. What may have looked like a gradual gradient, or ramp of colors and shades, becomes in the JPEG file a solid block of one particular color. Where two of these blocks meet, there can be an abrupt change known as a compression artifact.
Most people may not even see these artifacts, but they are there and they degrade the image quality just slightly. Every time the image is edited or re-compressed, these artifacts become slightly easier to see until they are ugly and irreversible.
Another drawback to JPEG's lossy code is that it immediately throws out any useable overhead captured by the camera's sensor. Digital sensors are able to "see" a lot more than the camera records and they don't recognize any white balance settings until after they are processed by the camera's onboard software. Recording in RAW format does exactly what its name implies: The recorded file is exactly what was seen by the sensor at the time of exposure, minus any camera processing. There is no sharpening applied, no white balance adjustment, and no noise reduction (though some cameras do have built-in noise reduction at the sensor-level). The RAW file has all kinds of detail that can be drawn out in post-processing software, but most of it is lost in the JPEG compression.
As an example, a key feature of RAW shooting is the ability to recover details in over-exposed areas of an image. In the last section, I mentioned blown out clouds as a measure of a photographer's skill level. A lot of shots with this defect can be salvaged by shooting in RAW format as the highlight details are still there, just hidden in the file's additional bits. Any software capable of editing RAW formats can extract some (but not necessarily all) of this detail, restoring those clouds to wispy or billowy goodness.
Let's say you're shooting indoors with lights near an open window during the day. The white balance of the scene is tricky and shooting with your camera set to Daylight white balance (WB) will result in the interior of the scene looking very yellow from the tungsten light sources. If you set for Indoor WB, the exterior light through the window will be very blue. JPEG photos will have this WB setting applied and there is only so much you can do to change this setting without loss of details and some strange color shifts. RAW files, on the other hand, are easily adjusted to Daylight or Indoor WB, or anything in between. You can adjust the color temperature to your own tastes until the scene more closely matches what you experienced on the set.
Aperture, or how to control your depth of field:
(The next few paragraphs pertain to all cameras, though the photographer may have limited control over these settings depending on the level of automation built into the camera.)
If you have a lens marked ƒ5.6, that is the widest opening of the aperture (iris) for that lens. (Each full ƒ-stop of a lens doubles the amount of light that passes through to the sensor or film. Unfortunately, they're not numbered logically, so you kind of have to know these things from experience, but ƒ4 is twice as much light as ƒ5.6, ƒ2.8 is twice as much light as ƒ4, and ƒ2 is twice as much light as ƒ2.8, and so on. An ƒ1.4 lens lets in sixteen times as much light as a ƒ5.6 lens at their widest respective settings.) The aperture numbers are actually a ratio of the iris diameter to the focal length of the lens. They represent the amount of light that enters the lens and they are the same for all lenses, no matter the focal length. In other words, ƒ4 on a 17mm wide angle lens lets in the same amount of light as ƒ4 on a 200mm telephoto lens. If lens manufacturers expressed the opening as an absolute diameter, the aperture setting would be different for every focal length and would require complex math to figure the correct exposure. ƒ-numbers aren't the easiest to understand, but they are much better than the alternative.
Most photographers use aperture to control depth of field (DOF), which is the amount of your scene that appears in acceptable focus (from near to far) while using the shutter speed to control the overall exposure. (Obviously, there are times when you need to use shutter speed creatively and aperture and/or sensitivity settings can be used for exposure control.) For wide landscapes, you typically want a very deep focus to get everything from the foreground to the background in sharp focus. An aperture setting of ƒ11 or smaller (down to ƒ32 or ƒ64 on some lenses!) is good for these times. For portraiture, you probably want to use a wider aperture and thus a narrower depth of field to exclude the background by blurring it out. ƒ4 or wider results in only the subject being in focus while anything in front of or behind the subject is heavily blurred. (Apertures wider than ƒ2.8 can yield a focus so narrow that only the tip of the nose is in focus while the eyes are slightly blurry.) The focal length of the lens is another factor that affects depth of field, but most "portrait" lenses are between 75mm and 150mm. Wider angles have wider depth of field and longer lenses have narrower depth of field for any given aperture setting.
Also note that the distance from the subject affects the actual DOF with objects further away having a greater DOF and objects close to the lens having a relatively narrow DOF. In other words, photographing a person standing next to a car results in a wider focus than shooting an insect on a flower.
Wide depth of field (narrow aperture of ƒ14):
Narrow depth of field (wide aperture of ƒ5.6):
Most lenses will perform best around the middle of their aperture range. Even the most expensive lenses will have some slight degradation of quality at either end of the aperture range, but the better the lens quality, the less noticeable this degradation is. The point is, when you use an inexpensive "kit" lens, you might want to shoot at ƒ8 or ƒ11 to get the sharpest possible image at the expense of a wide depth of field, whereas you can shoot a better lens at ƒ4 and still retain sharpness while keeping a relatively shallow depth of field. If you buy a zoom lens with a variable aperture (the lens is marked 28-80mm ƒ4.5-5.6, for example) the lens lets in less light when you zoom to the longer focal length, so the widest opening available at 80mm is actually 5.6 instead of 4.5. Since the lens is probably not at its sharpest at ƒ5.6, you should stop it down to ƒ8 or smaller to get the best sharpness at the full telephoto setting. This may or may not be acceptable for the given scene. A better lens will have a constant aperture throughout its zoom range and a prime lens (fixed focal length) is typically sharper and faster than a zoom lens. (As usual, there are exceptions. Some really nice zooms are sharper than a cheap prime lens.) Shooting with a lens with a maximum aperture of ƒ1.4 gives you the freedom to use a wider aperture and still stay in the middle of the aperture range. ƒ4 on a ƒ1.4 lens is typically sharper than ƒ4 on a ƒ4 lens.
Photographers have a word for the blur effect created by a wide aperture: Bokeh (pronounced BO-kah). The smoother and silkier the blur, the better the bokeh. As it turns out, a perfect lens would produce blurred elements that were evenly out of focus from their center to their edges with a clean edge. While in theory this would seem optimal, in practice this would not look good. So, lens manufacturers go to great lengths to design slightly imperfect lens designs that yield blurring that tapers off to smoothly graduated edges. The best lenses are often prized for their creamy soft bokeh effects. Things like the number and curvature of the aperture blades also affect the blur quality. You can often count the number of blades in a lens by looking at the out-of-focus lights in the background of a night scene—the pinpoints of light will adopt the shape of the aperture (usually five to nine blades). Star-shaped flares will also indicate the number of blades, but the number of star points will be doubled on an odd-numbered aperture (five blades will create a ten-pointed star pattern). This is due to refractions inside the lens and the effect is greater when the aperture is smaller.
I was going to add a few technical paragraphs about acceptable focus and hyperfocal distance, but that stuff is just way too complicated and not really something everyone needs to know. If you must learn it, Google the aforementioned terms and be prepared for a lot of dry reading. And really, if that kind of thing is interesting to you, you're probably not reading this primer anyway.
Last edited: