The Art of the Frame

By Glenn Campbell

An essential problem of art is where your work should begin and end. If you are taking a photo, what should you include in the frame and what should you leave out? If you are editing a movie scene, how long should it be? If you are writing a piece of music, how long should each passage be and how many times should it be repeated? If you are writing a book or essay, at what point have made your point and are now beating it to death?

Many works fail not because the raw material is inadequate but because it is framed wrong. We all have all set through movies that have some good parts but just goes on and on and on. The good parts are eventually drowned out by the rest of the movie and we walk away exhausted. The artist might be including too much or too little in the frame or perhaps he is ignorant of framing altogether. You could say that framing is half the job of art. It is not just what you produce but how you edit and present what you produce.

I could probably write a whole book on the subject of framing, but the core idea is pretty simple: Every work needs a human focus, a single point of reference within it that he automatically identifies with. Once you know what that human focus is, framing will be dictated by it.

To illustrate this concept, I will use photography, but similar analogies apply to almost any art.

Grand Canyon Landscape

Imagine the Grand Canyon. You have seen countless photos of it. If you go there, you are probably going to take your own photo, but it will be probably be no different than the thousand you can pull up on the internet right now. A photo that shows only the Grand Canyon is almost meaningless, because you have no perspective. You can't even tell how big it is! If you have an expensive camera and some fancy editing software, you can probably make it look dramatic, but even then the viewer's eyes glaze over very quickly. Beautiful but boring!

Now imagine a photo of the Grand Canyon with a single person in it. He or she is a tiny figure in the lower corner of the frame, looking out on the vastness of the canyon. Suddenly, the Grand Canyon means something! The tiny person is someone the viewer can identify with. The viewers is thinking, "That could be me!" Now the canyon is not just mindless erosion anymore. We can now measure it, get lost in it, be frightened of it, be in awe of it.. The simple addition of a tiny figure makes us capable of processing it.

Here is a simple rule for better photos: No matter what you are taking a photo of, make sure there is a human or human-like presence in it. Furthermore, there should be only one human focus, even if you are taking a photo of a crowd. The viewer's attention should be drawn to one specific part of the photo, and the whole rest of the photo is judged by that reference point.

The human reference does not always have to be an actual human. It can be an animal that looks human, like a dog or cat, or it can be an animal that humans hate like a spider. It can be a lone tree on a cliff, because that has a human feel to it. It can be something that implies a human presence like a tiny cabin in the woods. There are all sorts of things that can serve as stand-ins for humans—a car, a sign a cactus, an empty road—but the human center always has to be there. You can call anything "art", but I contend you don't have memorable and emotionally compelling art without a human in the picture.

Once you have a single human in your Grand Canyon photo, how do you know where to crop it? In this case it is pretty easy: You want to show as much of the canyon as you can while still allowing the viewer to quickly identify the human figure. (The photo above is a little weak because the humans are the same color as the background.) Obviously, if you show too much canyon and too small a figure, viewers are not going to see the figure. If you show too little canyon, then you've wasted all that canyon. The message of a photo like this is, "Look how big it is!" so you want to show as much of that bigness as you can while still preserving the human reference point.

Your cropping will also be dictated by practical real-world considerations, like the size of the final print. If you are preparing a wall-size mural, you are going to crop the photo differently than if you are displaying the photo on Facebook. With more attention space at your disposal, you are probably going to make the human figure smaller, just to maximize the apparent size of the canyon.

You are also constrained by your raw materials. You can't crop the photo any wider than the photo you originally took, and not all of the frame may be usable. Maybe there is a distraction on one side of the original frame that you would like to keep out—like some tourist picking his nose. You are going to crop him out, even though you would have liked more canyon.

Real world photography (and all art) is filled with these practical constraints—working with what you've got and keeping distractions out of the frame. Thanks to the magic of editing, the viewer is never going to know about that tourist picking his nose. You may have lost some of your spacious canyon but you probably haven't lost much in overall impact.

Why do you want your human subject in the lower corner of the photo rather than the center. Simple: You want the biggest canyon possible within the constraints of your frame. A reference point in the center sometimes works, but in this case you would be splitting the canyon into two smaller areas and it might not feel as big.

Grand Canyon Portrait

Aside from shooting the Grand Canyon without reference points, what other framing mistakes to tourists make? Here's one thing you see all the time: They take pictures of each other in front of the canyon, with the canyon in the background and themselves in the middle foreground where we can see them almost from head to toe. We can hardly make out their faces, but they are apparently smiling and obviously saying to the folks back home, "Ha, we made it to the Grand Canyon and you didn't!" (BTW: None of these Grand Canyon photos are mine. All swiped from the web.)

What is wrong with this picture? It is the same bland photo every tourist takes at every tourist attraction. While there is certainly a human reference point in the photo—the tourist being photographed—it is not being used to the best effect.

There are two ways you can use a human subject. The subject can give perspective to the background, as in our Grand Canyon landscape photo, or the background can be used to give perspective to the subject. Let's be honest, that shot of you at the canyon isn't about the canyon; it's about you, so let's frame it that way. Bring the camera to within two feet of the subject, the same way you would shoot a a close-up portrait with no canyon out there. The Grand Canyon will appear blurred in the background and will be almost unidentifiable.

"Almost" unidentifiable is the key. You going to frame it in such as way that just enough of the Grand Canyon appears in the background that it is identifiable and authentic. The intended effect is to understate your presence at the Grand Canyon, which paradoxically heightens its value. You are no longer a rube tourist from Iowa. You are a seasoned world traveler who just happened to be at the Grand Canyon when this portrait was taken. See the difference?

You see this in fashion photography: models cavorting at base the Eiffel Tower where you don't actually see the whole tower, maybe only a bit of the base, just enough to reveal subtly to the viewer where we are. The message is that we just happen to be in Paris. It's no big thing. The more you understate a visual asset like this, the more you imply to the viewer a whole big world beyond the frame—much bigger than reality!

Framing in Film

There are similar analogies in film (apart from the obvious ones of deciding how to photographically frame the shot). For example: When something disastrous happens or an evil character does something outrageous, you have to have a sympathetic human in the scene to give that action perspective. There is no point in Darth Vader using the Death Star to destroy Alderaan if Princess Leia isn't present to witness it. Leia is our human reference point. We are going to see what she sees, and the scene will end when the emotional impact has been made on her, not us.

In film, things shouldn't just happen, with an omniscient camera recording them. Most significant actions need to be witnessed by somebody. The audience is sees the story through that person's eyes (even if that "person" is R2D2). The scene is presented from their perspective, even if it isn't shot from their literal point of view. That is the reason Darth Vader has sympathetic human commanders under him. Even though they are his henchmen, we see the horror of what he does through their eyes. A crew of mindless droids would not serve the same purpose.

Natural Boundaries

In photography, here is another framing issue: Every original scene has natural boundaries where it can be most easily be cropped. Even if you are shooting a random forest, there are always gaps in the trees where it feels right to crop. Once you decide the general boundaries dictated by character focus, you can look for natural places in the frame you will make the actual incision.

For example, note the golden frame in the main illustration at the top of this blog entry. (This is my own photo taken yesterday in Puerto Rico. Album) Notice that I placed the frame just above the palm trees, so there was a little bit of blue sky above them. In general, you want to crop out any useless empty space, like too much blue sky above the palm trees. When you have useful information available, you keep it. When you run out of information, you crop it out. (You can see how I actually cropped the photo here.)

However, sometimes you do NOT want to crop at natural boundaries. Sometimes you want the opposite: you want details to bleed off the edge of the frame. This takes advantage of a powerful human illusion: Whatever you see bleeding off the edge of the frame is assumed to continue indefinitely. Look at the still of Princess Leia above. Notice the simple fact that the background continues off the frame. This leads the viewer to believe there is a whole Imperial Star Cruiser beyond the frame. In fact, the set may end six inches beyond the camera's view, but because we don't see it end we assume it goes on forever.

If you are shooting the Grand Canyon, you want to give the impression that it goes on forever, so you let it bleed off the edges of the frame. This also how a filmmaker can make a group of 20 extras seem like a mob of 10,000. They cut people off at the edges so you have only half a person on either side. This is not the natural way to crop, but it is the way you do it to imply infinity.

This is one of the ways photos can lie. They imply an infinity beyond the frame which may not be real. Even reputable photojournalists will show you a close up of 10 protesters, making you think there are 1000. I am unashamed about using this illusion myself. What I do is art, not journalism. While I rarely use Photoshop to change details, the impression created by a photo almost never matches the original reality. That's just the way art works. You can't let reality get in the way.

The framing concepts of photography also apply in analogous ways to other arts like writing, music and public speaking. You have to have an emotional focus—some person you are playing to. You're not just stringing words or notes together because they sound good. You working within an emotional frame of reference that was established by sticking a human in there someplace.

Only humans have meaning. Without them, all you have is erosion.

Glenn's Photo Management System

One of my Facebook friends, Gavin Payne, writes:

How about a new blog entry on how you process all of your images, index them, make the libraries, add text to them etc?

As you wish. Let's see if I can give you the highlights in 15 minutes...

1. I take a LOT of frames, only a tiny percentage of which I ever show to the world. The pictures are all free on a digital camera, so why not?
2. On my laptop, I store photos in directories by year and month and setting. e.g. inside directory "2009" is a directory "September" which has a directory "BAR HARBOR".
3. Inside the setting, I have four working directories: "raw", "web", "UPLOAD" and "UNPROCESSED".
4. When I download the photos from my camera, they go into "UNPROCESSED" and are immediately deleted from the camera.
5. I go through UNPROCESSED at my convenience, looking for good photos. I'll first "cherrypick" the very best photos, then I'll go through the rest of them as I have time.
6. Each photo I choose will be cropped, corrected for color/darkness/etc and resized to 604 pixels across, as suitable for Facebook. (I use Corel Paint Shop, the cheaper equivalent of Photoshop.)
   
7. For the clear, crisp quality, I "sharpen" at 604 pixels. (Makes all the difference in the world.)
8. I save the edited photo under the same name in "UPLOAD".
9. I move the raw photo I just edited into "raw", along with any original photos I know I won't be doing anything.
10. Facebook is my main album medium. (I once had my own album system, but Facebook does it better.) After I upload the "UPLOAD" photos to Facebook, I move them into the "web" directory.
11. If I have time to edit the whole batch, I'll end up having all the original photos in "raw", the upload photos in "web" and the other two directories empty. Then I delete those two directories and have only "raw" and "web" left. I'm done!
12. More likely, however, I will still have some "UNPROCESSED" photos left by the time I move to the next project. I could come back to these later, but probably not. C'est la vie!
13. As I pass through my parent's house once or twice a month, I back up my new monthly directories onto some terrabyte hard disks I have. Once I have backed each directory up on two or more media, I can delete the "raw" directories. (I keep the "web" directories because they are small.)
14. My on-line index (http://roamingphotos.com/main) is of my own construction using Perl. (Remember that I used to be a programmer.) I can't easily explain how it works, but it all routes back to the albums on Facebook.
15. For every album, Facebook provides a public URL that anyone can use to access the album, even if they are not on Facebook. (Look at the bottom of the album's page.) I use that address but don't have complete confidence that the address won't change. (It has in the past.) To protect myself from future address changes, I have an intermediate system that translates my own preferred address into Facebook's address. For example...
    
    
    http://roamingphotos.com/a?barharbor
    
    ...is redirected by my system into...
    
    http://www.facebook.com/album.php?aid=2008128&id=1003315385&l=410397cfbd
    
    
    I have a spreadsheet table I maintain that has both addresses, as well as some other info about each album, and this is what my online indexes are generated from.
16. My Facebook albums roughly correspond to my monthly directories (e.g. "Bar Harbor"). At the end of each directory, I have a bumper image...
    On that page, I provide my preferred public URL, as well as links to my photo home page and any other albums that are related to this one.
    

That's my system (or at least all I can think of).

Most people can do everything I can do except the fancy index, but most people don't have hundreds of albums like I do, so it doesn't really matter. You can always create a similar index in html using Facebook's public address, since it has been stable now for over a year. (You could just copy my table at http://roamingphotos.com/main, edit the html and plug in your own album information. I won't object.)

Glenn's Photo Tips

Here are my tips for taking good photos, as expressed succinctly on my Twitter feed:

#1: CROP. Cut out extraneous data along edges of photo, either in-camera or in post-production.

#2: SEPARATION OF ELEMENTS. Each person or other subject should be distinct from its surroundings. Move to make that happen!

#3: SINGLE FOCUS POINT. Each photo should have only ONE center of attention, no more. If two things are competing, cut one out.

#4: COMPRESSION. Select a viewpoint that compresses the scene into a tight area. E.g. A whole mile-long train seen from the front.

#5: ILLUSION OF DEPTH. Always put something in the foreground and something in the background.

#6: HUMANIZATION. Every photo needs a human or human-like character to give the scene perspective.

#7: HIGH CONTRAST. Search for bright colors and high contrast between colors. Avoid dull grays.

#8: IRONY. Seek the outrageous and that which is unexpected for the situation.

#9: FIND HIDDEN MESSAGES. Look for messages in the juxtaposition of objects. Change your viewpoint to bring these items together.

#10: TAKE A LOT OF FRAMES. Shoot first, ask questions later. In the digital age, it's all free, so why not?

#11: REMOVE DISTRACTIONS. Frame or crop to exclude distracting objects, or Photoshop them out.

#12: KEEP SUN BEHIND YOU. Whenever possible, stand with the sun behind you for best light.

#13: ILLUSION OF MOTION. Every photo should be "going someplace" with its main character engaged in an action.

Most of these elements are present in the photo above (from Rome, see larger version). The girl in pink is the reference point. There's depth. There's motion. Most of the people in the photo are nicely separated. The "irony" element is that this place looks surreal, yet it is real.

Also see my Guide to Photo Cropping.

Permission to Photograph

This lady in Dublin said, "Take our picture!" so I did.

One of my Facebook friends writes: "Hey Glenn, I was wondering if anybody got upset about you taking pictures of them without their consent... Have you ever been in a situation like that before? If so, what happened?"

This is a complex and interesting question: Do I have a right to photograph people without their permission, and how far am I willing to go?"

(My photo albums are found at RoamingPhotos.com.)

I have basically two kinds of human subjects: (1) people deliberately making a spectacle of themselves—say, by dressing in costumes and participating in a public celebration; and (2) people going about their daily business in the place where I happen to be photographing.

A good example of the first case is a renaissance fair, where people dress up in outlandish medieval customs intending to be noticed. That, to me, implies a permission to photograph them. I am careful not to make people uncomfortable, but I have no discomfort about shooting.

The second case is more complicated. If I photograph, say, a street vendor without his permission, am I violating his privacy? In the U.S., at least, there aren't any specific laws against it. It is well-established, for example, that celebrities can be photographed whenever they appear in public, and they don't have to be paid royalties for it, which is the legal principle supporting the paparazzi. The same general idea applies to anyone appearing in public. Legally, I never fear that I will be arrested for what I do.

The more important issue, however, is making people uncomfortable, and I do my best to avoid it. The fact is, the vast majority of the people in my photos don't even know there are being photographed. It happens so quickly (in a fraction of a second), that they are only dimly aware of it at most, so there's no opportunity for them to become uncomfortable. Usually, I appear to be photographing something else, like the Grand Canyon, and the people in the foreground aren't aware that they are my main subject.

A lot of the apparent intimacy of my photos happens in post-production. My original photo shows a much wider scene, but I crop it down to the small part of the scene that is most interesting. When it appears in the photo that I am close to a person and interacting with them, I am usually very far away and just happened to catch them when they are looking in my direction. They may see a guy with a camera, but they probably don't know he is looking at them.

I don't feel uncomfortable about "stealing" photos like this, because the chance of the subject or anyone he knows ever seeing the photo is very slim. And I never portray people in an unflattering light, anyway, so it shouldn't be something that would upset them if they saw it. If anyone did see themselves on one of my websites and asked me to remove them, I probably would, but so far no one has asked.

Occasionally, though, I may actually engage with my subject. This may happen, for example, when a bunch of kids (or drunk adults) see that I have a camera and they start hamming for it. If you start performing for my camera, this implies a permission to photograph, and I will take advantage of it.

Generally, I only photograph people under circumstances where photography is normal, like at tourist attractions or public events. This gives me the "cover" I need to focus on what really interests me: the people. I don't photograph in circumstances where I would be drawing attention to myself by doing so. The essence of candid photography is that the subject be at ease, so I don't want them distracted by my presence.

In practice, it is virtually impossible to get "permission" to photograph. Once I ask, "May I photograph you?" people become self-conscious, and it destroys the spontaneity of the moment. It also starts raising questions, like "What are you going to do with these photos?" which requires a long, complicated negotiation. I couldn't take a lot of pictures in an amusement park if I tried to ask permission from each person who appears in my viewfinder. Instead, I just shoot and move on. Usually, it happens in a split second, and I have completely left the area five seconds later.

I do obey some general rules:

(1) I don't photograph people in undignified positions. I try to draw a line between "capturing" people and "exploiting" them.

(2) I don't photograph people where there is a "reasonable expectation of privacy."

The last concept is a vague one, but I generally don't photograph people in busses or subways or engaged in activities that are generally closed to the public. I wouldn't photograph people at a funeral or at private events I haven't been invited to. A general rule of thumb is that if a regular tourist wouldn't be photographing the scene, I won't either, but if a lot of other cameras are present I will feel free to snap away. This is why the majority of my venues end up being tourist attractions.

Occasionally, people will ask me not to photograph them, but usually these are very paranoid characters who are hyper-vigilant for any perceived threat in their environment. Usually, I don't give them the chance complain. If I dwell too long in one area or point my camera in one direction for an extended time, I might get a complaint, in which case I promptly leave the area or put the camera away.

In my photography career, only handful of people have asked me not to photograph them. These incidents stick out in my mind, and I have adjusted my methods so I don't trigger this kind of response. I may walk down the street going snap-snap-snap, but by the time a paranoid has a chance to react, I'm already long gone.

I feel that the benefits outweigh the concerns. I usually believe that something important is being captured that would slip away otherwise.

The Fraud of Photography

By Glenn Campbell

Photos are full of lies! As a photographer, I know it well. Even if it isn't doctored in Photoshop, a photo can seriously deceive the viewer. A photo is the equivalent of interviewing someone for an hour, then broadcasting only a few words of what he said, giving an impression that could be entirely different from his actual opinions and the rest of what he said.

The essential fraud of photography is that the viewer assumes the photo is a representative sample of the whole of something, when in fact it is a highly selective. The objects or events being portrayed may be exceedingly rare, but the photo makes us believe they are common.

If you see a photo in the newspaper of the President frowning at a news conference, you can rest assured that he really did frown—for at least 1/250th of a second—but that expression may not have been representative of his demeanor during the rest of the news conference. Thus, the photo, although recording real events, may not be truthful journalism.

The fraud can take place over both time and space. In time, you are taking a slice of life in a fraction of a second, which might have nothing to do with what a person is doing the rest of the time. In space, the camera is looking in only one direction, while something totally different could be happening in all the other directions.

In both cases, the viewer invariably assumes that what is going on outside the frame is merely a continuation of what is happening inside the frame. Thus, he may be shocked and disappointed when he experiences the reality for himself. It could be nothing like the photo!

When someone publishes a photo of something, they are showing only the most appealing or unusual part of it, intended to serve their economic, personal or artistic agenda. Every photographer has an agenda. Photos rarely show the dull and ordinary parts. A photo is like a tourist brochure that shows only the most appealing (or least appealing) parts of the city and tells you nothing about the rest.

The photo CAN be representative of the whole, but more often it is not. This is because the photographer is mainly concerned with the aesthetics of the photo itself and not what it represents. The photographer is selling a product and has an incentive to create drama and interest in it. His product is "reality-based" rather than reality itself.

If you want reality, a photo can be a helpful starting point, but to see the Big Picture, you'll probably have to be there yourself.

Glenn's Guide to Photo Cropping

"Before"

"After"

Cropping is the most important editing you can do after a photo is taken. When you take the photo, you don’t have much time to think, so you just want to make sure you’ve captured everything you might need later. (Shoot first, ask questions later!) Later, in “post-production,” you want to focus the viewer on the most important elements in the frame and cut out all extraneous material. Cropping alone can turn a mediocre photo into a fantastic one.

How do you crop? You can use Photoshop or Corel Paint Shop, but the software package that came with your digital camera can probably do it, too.

Some rules of thumb when cropping:

1) Make the photo just big enough to include the essential elements of the shot. Cut out everything that isn’t necessary around the edges.

2) Crop to remove distractions from the periphery of the photo (like parts of another person’s body). Sometimes, it is better to cut off a non-essential part of the subject than to let a distraction intrude.

3) In general, the more you fill the frame with the subject, the better. Why should the person occupy 1/4 of the frame, when he can occupy almost the whole frame? This is true even with tourist photos. At the Golden Gate bridge, don’t focus on the bridge but on the subject in front of it!

4) The main subject should usually be centered, unless there is a good reason to place him off-center.

5) Crop for the resolution the picture will be displayed in. The crop for a thumbnail or an online photo is going to be different for a poster-size print, where more detail is available.

6) There is usually no need for “headroom” (extra space) above the subject (as some photo books advise). Crop just above the top of the head—sometimes even below the top of the head for maximum impact. (The only things that are really essential about people are the eyes!)

7) If the subject is looking to one side, you SHOULD provide some empty space for him to look into.

8) The same applies to an object moving in a certain direction (like a race car): You want to give the object space to move into, thus implying motion.

9) Don’t be afraid to cut off non-essential parts of a subject or photo element. (If looking at a warning sign, for example, it is okay to cut off the corners of the sign where there is no print.)

10) When you are planning to print a photo, always crop for a 4x7 print (the standard at most photo labs). If you want to blow up the photo to 5x7 or 8x10, you can usually still use the 4x7 image file, but keep in mind that portions of the left and right sides of the photo are going to be cut off in the larger print.

11) When cropping for BOTH prints and the web, crop for the 4x7 print first, do any correction for color, etc., then crop again and resize for the web, saving two separate files.

12) When cropping for the web, start with the 4x7 aspect ratio (1.5:1), but then consider cutting off portions of the left or right to make the image bigger and more square. (Facebook, for example, limits photos to 604 pixels wide or high, so if you make the picture more square, you can squeeze more pixels in and essentially make your picture bigger.)

13) Try to avoid vertically oriented photos (where the vertical dimension is longer than the horizontal one). It is more natural for the eye to move side-to-side than up and down (which is why movie screens are horizontal).

14) Crop to imply “infinity.” For example, if you take a photo of a crowd of twenty people, and you include empty space along the edges of the group, the viewer will think the group is small, but if you cut off the edges of the group, the viewer doesn’t know where the crowd ends and assumes it goes on forever. Wherever the photo edges end, the viewer is going to assume that the pattern seen there goes on forever.

15) Crop to make the image more universal. Sometimes, you don't want elements that reveal too much about time, place and circumstance. This gives the viewer the opportunity to fill the scene with his own experiences and emotions.

In the example photo above, the core image of the father and his daughters is good, but the impact is diluted by all the extra space on each side. At the same time, there were two distracting elements: the minivan on the left and the street name above the man's head. (I initially wanted to keep the sign, but I soon realized it was too specific and interfered with the universality of the scene.)

The minivan forced me to place the subject off-center, but as soon as I did, I realized that the road should be the frame of reference, and I pushed the subject even further to the left. My crop on the right was determined by the yellow trees in the background: I cut off the photo just before the trees became duller.

Of course, I also pumped up the color a bit (by adjusting the hue and brightness). Overall, the cropped scene is more perfect and idealized than the original one and thus more appealing.

Art and the Sensory Sphere

As a prolific photographer, I have had plenty of time to contemplate what makes a good photo. It isn't the subject matter but your selection of the subject matter.

If you put me in the middle of, say, a kindergarten classroom and give me permission to take photos, I guarantee that I will come up with 100s of great shots in an hour or less -- provided I have a zoom lens and an opportunity to crop and edit those photos later. Furthermore, I don't have to move from my location in the center of the room to find these compelling images. They are all around me; I just have to select them.

I call the input all around me the "sensory sphere". It is a little like lying on the ground at night looking up at the stars. In theory, there are a million stories up there; you just have to have a powerful enough telescope to see them. If you are in a crowded room, there are also a million stories. You just have to focus on the right places.

To be a great artist, you don't have to go out and seek stories. They are already all around you; you just have to select them from the environment. The real challenge isn't usually in the material available but in your own skills in recognizing and editing it. That's essentially what a photographer does: He edits the sensory environment around him. It's also what a writer or filmmaker does. If they are skilled enough, they can find a story anywhere. Only the unskilled need to travel the world or have huge production budgets.

Some environments are richer than other, but there is always good material in front of you. You can say, "I live in a boring town. There are no stories here," but that's not true. There are always stories right under your nose; you just have to have the wisdom to see them.