THIS IS A SUBJECT near and dear to my heart — and my wallet — because it’s essentially what I do in my day job. I don’t design book covers, but do make objects of similar purpose in support of a creative field where the content providers have recently won manumission from the corporate middlemen who heretofore ruled the roost. Eerily similar to the book world, as a matter of fact. I don’t want to be more explicit about my day job because my employer would rather I not associate my (sometimes toxic) political views with his business. Entirely fair. So you’ll just have to imagine.
I’m assuming you understand at least basic composition and design, and have some kind of tools — whether a full-on version of Adobe Creative Suite or a cobbled-together collection of the free, cheap, and bundled. Let’s step on from there.
Design principles I long ago learned to hew to: Size, frame, focus, readability, color, texture, negative space, and fit and finish.
Size: remember you are essentially designing a postage stamp. Your whole design must convey your intent from a tiny canvas at an impossible distance. Everything in it must serve the purpose with nothing extraneous.
Frame: your design must stand out from the background. Place it in a frame that permits stark contrast across the boundary of the frame. (By frame, I mean the “box” in which your image is contained, nothing more. You should eschew most framing devices for a simple, rectangular crop. God forbid you should attempt to use an actual picture frame. That way lies disaster.)
At the same time, you must select elements and size them in the frame for maximum impact. Ex: you want human figures in your design. You should rarely have more than one, though. And you should crop the figure as tightly as you can and still permit the viewer to identify it. George Lucas once said that, in editing Star Wars, he sought to show objects and actions for the least amount of screen time possible that still permitted the audience to see what was going on. Proper framing should also be dramatic. You want to attract attention to your design. Your chances of doing that with undramatic images are slight. If you have a face, crop it tightly to the features — eyes, nose, mouth. Do not obscure features with type, but don’t hesitate to run type over hair, forehead, ears, or below the neck.
Focus: do not clutter your design. If possible, it should be limited to a single object — person, thing, or device (logo, badge, etc). Fancy type faces should be avoided at all costs. Unless you are a professional typographer, you’re probably better off sticking to the basics — Helvetica, Univers, Gill Sans, Franklin Gothic in sans-serif, and Times, Palatino, Century, Bodoni, Garamond in serif faces. Do not use light weight faces. Keep the typeset copy simple — author’s name and title will suffice. A tag line is OK, but it should be puncy. For It’s Dolly’s Birthday, I intend to use a line the doll delivers in the denoument: Sometimes, you have to go to war in the underwear you have on. I believe I am sure I know why and that my reasons are compelling. If I should have doubts down the road, I will cut it.
And, unless you’re looking for a ransom note effect, use no more than two typefaces in a single design. Seriously. I’m warning you! This is one even non-pros know, and they will ding you for it! Which brings us to:
Readability: Has mostly to do with contrast. Black on white or white on black is best. When you move away from these, make sure you are maintaining that kind of visibility as far as you can. In selecting colors of type to go over a background or an image, remember that colors closer together than about 60 degrees around a color wheel of the spectrum are hard for the eye to distinguish one from the other in less than ideal circumstances. Yes, when you have a life-size image in front of you and you can turn it one way or another to look at it, you can tell the differences in shadings of color from fuschia to red to orange, but if you need to rely on a color contrast to render a title readable at thumbnail size on a 72 dpi computer display (or even 96 or 120 dpi), why stick your foot in a bucket to boot? Pick colors that contrast with one another. Not only that, but you must also assure contrasting tints as well as hues. You want a minimum absolute gray scale delta of 25%. That is, when you convert your image to grayscale, you should be able to eyedropper both the glyph shape fill and the immediate surround and find at least 25 points different on a 100-point scale (64 on a 256-point scale). And that is chancy. 50% (128 levels) is better, if you can get there.
Also, though I do recommend you use texture and variability, do so with restraint. Do not make your design too “busy”. Complexity is all very well. I’m not even saying you should escew it utterly in book cover design. I am, however, urging you to ensure that your dominant elements stand out from one another at the extremes. You can render even the cleanest type unreadable by surrounding it with an image or texture with too much high-contrast detail in it.
Color: You should work color the way pros do. That is, work from a limited palette. Include a color in your design for a reason. Pick colors that work together, either by complement or by contrast. You should also be aware that colors render differently in different media. An image printed offset in CYMK will look markedly different from one viewed on a monitor in RGB and still differently when printed on a modern digital press, such as might be used in print on demand. Most of the time, no one will notice. But every once in awhile, the color gamuts of open-loop devices will surprise even seasoned pros, and results will be the veriest definition of not-pretty.
Unless you have a solid creative vision that demands it, I would strongly advise you avoid large areas of flat, solid color. No colors in nature, and none in well-done art, are precisely the same tone from pixel-to-pixel. Pick your colors for natural variation. Gradients, for example, help break up a large flat area and can add drama and intensity to colors. In the realm of visibility, you should avoid what I call “browns.” These I define as colors between about PMS400 and approximately PMS600 in the Pantone system. If you don’t know that system, you should watch the CMYK equivalents of whatever colors you pick. If you see values above 10% in more than two colors, beware: you’re getting into territory where colors can look dull and muddy. Earth tones. It’s not necessary to avoid earth tones, but you should use them advisedly. And be aware that a CMYK spec of 20,85,95,10 is not red, no matter how bright and bloody it looks on the screen. It will come out of the printer as dull and lifeless.
(And, if it does look firey or deep and rich on your monitor, you should do some calibration work or get a new monitor.)
This is all less important if you are designed for an exclusively electronic product. But, if you are working on something that will eventually have to put pigment on paper, pay attention to clean color, and make as many tests as you can to ensure it. Finally, you should work in RGB. Select colors by Pantone number or CMYK spec if you want, but translate them into RGB. There are myriad reasons for this, but here are two: RGB is most likely to get you the colors you see on your screen from your output. And: RGB is a color space, while CMYK is a device-dependent separation spec. If you do not understand the difference (and trust me: many self-designated pros do not), working in CMYK will get you into trouble sooner or later.
Texture: texture is related to color in that it permits variability and takes your design away from the flat and lifeless and toward the natural and professional in appearance. Texture can be achieved in many different ways. Perhaps the simplest is to add an overlay of the Photoshop Clouds filter (or your local equivalent). And, of course, if you have the budget, you can render textures in a myriad of plugins. However, if you are of a more-limited budget, you can find textures all around you. Closeup photos of clay bricks, for example, can yield source images that, with some creative manipulation, can add interest to any design without necessarily revealing the source. A long time ago, before the advent of microcomputers in the graph arts, textures were more graphic and rarer in actual use. As a result, you didn’t see much of them. This can actually be a telltale in judging the age of a design. Flat, untextured designs tend to look a little dated these days, whereas dimensioned, textured designs — even from the elder days — look fresh and up-to-date.
Just as with type faces, it is possible to have too many textures in a design. I do a lot of logo work in 3D CGI applications and quickly learned to use no more than two textures of any type — no more than two stone textures, no more than two metals, two plastics, etc. No matter how you build your design, textures are one area where I’d urge more restraint rather than less. For one reason, textures can be a detriment to readability. Multiple textures can multiply the problem.
And here’s a pro tip. Your texture should be large enough that it fills your entire frame. That is to say: Don’t Tile. Tiles are strictly bush league and are a dead giveaway that there’s an amateur at the controls. Even so-called “seamless” tiles have obvious repetitions that, when spread across large areas, stick out like a lemon in an orange crate.
Negative space: this is one I see even pro designers mis-step on. Back in the ’60s, artists had this drive to fill the field of view with images. Type was stretched to fill all areas that weren’t images. Images were blown up. Full bleed was the watchword. Nobody used borders at all much any more, it seemed, and when they did, the border threatened to overwhelm the design — like Mucha on steroids. Readability went out the window in favor of the eye kick. What was missing? Negative space.
There’s a rule to be derived from this: if your image bleeds, don’t put a border on it. I will be the first to admit I break this one almost daily. But it is something to think about. And the more complex your border, the less you should want to use it with a bleed image.* On the other hand, a clean, white border never hurt anyone. Just sayin’s all.
More egregious is the failure to hew to good margins. Margins are defined as negative space around the content of a page.
What is negative space? Well, it’s basically nothing. That is, when you take all of the things in a frame that are something and eliminate them, what’s left is negative space. Not quite that, because a background with color and/or texture can be negative space, depending on the intended focus of a design. But close. In photographing art mechanicals for prepress in offset lithography, the black areas of the negative — which would not permit light to hit the plate, thus making an image — were referred to as negative space in some contexts. In an ideal situation — the textbook example — negative space is white and the image is black. This is generally true for text, which is usually printed in black on white paper.
And a margin is a special kind of negative space around the outside of a page of content, or a block of it, although that usage is rare.
It should be noted that what HTML refers to as a margin is not. A margin in print design is what HTML code calls padding. Don’t ask me why they got that wrong, but it bugs me, so I point it out here. It’s only one of myriad examples where computer folk don’t get graphic design.
Margins serve several purposes. First and foremost, they contribute to the readability of blocks of text. You can test this by looking at a badly-designed Web page — one designed, say, by and for computer folk. The paragraph tag does not in and of itself have a default margin spacing. A margin or padding needs to be added to the tag in order to lay out negative space around a block of text. If the bare tag is used, the paragraph will butt up against the browser window frame — at least on the left — making it maddenly difficult to read. A very slight margin of 5 or 10 pixels would pull the text away from the frame decorations in the browser window, eliminating the visual conflict between the two, making the text more readable.
Second, they provided printers a clear area of paper by which a page or a flat (of several pages) could be picked up and handled without touching wet ink. Of course, modern best practices obviate such handling and this purpose is not so urgent any more.
Third, they provide a safety margin in both cutting and folding. Pages are printed several to a sheet of paper. (There are many reasons for this, mostly having to do with efficiency and economy in process. If you’re interested, any beginning textbook on the graphic arts can explain it to you. I don’t have the time or space.) When several groups of pages are folded down and lain together in a saddle, there are multiple thicknesses of paper between the end of the type and the edge of the paper at the spine of a book. Control over the size of this inner margin (also called a gutter) permits all of the pages to line up so they don’t flicker when fanned through like a flip book animation. This is a readability issue. (No. No time to explain. Just trust me.)
In cutting, which is relevant to our task, where folding was not — quite — the margin also provides a safety … margin. When any printed job is printed, it is done on a sheet of paper that is somewhat larger than the finished piece. (There are many reasons for this, the explication of this would turn this into a teal deer (tl; dr.).)
You mean it’s not already?
I’m trying to keep it short.
The mechanics of cutting are thus: a stack of paper is jogged up by bouncing it on its edges and fanning air into the stack, allowing the sheets to slip past each other into (more-or-less) perfect alignment. This is one of those lost fundamentals, like those German apprentices who used to spend two years just filing metal blanks. Nobody knows how to jog paper any more. The alignment is critical, because of the next step, which is to slice the stack with a razor-sharp knife, (No hyperbole: lots of writers use “razor sharp” when it doesn’t quite apply; these knives can cut your finger clean off just sitting there.), driven by several tons of pressure (in excess of 10,000 psi). Even with such a sharp blade, the friction of the knife within the stack can cause “draw” — which is the sheering motion along the direction of force of the moving blade. This can cause inaccuracies in the cut. In order to hold the paper steady, it is first put into a clamp, which presses down with even more pressure than the knife (5-9,000 kg). This pressure can be “soft” or “hard”, depending on the compressability of the stock and the amount of air in the lift. If the pressure is too great for conditions, the stock can “belly”, or bulge out of the clamp.
Thus, when the knife cuts through — straight — it actually follows a curving path which, when the clamp is released, will result in a concave edge to the stack. And, on the individual page, the distance from the edge of the content to the edge of the paper will vary slightly. The closer the content is to the intended edge — the tighter the margins — the more noticeable this effect will be. If the design has not been made with an awareness of these factors, it is quite likely that parts of the content will “bleed” off the edge of the paper — whether intended or not. (This can be a quite disconcerting sight when encountered in actual production, with many thousands of dollars on the line.) The wider the margins allowed, the less noticeable this effect. Thus the wise designer will allow for these effects in laying out the margins.
Over the centuries, all of these marginal effects have become ingrained in the tastes and habits of the reading public. That is to say there is no absolute reason in natural law that margins should be left, except that work without margins tends to look unfinished or badly-finished, and thus unprofessional. At life-size on a cover, (say… 6″ x 9″), you can probably get away with a 1/8″ margin, though wider would be advisable. But, when that margin is shrunk down to Amazon-standard thumbnail size, it will look cramped. So you’re wise to leave a slightly wider margin (I’d say at least 1/4″) at life size, so that, when you shrink your design down, there will still be a visible — if not measurable — margin.
Now. Is this to say you should not run type right to the edge of your design? Of course not. It works. It’s quite dramatic. It’s highly visible. But you should make sure that it looks like you intended it that way, and not that you just ran the type to the edge with no rhyme nor reason for it.
Fit and finish. Just as your prose should have a solid “thunk” to it, like the closing of a door on a new car, so, too should your visual design. Almost any sin can be forgiven if the execution of the whole is clean, clear, and has the right finishing touches. All of the covers Sarah calls out in her post fail on this score, no matter how else. Yes, the typeface selected for Neewa the Wonder Dog is an abomination of inappropriateness, the frame is all wrong, the focus is horrible, but the worst sin is it looks like it was thrown together at random. Even as a rough schematic, it fails miserably. If you’re going to set type at random, with no margins and no justification, no centering or alignment, then you should at least try to make it look as though you had a reason for doing otherwise. I’ve staggered type across a frame a brazilian times. But I always did it to balance other elements, or to draw the eye down a specific path through the design.
In short, you should be aware of everything you put into a design, have a reason for its size, conformation (color, etc), and be certain you can’t find a better way.
And, still, somebody will call you out, but at least you will know you have given it your best effort. And one of the joys of indie e-pubbing is, as Sarah points out, you can go back and change things if you’re not happy with them. And you can do it without breaking the bank.
Obviously, I cannot core dump all of the lessons learned in a 30-plus-year career that is not over yet. However, I don’t mind showing off what I know. So, if you’ve got questions, ask in comments. If you disagree with my conclusions, say why — in comments. If you have something to add, pipe up. In comments. Just, please, make your comments substantive.
*(“What’s bleed?” I hear you say, and can imagine you picturing the colors all running like a watercolor in the rain. But no. It’s where the image runs off the edge of your design. It “bleeds” out of the frame. (In actual printing, the image runs past the crop mark for a specified distance — usually 1/8″ — and the paper cutter actually cuts it off and throws it away.))