IN THE LAST post, you met Dolly’s eyes. Now I get to work to develop this basic image into a book cover illustration. I open the original JPEG file, downloaded from Dreamstime, in Photoshop and immediately resave it under a new name eyes.psd. This is a PhotoShop Document, which uses pack bits (lossless) compression and supports layers with variable-transparency and various built-in blending effects. I will be using these tools to alter the image to suit my needs.
Warning: icky sex talk. If that squicks you, skip the next paragraph.
The original idea is to focus on Dolly’s eyes, looking up in love at Drummond (whom we do not see) at the moment he penetrates her while they are making love. It is an expression most men will know and most women have assumed — whether or not either is sincere, though in this case, I think Dolly is. It is a wide-eye expression, almost one of surprise, with the eyes turned up, as the man is usually looming over the woman at this moment. (Both of them in the horizontal position.)
OK. It’s safe to come back, now.
The image is to focus on and isolate the woman’s eyes, even stylize them a bit, and render them in what’s called a high-key image on a white background. I figured that I would have to have a source image to modify by painting in a pixel editor. I don’t draw well enough to satisfy my perfectionism and don’t-have-slash-can’t-work-with a model. So I did the stock search described in the previous post.
And came up with something somewhat different. But close enough. I think I can modify this image in Photoshop (perhaps with an assist from a 3D app — modo or Poser — to get the eyeballs themselves JUST right, but we’ll see). But the really interesting part is the torn paper frame and the silver paper background. It intrigues and offers new directions. I may abandon both, but I will proceed in a fashion that permits me to do either ad lib and only once I have seen both images and chosen between them.
It is experience and practice which allows me to A) do this, 2) see that it can be done and 3) how to do it so as to preserve my options to the end and D) — perhaps most importantly, that it’s desirable to do so. I’ll refer back to this a lot. I’ll call this concept preserving my options.
So there are two basic options — work with a very high-key image on white (I’ll show you what that means in a bit) — or with the background, either as it exists, or with an entirely different look. No, that’s not three. That’s One and Two, with a possible Two-B. The reason to think of the either-or should become clear down the road.
The first thing to do, then, is to save under a new name, which I’ve done. Then the next is to prepare the image to be worked over. In order to do that, I’m going to establish my frame and pull the thing apart into layers.
Experience teaches me that, if you don’t start out establishing two factors at the very beginning, they will devour you in the end. The first, as we say at the Patch Factory, the most important specification, is the deadline: when do you need it.
Weren’t expecting that, were you?
I’m not surprised. Even experienced, sophisticated, and knowledgeable people come acropper of this problem. But time is a key factor in any creative project. How much time there is to do the job determines each choice made in a binary fashion — yes or no, do it or not — and most fundamentally at every step of the way.
In this case, the goal is to have an ebook on sale in the KDP Select program by September First. Now, there are a lot of stumbling blocks, not least of which is that the text may not be ready. BUT… if it is ready and the cover is not, then the cover is holding up the release, whereas if the opposite is true, well, nothing to do but push the deadline back, but that decision is being taken for the right reason — that the product itself is not ready, and not that the packaging needs work. Remember that; it’s key. The book is the thing. The cover is packaging and of secondary importance UNTIL it becomes the deciding factor in a sale. In any case, the cover cannot be permitted to drive or drag on the on-sale date.
(Aside: A friend who is trad-pubbed did just that — she and her agent forced her publisher to change the cover. The publisher pushed her book back a whole year. Not in retaliation, mind. It’s just that that was the first available slot.)
Now, the way to establish deadlines and schedule is to find a delivery date and work your way backwards until you get to now, at the start of the job. It is (as I write this) the Second of August, so I have twenty-nine days. I need to lead my on-sale date by an as-yet-unknown period. Dean Smith asserts he can have a book up in minutes. (Fifteen, if memory serves.) That assumes, I imagine, that you already have your account set up at Amazon and know what you’re doing in clicking the “Publish” button. But it doesn’t seem as though the lead time should be more than a day or two at most. I know that I can do the work necessary to producing this cover art in under two weeks. My editor claims that he can have the MS back to me with notes and edits in two weeks. That seems to indicate a timeline that has the final prep of the work happening between the Fifteenth and Thirtieth of the month. Shouldn’t be stressful.
So the deadline (checks the calendar) is the Sixteenth. Mark it on the calendar and set a reminder five days prior. We probably won’t need the reminder, since this project will be pretty much our sole focus this month, but you should never rely on that kind of stuff. Always be disciplined and organized about your work and you’ll avoid that overwhelmed, too-much-to-do feeling.
OK. That’s the deadline. But you’ll remember, I said there were two factors. The other may be more obvious, but, again, you’d be surprised by how many putative pros get this wrong.
In — scorn quotes — “fine” art, you can wing it. You can frame your image or object ad hoc as you go along, as the mood moves you. The canvas can be any size, and the image wherever and whatever size you plunk it down. In commercial art, however, what you’re doing, first and foremost, has to serve a purpose. And, in most cases, it also has to be made — usually by someone else. The image has to be producible in a medium that costs thousands of dollars to throw the on switch on. And, all the digital cheerleaders notwithstanding (I kid you not) digital costs the same to operate whether you’re printing something or not. Digital printing costs pennies per copy. The cost per widget only adds up when you have multiple copies per widget. The prep cost, far from being two thirds of the cost of a printed project old-school, is ninety-to-ninety-five percent of the cost new school.
This means you have to get it right, and you have to get it right the first time. Your trim size has to be right, as do the bleeds and margins, and the creep allowed for in folding signatures, and… and … Or the job comes out — in technical terms — fucked up.
Now, here again, we’re having to push against some unknown factors. Of course, there is no size for an ebook. Every ereader and ebook format assumes a different screen size. So what size is the cover? I’m going to side-step that for the moment and assume I’m going to eventually do a paper edition and will want to use this image for the cover of that. And, at CreateSpace (and from other sources) I see that the most common advice seems to be to think in terms of a 6″ x 9″ layout. No problem. I can deal with that. So my trim size is putatively 6″ wide and 9″ high. That makes my bleed size 6.25″ wide and 9.25″ high.
(A bleed, for the uninitiated, being 1/8″ of an inch margin OUTSIDE the trim. And, if one side bleeds, then a bleed image — even if it’s just blank white — must be provided and accounted for on all four sides in the engineering of any job.)
But my image is square. Ish. That means I need to either crop it or add image to make it taller. But before I do that, I realize that, because I have (at this stage, pretending we haven’t already played with the type) no idea how much space the type will take up, I need the ability to “float” the image in the frame, I need to separate the image from the background and make it a layer. In Photoshop, I do this by finding the layers palette and double-clicking on the layer. I get an option to rename the former Background Layer as Layer 0, which I accept. The layer the image is on (the only one in the file so far) thus becomes Layer 0, a floatable layer, which I can work on independently of the other layers in the document (as they come into being). It can also be blended with other areas, depending on the relative modes selected. For now, it is in Normal mode, which makes it the effective background.
OK, now. If you’re following along at home, or have a really good grasp of what’s going on, you may have noticed that, once we get the image settled in with right and left bleeds indicated, that the margin to the corners of the model’s eyes is only a quarter of an inch. Which is acceptable, but it worries me. I don’t know HOW much time I’ve wasted over the years faking in image to meet an acceptable margin and/or bleed because the important image was JUST that much too big for the live area, had to be shrunk to fit, and left insufficient, as I say, margin or bleed. Now, what I could do would be to fake in bleed left and right now, at the outset. But instead, I’m going to cheat.
Well, I’ve already cheated.
Remember that I told you I’d bought three images? Ever wonder why? Well, it’s because THIS one is not the original This has been modified by the photographer with the addition of the torn paper frame. It indicates what his vision was in making the original photo, but it’s not the thing itself.
That is an image of the model’s face, full width, from about the bridge of her nose to just below her hairline (which is how I could tell fersher she’s a redhead), with her hands held up flat either side of her face like she’s shading her eyes against oblique light.
This is valuable because there is more image in the width, but also in the height, either of which can prove useful in my repurposing of the image. Also, even though it probably won’t change my crop, that there is additional image to the sides can potentially save me a lot of trouble.
What I want to do is use the original, but treat it in some ways like the modified image. In order to do that, I need to drop the original down on the modded pic and match the two of them pixel-for-pixel in size and position. Since I really want to work with the original — always work with the original if you can — I drop the mod onto it. And realize that, in pixel dimensions, the mod is smaller than the original. This is good, because it means I’ll be able to work at an even higher resolution than I had originally thought, and at this stage of the game, I’m hoarding resolution like a gamer hoarding strength points. The final piece will be resolved more coarsely, but for the moment, it’s like speed or gasoline: it is life itself.
I slide the transparency of the mod to 50% and stretch it to match the original. (I use the corners of the eyes as guides.) As I do so, I discover that the mod was stretched anamorphically (differently in x and y axes). The delta is slight — perhaps three or four pixels over several thousand, but noticeable when overlaying as I am. I finally get the two to match and resize the final piece to fit the frame (greedily retaining the extra bleed just in case).
At this point, I can de-couple the process from the modified image and work exclusively on the original, saving the mod only for reference. Next time, we inspect the image even more closely.
This post is part of a series of posts on the subject of books covers, directly primarily at self-published and independently-published authors seeking to design their own covers. It is in the category “Covers” and can be seen with all of the other posts in that category by clicking on the link in the right sidebar. The sophistication of that display page will improve as time goes on. The next post in the series will appear here within a couple of days.