Anatomy of a book cover

Now that my release is out, I thought I’d do a fun post to help out fellow indies looking to do their own book cover work. Here’s a rough breakdown of everything I did to put together my most recent cover and a general process you can use yourself.

Know your limits

I am not an artist, and I am not a designer. But I have at least some eye for design and for typography. Given those abilities, I can put together a cover on my own—sometimes. But it’s important to know when your vision exceeds your grasp, or if you lack a vision and need a professional to help.

For the cover of Below, I knew I could never come up with artwork good enough for the kind of cover the book demanded, so I sought out an artist on DeviantArt. I posted on their jobs board with info about my target dimensions, the general kind of image feel I wanted, target budget range, and a brief description of the  book. Also, I mentioned that I could handle the typography on my own, and only needed the image itself. A lot of specificity when hiring an artist, and knowing their terms and process up front, goes a long way.

I wish I had hired an artist for my first book cover, and someday I may redo it.

But for The Affix and The Well of Moments, those fell in my wheelhouse.

Have the right tools

When I work on a cover I need a good photo-quality image editor, and I also need a good vector art program.

My go-to photo editor is Photoshop Elements, which has a lot of power for the price. It can’t do all that full Photoshop can, or more specifically it’s been nerfed but a lot of that is available under the hood, but what it can do is typically more than enough. Some people swear by the GIMP; I used it many years ago and despised it, but people keep trying to convince me it sucks a whole lot less these days; however its lack of ability to do non-destructive layer filtering makes it a bad choice for these purposes anyway, so do yourself a favor and buy Photoshop Elements. Anyway, the important features you need in such an editor are the ability to work with layers, and save the document in a form that keeps separate layers intact.

For vector art, which includes the text for titles, the back cover, spine, etc., I use Inkscape. It can do text, shapes, and all kinds of things.

Also I have occasionally used POV-Ray, which  is a ray-tracing engine to produce photorealistic images. The cover art of The Affix was produced in POV-Ray. It was not quick; the program had to handle things like internal reflections, refraction, lighting, etc. The full-size image took about a day to render the first time, and slightly less when I redid it a few years later on a newer computer.


Before you ever work on your cover, you need to have an idea in mind of how big you want the paperback to be. Even if you don’t plan to release a paperback edition right away, trust me, you’re way better off planning ahead! If you create a full wraparound cover, you can always trim it to front-only when you’re ready to assemble the ebook.

One Woke Up and Below are both 6″×9″ in paperback. The Paranormal Curio series goes for the smaller 5″×8″ size, with matte-finish covers instead of glossy. If you’re not sure which to go with, 6″×9″ is probably the better choice.

All your artwork should be at least 300 DPI, so 300 pixels to every inch. That’s the standard size for most printing and you’ll get a lot of detail. Yes, some printers can do even higher resolutions, but you should be good with 300 and I don’t think there are print-on-demand publishers who will take anything higher anyway.

So your main artwork file should be big enough to include the front cover, the back cover, the spine, and bleed space around the edges. I always work with ¼” bleed which is 75 pixels in every direction, but CreateSpace wants half that so I always trim it down to 38 pixels of bleed before uploading to them. More bleed can’t hurt you, so stick with 75 pixels.

Your dimensions for the working image file are:

Width: book width × 2 + spine width + bleed × 2
Height: book height × 2 + bleed × 2

What you probably don’t know at this point is the spine width. Unless your book interior for paperback format has been totally finalized, you don’t know how big the spine will get. And even if you do have it finalized, there’s value in adding some slack space for future editions. Unless you’re writing a behemoth 200K-word epic, you probably won’t go anywhere near two inches, so that’s a good upper limit. You could do 1½ if you think that’s good enough; for The Well of Moments, I knew I’d never get to 1½ so that’s what I used, and calculated my spine at 450 pixels. So for that book, the dimensions were:

Width: (5″ × 2 + 1.5″ + 0.25″×2) × 300dpi = 3600 pixels
Height: (8″ × 2 + 0.25″×2) × 300dpi = 2550 pixels

The front cover portion of the image was set 75 pixels from the top bottom, and right edges of the image and was 1500 pixels wide.

So now you have the basic dimensions setup. Be mindful of the following

  • Your cover text needs to make use of some margin, even if it’s a small one, to look nice, and ideally most or all of the images will be kept within a margin as well. The bleed area doesn’t count; you need a margin from where the cover meets the bleed.
  • If you’re using a 6″×9″ size, remember that Kindle prefers a 5:8 ratio which is slightly narrower, so give yourself just a smidge of extra margin on the sides so you can crop the ebook cover.


If you’re doing your own art and not working from an illustration or entirely from stock photos, you’ll need a background for your document. A flat color is likely to be boring, so consider something textured or with a gradient. But beware: gradients tend to produce banding. (If you have a gradient with banding, create a new medium-gray layer, set your foreground/background colors in the editor to black and white, and then create 2% monochrome Gaussian noise. Change this layer’s blend mode to screen or overlay. This interferes with the banding pattern and makes the gradient look better.)

For The Well of Moments, I wanted a stone texture like the Well itself. I made a brown background, then created a new document of the same size. I rotated the document by a certain angle to enlarge it, then used the Render | Fibers filter to create long striations. After that I rotated back, cropped back to the original size, and copied that image into my original cover image as a new layer. Changing the blend mode to overlay and lowering the layer’s opacity made it look like subtle color variations in the brown rock.

After that I also created some noise layers and used the Crystallize filter to make little flecks. I applied a new adjustment layer to limit how many flecks I got based on how bright the color was. Then a few more adjustment layers changed the color of those flecks to what I wanted. I did this a couple of times for different kinds of flecks. I’m going sparse on detail here because it’s really not the point of this tutorial.

Suffice it to say I created a nice brown stone background for the entire document.

A word on fonts

Don’t use any old font on the cover, and DO NOT USE TIMES NEW ROMAN. Find a font that suits your genre and use it. You can use multiple fonts, but don’t use more than two fonts on the cover; it never works.

Also, make sure you don’t just assume any font you download is usable. You can find free-for-commercial-use fonts at There are other font sites online too, but I don’t trust them as well; they may serve up fonts that have more restrictive licensing or even have misleading/incorrect info about that.

Titles and text

This is where Inkscape comes in. You’ll want to create a document for your front cover, one for your back cover, and later one for the spine. You may not know the exact spine width but you may know a minimum of what it will be; if not, you can always do the spine last.

For both covers, front and back, you’ll need the document to be the size of just one cover for your book, and since it won’t be extending into the bleed area (if it does you’ll need to deviate from this tutorial) you do not need to include any bleed in your size calculations. Using The Well of Moments as an example, that’s a 5″×8″ book, so you should tell Inkscape to use pixels: 1500×2400. Yes you can set a document size in inches or whatnot, but you really want pixels for this.

For front-cover titles I created some text in Inkscape and moved it into position using the alignment and sizing tools. Special effects like bevels, and a black blur around the letters to outline them, are done later in Photoshop Elements. For now the main thing is just the shape and positioning of the text.

The back cover gets special treatment: I take the blurb I’ve written and make a text object, and I also make a transparent rectangle that fits the box I want the text to fit into. Then I select both, and flow the text into the rectangle. Turn on full justification in the text so it lines up with the left and right sides of the box. I also write a little header/tagline and put it above that box, and may size it to match the box’s width.

These documents can be exported to PNG format, and imported into your wraparound cover document via Photoshop. Now the PNG export will say it’s 90 DPI, not 300, but what the image says and what it’s actually used for are two different things. If you were to export at 300 DPI it’d just blow your image up to a bigger pixel size than you actually want.

Usually for the spine, the image I create will have a width of the book’s height (2400 pixels in this case) and a height of about 0.75″, or 225 pixels; maybe 250 pixels for an even number, which is easier to work with. Hopefully your book will be more than half an inch (150 pixels) in paperback form, but  less than an inch is common. I take my title and author text from the front-cover document and rearrange them here as smaller copies, usually with the title a couple inches from the top and my name a couple from the bottom. Ultimately, the image exported from this file will be rotated 90° clockwise and placed in the wraparound cover document.

Also there’s a little matter of outlines. Outlines can make your text pop. I’ll take my text and give it all a stroke width of something like 5 or 10 pixels, then make the text and stroke black. Exporting these to PNG gives me something to use as an outline layer behind the text, and once it’s in my main document I’ll typically apply a blur to the outline layer.

Positioning front-cover text and images

To bring your front cover text into the full image, this is how you do it in Photoshop Elements.

  1. Open the PNG you exported. In this example, it’s 5″×8″ so it’s a 1500×2400 image.
  2. Resize the canvas. You’ll see a 3×3 box that shows how the original image will be positioned in the enlarged canvas: click the left middle box, so any extra width goes to the right and extra height is shared on both sides. Add 75 pixels to width for the bleed on just one side, and 150 pixels to the height for bleed on top and bottom combined.
  3. Resize the canvas again. Click the right middle part of the 3×3 box so extra space goes to the left. Now choose the exact dimensions of your full image.
  4. Duplicate the layer(s) of this image into the full-size document.
  5. Close the image you just resized without saving.

You do something similar for the back cover, but that’s usually just text and you don’t have to worry about centering images, etc. there. Usually, you’ll work on back-cover text and spine text/art in a separate file, and only apply them to the full wraparound cover and position them once the exact spine width is known. More on that later.

Other front-cover items

Very commonly, I’ll use Inkscape for other design elements on the cover. Just like with the cover text, I create a document of the one-cover-only size, and export it to PNG. For The Well of Moments this took a couple of forms. First, I wanted to create the runes. For that I had to download a font that included the runes, and then I had to get the text to flow along a circular path—which also ultimately involved having me extend the circle to a spiral path so I could fill it in properly. The runes were imported into the main document in Photoshop Elements, positioned just like I did with the text, and then various effects and layer styles were applied to make them look inset.

Also in Inkscape, I needed a shape template for the opening of the Well, and for the various inner circles within it. Each of those were exported as new documents. Another, larger circle was done for use as a “spotlight”. I calculated the circle sizes based on the size I had used for the runes, and the smaller concentric circles were sized so that they got slightly more spaced out as they moved outward.

The outermost circle of the Well went through a lot. It was used to create the “lip” of the Well through various effects and filters and shading trickery, plus a black nothing behind the inset images if anything leaked through.

I won’t go into too much detail on how I did the ripple effect in the middle, but suffice it to say I built an image in Inkscape with gray circles and some alternating black-to-white gradient rings around them, brought that into Photoshop and used the original for the red channel and a rotated version for green, and then used that as a map for the Displace filter. It was tedious but I was happy with the final effect.

The photographic parts of The Well of Moments‘ cover are obviously stock photos. I purchased them from Finding the right model was the hardest part, and I had to separate her from the background she was up against—even though it was mostly off-white—which involved challenging selection brush work; that’s a subject for somebody else’s Photoshop tutorial. I used the concentric circles I had made as clipping masks for these images, and then combined them into a single layer so I could apply that ripple.

The ankh and its chain were Inkscape work too, at least their shapes, but it’s a little too complex to explain my process for those. I actually used a partially finished mockup cover in the background to help me create the chain with the right positioning.

Finally, I mentioned that I created a spotlight circle. This was a white circle larger than the Well and the runes. Once I had it positioned, I applied a powerful blur to it and merged it on top of a solid light gray layer, then applied a layer style that used the gray to darken everything beneath it. All of the text layers went above this spotlight layer.

Bringing it together

So you’ve finally formatted your paperback and you’re ready to assemble the rest. The spine and back cover text haven’t been added yet.

First, you need to calculate your final spine width. If you use CreateSpace and you’re going with black-and-white printing on cream paper, they calculate 0.0025″ per page, which is actually 0.005″ per leaf. So if you have an odd page count, round up to a multiple of 2. Then multiply that by 0.75 (0.0025″ × 300dpi) and round up if need be, and that’s your spine width in pixels.

Make a copy of the wraparound cover document you’ve been working with so far. However much space you allotted for the spine, you need to trim it now. If your spine had a maximum of 450 pixels and it actually comes in at 226, then you need to trim off 224 pixels from the left of the document. Open Resize Canvas, click the right-middle square, click the checkbox that says you’re using relative dimensions (i.e., adding or subtracting), and put in -224 for the width. Confirm that, and boom: you’ve cropped successfully.

Now let’s add the back cover text. Open your back cover PNG and follow the same instructions for positioning the front cover, but reverse the left and right. That’s all.

For the spine, open the spine document and rotate it 90° to the right. (Before rotation, you may need to apply special effects like bevels to the text if you want them.) Resize the canvas to the full wraparound size, leaving the 3×3 positioning box alone so this image will be in the middle of the new canvas. Duplicate this as a new layer into the wraparound.

If your back and spine text has outlines, you’ll want to follow the same procedure for the outline PNGs too.

NOW, your cover image is mostly finished—for this edition. Save it as a new document and give it a name that tells you this is a nearly finalized copy.

Now let’s make a PNG. Flatten the image. If you’re using CreateSpace, this 75 pixels of bleed we’ve been working with is too much, so you want to back that down to 38 pixels. Resize the canvas, relative mode, and choose a width and height of -74 pixels: -37 on every side. Once you’ve cropped the image, save as a PNG.

CreateSpace is stupid about not accepting PNG files as covers, so now you’ll need to make it a PDF. I use LibreOffice Draw for this. Create a new document in Draw, or a similar program, and change the page size to the exact size of your image—but in inches, so divide the pixels by 300. Get rid of any margins. Drag your PNG file into the document and position it at the edges. Now export to PDF, and you have a file ready for CreateSpace.


About Lummox JR

Aspiring to be a beloved supervillain
This entry was posted in Uncategorized and tagged , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s