Outbreak – Cover Art

First, right to the book cover, explanations below:

Just a quick note that Book 7 of Raven of Iskandar came back from the editor’s first round and I did pretty good this time. Most of the corrections occurred late in the book and during the exciting sections. One of the ironies was that my Scottish editor corrected my spelling from British versions back to American versions. There is a slight chance I’m being infected.

I’m not an artist and I’m not thrilled with paying multiple hundreds of dollars for ‘professional’ cover generated from their vast library of image elements. However, I’ve just gone through a week playing with AI image generators. Note that the book cover is the combination of three different AI generated elements.

I initially started with Microsoft’s Copilot. Mainly because it was free.

Copilot uses Dall-e 3 to generate images but the results were not great. First, if you use a long, specific description, it will ignore some of your elements. I tried generating a Raven picture and try as I could, I had a hard time getting it to show full body and then it insisted on high heels rather than boots.

If you started with a basic description, it was difficult to refine the image. It would give you options, usually two or three, and most would be the opposite of the direction you wanted to go. Finally, you’d get one to two tries to refine by text, then it would stop responding.

However, then I found OpenAI. First, they are one of the only AI image generators that allow you to demo the program without demanding a credit card number. And the results were much more lifelike than Copilot and you had endless possibilities for refinement. Oh, yeah, they got my money. Currently a sale is going on and you get one year at $6.00 per month for 5000 credits. They charge you a single credit for the initial image and then 1 credit for each refinement. When you do a refinement, you a selectable number of tries, usually 4. So you can chew up 10 to 20 credits an image. That works for me as I’ll only be doing images for book covers and the Raven website.

There are higher tiers of service for more money per month.

Here are a couple of images I’m working on. Note that with OpenAI (and not Copilot), you can upload an image and use the AI and editing tools to modify it.

First, a Raven picture:

I’m still working on the transparent background. Of course, I’m wondering how it will go with trying to create an image of Skylla.

Here’s a second image with the idea of Central City:

So, what do you think? You can leave a comment.