Hey guys, I’m releasing my second e-book of children’s stories. It features a meeting between the Big Bad Wolf and the Little Mermaid. This book is better overall, especially in terms of the plot has also been improved, trying to make everything as simple as possible and with a much clearer and more objective moral lesson and the vividness of the background colors (I’m getting the hang of Stable Diffusion and the other tools used to produce the books).
The link is here:
https://www.amazon.com/dp/B0C27GKXY5
All the books in this “saga” have the Big Bad Wolf as the antagonist. In traditional children’s stories, the Big Bad Wolf appears in only a few stories, but taking the hype from the second Puss in Boots movie, which introduced a tough wolf, I had the idea of creating new children’s stories where the wolf interacts with other well-known characters from children’s stories. However, I already have ideas for other children’s stories without the Big Bad Wolf.
I chose to focus on the Little Mermaid now (I was going to do it later, but I moved up my plans) to try to capitalize on the controversy that will arise from the live-action Disney Little Mermaid film and the Dreamworks’ Ruby Marinho movie, both coming out in 2023.
Now that I’ve got the hang of book production and have all the necessary skills, the tendency is for the books to come out faster and faster. The first one was a laboratory of experiments, and this second one confirmed that the production process is faster.
About a week ago, I released “The Big Bad Wolf and the Winnie the Pooh,” which can be found here:
English version with Winnie the Pooh (US restricted, because only in the US is Winnie The Pooh already in the public domain):
https://www.amazon.com/dp/B0C1L5JJBM
International english version with Duuh the Bear (with a different name and color than Winnie the Pooh, because in the rest of the world the Pooh is not in the public domain - but it’s the same story):
https://www.amazon.com/dp/B0C1MYZ8BW
The general workflow is the same as before:
-
I preconceived several children’s stories with characters that are already in the public domain and are known in children’s literature and media, I made a selection that resulted in the 8 most promising stories. I defined the amount of characters that I should have per story, in the vast majority of them there are 3 characters that I will need to model per story, a single story will need to model 5 characters. I’ll leave this as the last thing to do.
-
Then I refined and created a script for each of the stories and asked ChatGPT to help me with the script. Most of the scripts I made some small adjustments with ChatGPT suggestions, others not so much and ChatGPT gave me ideas to build better scripts. In the end, none of the stories were 100% the same as the ChatGPT suggestions, all the final stories were properly adjusted and assembled by me with such suggestions.
-
Then, I started the process of modeling, texturing, etc., the characters in Blender 3D. I’ve been working with Blender for over 10 years as a hobby, but I spent almost 10 years without practicing it frequently, despite the time, most of the basics I never forgot and I revised/relearned and learned new things over time. I’m just a beginner in 3D at the moment.
-
Based on the script, I defined the amount of images that the book would need to be illustrated and the frames/positions/actions needed to represent the book’s story well.
-
I then used Stable Diffusion on my computer (using my old but useful GTX 1060) and used a model for the image generation to illustrate the cartoon looking background. I generated 512 x 512 images by saving the best image data and then upscaling the images to the desired resolutions.
-
I used the characters created in Blender 3D and positioned them the way I had thought of the script part, rendering images with a transparent background and a shadow catcher (a transparent plane that retains shadows).
-
I mounted the two images (3D characters rendered image + Stable Diffusion scene images) in GIMP. In both Blender and Stable Diffusion, the generated images had high resolution to be placed in the book without pixelation problems.
-
Assembled images, I took it to Inkscape to place semi-transparent white square shapes with rounded ends + text to place the script texts. Then I generated the final images for each of the pages in the book.
-
I configured everything in Libreoffice to convert the ODT to PDF to be converted into a children’s book by the Amazon application ‘Kindle Kids’ Book Creator’. Only the digital version will be available at the moment.