Generating art on Midjourney, Dall-E, and other platforms may be just a fun hobby for you. But if you’re like me, you create this work for specific purposes. One of those purposes for me is to create artwork for music albums which, lately, I have also been producing through generative AI.

I’ll tell you a bit about my workflow so you can have fun with this, too.

A Collage

The album itself is “nature” themed. I worked in ChatGPT to create lyrics using themes like a storm, rain, the ocean, etc. as metaphors for life. So I wanted the cover to look like a collage of sorts – a mix of different pictures that created a mishmash of nature images. Below are the prompts and results.

Prompt: “collage of ocean, forest, desert, rain, and volcano, album cover” stylization 300

This was cool, but not quite what I wanted. The way these systems work is they pull from all the images of a particular category to create something. This is something you might want to think about as you try creating images, and this might explain why you don’t always get what you want. My guess is that Midjourney has a data set for the world “album” that includes a lot of old albums (totally speculative assumption). So I dropped that in subsequent prompts and switched it to this.

Prompt: “collage of ocean, forest, desert, rain, and volcano”

This was a little closer to what I was looking for. I eventually landed on this to be the artwork for my album cover:

I threw it into Canva, added some lettering and presto:

You can listen to the songs I generated on my official Bandcamp site. Hope you enjoy, and I hope this description of my workflow inspires you to make your own awesome art. 🙂