The Girl from Ipanema

An AI generated image of a tall and tan and young and lovely woman walking the streets of Ipanema, Brazil.

It’s sort of crazy how quickly I’ve become addicted to using Easy Diffusion. Just a few days ago I found out about it and less than a week later I’m able to turn out images like this one.

Reading about the passing of Astrud Gilberto a few days ago, I used the first verse of her song The Girl from Ipanema as an image creation prompt in Easy Diffusion, and the result was great.

Tall and tan and young and lovely
The girl from Ipanema goes walking
And when she passes
Each one she passes goes “ah!”

Writing prompts for Easy Diffusion seems to be a bit of an art form. They’re either literary (or lyrical, in this case) or a bunch of keywords and measurements. The AI model also plays a part, so I went with the A-Zovya Photoreal checkpoint to give it a more photographic look. To get the desired image, I also added a couple of keywords to ensure the image was in black-and-white and that the girl would be solo. The decision to make it a black-and-white image was because my memories of watching old black-and-white videos of Astrud singing the song on TV, such as the one below:

There’s sort of a melancholy wistfulness that comes from watching old videos or looking at old photos and I wanted to capture that with the image above. I also wanted her to be walking on her own since that was how I imagined her, even before I knew all the lyrics.

As I’ve gotten deeper into the world of AI image generation, I found out there’s a sort of knockoff of OnlyFans, Fansly, and ModelCentro for AI artists called FansLora, I setup an account and posted this image. This is the only one so far, but I plan to add more in the future, though they’ll probably be a mix of SFW and NSFW.


  1. That came out nice. I’d be interested in seeing what you come up with. Do you have a decent graphics card? After my initial try I haven’t had much luck getting close to what I want

    1. It seems hit or miss. There is definitely an art to writing a good prompt, and from what I understand, it often takes multiple iterations for creators to come up with what they want. I have a decent graphics card; an XFX Radeon RX 580 GTS purchased three years ago. Not top-of-the-line, but not bottom-shelf either. Radeon is an AMD GPU line, and I think with AI and ML it might be better to use an Nvidia line, but the Radeon is working for me so far, so I’ll stick with it for the time being.

Leave a Reply

Your email address will not be published. Required fields are marked *