The Monster’s NOT Under the Bed image was created using NightCafe AI Image Generator. IMHO NightCafe Studio is one of the best AI image generators for newbies.

The Monsters Not Under The Bed

The Monsters Not Under The Bed

The Monster’s NOT Under the Bed AI image is an interesting prompt. The first part is like a story:

“The little boy and girl feels afraid and covers themselves completely with the blankets because there is a huge monster under the bed.”

It’s interesting because if the AI had created exactly what was asked for, there would be no visible children. At best it would show a couple of lumps under the blankets and a monster peeking out from under the bed. That would have been an interesting result, but what we see is much better. You’ll also note the children are sort of under the bed rather than in the bed under the covers.

I guess a way of interpreting this is the AI is trying to deliver at least two children (a boy and girl), the children hidden under the covers, a huge monster and something happening under the bed. It’s all there, but not exactly what the prompt asked for.

Note: Stable Diffusion XL 1.0 appears to have major problems with numbers. Ask for the 3 brides of Dracula and receive 4 clones! Ask for 3 little pigs, get anywhere from 1 to a whole litter of pigs!

My wife and I have run this prompt and derivatives multiple times and the monster is never under the bed and children are always visible. Sometimes it’s just scared kids covered by blankets with no monster, most times the bedding has morphed into a monster that’s eating children.

Example with minor changes to the prompt: “The little boy feels afraid and covers himself completely with the blankets because there is a huge monster under his bed.”

Monster Under The Bed

Monster Under The Bed

The AI generated image above is much closer to the intention of the prompt, but isn’t as interesting. It shows how the AI can’t yet interpret what we humans really mean, but can still generate interesting art.

The second part of the prompt which is broken into two prompts is:

“Brush strokes with volume and transparency, alcohol ink and oil paint.”

and

“Hyperrealistic, splash art, concept art, mid shot, intricately detailed, color depth, dramatic, 2/3 face angle, side light, colorful background”

The final part of this is a preset NightCafe prompt called Hyperreal which is generally used for shots of faces. Added to the “Brush strokes with volume and transparency, alcohol ink and oil paint.” makes for an interesting creation.

AI Image Creation Settings

Stable Diffusion XL 1.0 (SDXL 1.0).
NightCafe Creator model RealVisXL v3: An SDXL model for photorealism. Great at faces. By SG_161222.

AI Image Text Prompts

Primary Text Prompt
“The little boy and girl feels afraid and covers themselves completely with the blankets because there is a huge monster under the bed. Brush strokes with volume and transparency, alcohol ink and oil paint.”
Weight: 1

Secondary Text Prompt
“Hyperrealistic, splash art, concept art, mid shot, intricately detailed, color depth, dramatic, 2/3 face angle, side light, colorful background”
Weight: 0.9

Note: I didn’t use a negative prompt because I wasn’t expecting to see a lot of hands and feet or other things Stable Diffusion tends to have problems with.

Other AI Prompt Settings

Initial Resolution: Medium
Runtime: Short
Seed: 2850387242 the seed number adds randomness to the AI image creation process.
Overall Prompt Weight: 60%
Refiner Weight: 5%
Sampling method: K_DPMPP_2M