Rave at a dentist office

- I want to create an image where the mirror ball is spinning, there are neon signs, and the doctor and patient are enjoying a rave atmosphere.
- Draw a sketch of what you would like the final image to look like.

3. I choose the CANVA for making Ai image

4. Added my prompts to generate image #1

I needed to understand the program, so I obtained images through a simple keyword search, but I was not satisfied.
5. Image #1

6. Refine and tweak my prompt for Version #2

7. Version #2


8. Last chance to refine and tweak my prompt for #3

9. Finally I got these images

I tried to find something that looked close to my prompt. Then I got the image and I added a title.

This is my first time using an AI image program.
It was fun because I got many kinds of images,
but I think it is hard to get exactly what I want.
I thought that the more specific the description, the more accurate the results would be, but when I saw the results, I felt like the image was created with only partial data. Also, as Korean is my native language, I have limitations in expressing my thoughts in English, so it took several attempts to get the result according to the image I had in my head.