Your first image in 15 minutes
You open Gemini or Google AI Studio, write a four layer brief (subject, place, light, camera) for a cup of coffee on an Alexandria balcony at sunset, and make a real 4:5 image yourself. Then you save the prompt that made it for the rest of the course.
Chapters
- 0:00What we are making today
- 1:15Opening the tool
- 3:10The four layer brief
- 7:00The first image
- 10:00Changing the light and comparing
- 12:50Save the prompt
Lesson notes
Goal: a real image, made by you, in 15 minutes
We are not starting with theory. In this lesson you make a real 4:5 image you could publish, and you understand why it came out the way it did. All you need is a Google account and a browser.
Open gemini.google.com or aistudio.google.com and choose image generation. Both use Google's Nano Banana.
A small cup of Arabic coffee on a round metal table on an old Alexandria balcony, the Mediterranean Sea in the background. Golden hour sunlight coming from the left, warm and calm, long soft shadows. Close up shot from slightly above, shallow depth of field, the sea softly blurred. Vertical 4:5 photograph.Copy the prompt as it is the first time, then try changing just one thing.
Steps
0 of 5 doneThe mistake
Writing "a nice cinematic coffee picture" and expecting the model to know what you mean.
The fix
The model fills every vague word with its own decision. Say what it is, where it is, where the light comes from and where the camera is. Four decisions change the result.
| Step | Credits |
|---|---|
| First generationnano-banana | 1 |
| The second light testnano-banana | 1 |
| Margin for one extra attemptnano-banana | 1 |
| Total | 3 |
These are approximate generation counts, not prices. Free limits change, so check them in the tool itself.
Transcript
Welcome to the first lesson at U3 Academy.
We are not starting with theory. We are starting with a real image, made by you, in fifteen minutes.
The image we will make is a cup of coffee on an Alexandria balcony at sunset.
By the end you will understand why it looks the way it does, and how to change it.
Open your browser and go to gemini.google.com or aistudio.google.com.
Both use the same Google image engine, called Nano Banana.
Sign in with your Google account and choose to generate an image.
Now the most important part. We will not write "a nice picture". We will write four decisions.
Decision one, the subject: a small cup of coffee on a round metal table.
Decision two, the place: an old Alexandria balcony with the sea behind.
Decision three, the light: sunset sun from the left, warm and calm, long shadows.
Decision four, the camera: close, slightly above, the sea soft in the background, vertical frame.
The prompt is written below the video. Copy it exactly.
Paste the prompt and generate. The first result takes a few seconds.
Look at the image with a critical eye. Is the light coming from the left?
Is the sea soft behind the cup? Is the cup itself clear and sharp?
If the answer is yes, the model carried out your decisions. That is your first image.
Now a small experiment that will teach you more than any explanation.
Change only the words golden hour to blue hour after sunset.
Generate again and put the two images side by side.
One phrase changed the color, the mood and the time of day. That is what it means for a prompt to be decisions.
Last step: download the image and save the prompt in a file.
We will use it again at the end of the course when we turn this image into a video.
See you in the next lesson.