Underwater Portrait Photography

At a glance

Model

Midjourney

Use cases

RED covers, avatars, and personal homepage visuals

Audience

Social media operators, avatar creators, portrait photographers

Problem solved

Having a portrait theme but lacking a sufficiently premium cover composition

Subject

Character subject and layout structure

Core style

Cinematic

Aspect ratio

--ar 16:9

Difficulty

Medium, replacing clear variables first will make it more stable.

Variable replacement guide

Best use cases

A close-up portrait of a woman half-submerged in water, with shimmering light and goldfish swimming around, presenting a dreamy and ethereal aesthetic. It is highly suitable for scenarios requiring quick turnarounds like RED covers, avatars, and personal homepage visuals, targeted at social media operators, avatar creators, and portrait photographers. The core value is not stacking style words, but solving the problem of 'having a portrait theme but lacking a sufficiently premium cover composition': first lock in the subject and composition, then replace the variable information to allow the same template to serve multiple topics.

Visual style

The stylistic tone of 'Underwater Portrait Photography' is cinematic. When reading the image, prioritize the clear light and shadow, character posture, and framing, followed by the relationship between the main character, the layout structure, and the background; whitespace, hierarchy, and color only serve this main relationship. The final result is more like a finished product that can be directly used for RED covers, avatars, and personal homepage visuals, rather than a mere stylistic experiment.

How to generate a new image

1. 1Copy the entire 'Underwater Portrait Photography' prompt, preserving the descriptions of style, subject, composition, texture, light/shadow, and aspect ratio.
2. 2Prioritize changing the Portrait Subject, Aquatic Creature, and Aspect Ratio slots, leaving style and camera descriptions untouched for now.
3. 3Generate 2-4 images in Midjourney or your preferred image tool with default parameters, using --ar 16:9 as an aspect ratio reference.
4. 4Check against the target portrait visual: is the subject clear, is the information hierarchy readable, and are the colors and style close to the reference image?
5. 5If it deviates from expectations, reduce background noise first, then reinforce core style words or adjust the ratio without changing too many variables at once.

Prompt remix example

Portrait Subject: a beautiful young Chinese girl Aquatic Creature: goldfish Aspect Ratio: --ar 16:9

Portrait Subject: a boutique perfume bottle on a clean display table Aquatic Creature: goldfish Aspect Ratio: 4:5

After modifying this way, the image still retains the original style and composition logic of the portrait visual, but the subject, text, or frame changes to match the new publishing scenario, making it more suitable for directly generating another usable image.

A highly cinematic close-up shot, a boutique perfume bottle on a clean display table with her head half-submerged in water, face close-up. There are swimming goldfish in the water, clear water and bright light and shadow refraction on the girl's delicate face, aesthetic and ethereal, premium colors, highly artistic, high resolution, masterpiece. 4:5

Model and parameter tips

• Midjourney：Suitable for reinforcing stylistic atmosphere and composition reference; aspect ratio parameters can be added, but the stability of Chinese text generation may fluctuate.
• GPT-image-2：Excellent for Chinese understanding, text layout, and complex image relationships. When generating portrait visuals, we recommend directly writing the subject, aspect ratio, and intended use clearly.
• FLUX：Perfect for a clear subject, photographic textures, and graphic structures. Use natural language descriptions and minimize command-style parameters not recognized by the platform.

How to improve weak results

• If the subject in 'Underwater Portrait Photography' is not prominent enough, write 'main visual centered', 'clean silhouette', 'large areas of negative space', or directly name the core object of the portrait visual.
• If the style drifts, adding terms like 'core style', 'high contrast hierarchy', and 'clear texture' is more effective than just adding emotional descriptors.
• When the image is cluttered, strip away background elements, edge annotations, decorative symbols, and subtitles first, keeping only the most important layer of information.
• If the portrait looks like a cutout sticker: add realistic edges, natural shadows, skin details, and obstruction relationships between the subject and the foreground.
• If colors are incorrect, clearly write out the primary color, base color, brightness, and saturation, avoiding vague terms like 'premium feel' or 'atmospheric feel'.
• Adjust the aspect ratio to 1:1 / 4:5 / 16:9 / 9:16 if the platform's layout is incorrect, and reduce the density of fine print accordingly.