What makes a great portrait? What methods have artists used to capture the essence of a person, their character, and their emotions?
#Portraiture as Storytelling - Capturing personality
jovial merlin
Prompt: “Rubens Peale with a geranium. 1851.”
The prompt takes inspiration from an actual painting, 'Rubens Peale with a Geranium, 1801,' but nudges the date half a century ahead. The 17-year-old original sitter, Rubens Peale, was a botanist, and portraits of his era commonly used symbolic touches to hint at the subject's personality. The color scheme emphasizes the flower, making it pop against the solid, earthy tones of the portrait. The red of the flower is mirrored in the red of the young man’s lips. And yes, I’m pretty sure this painting is in Dall-E's training data after some initial testing. Made with v2-exp model.
unborn wadi
📝 Prompting Tip
When you're prompting for a portrait, you will always get better results when you are more specific about what you're envisioning. Instead of "a portrait of a beautiful woman" (image 1), try defining as many physical and/or emotional characteristics as you can.
If you sometimes enjoy the randomness of generating vague prompts, like I do, you can avoid specific characteristics by only focusing on emotion and action to describe facial features: "a portrait of a young woman, loose hair, friendly eyes, smiling at the camera" (image 2)
However, I find that simple descriptions of the physical characteristics will yield better results: "a portrait of a young woman with freckles, curly red hair and warm brown eyes, smiling for the camera" (images 3 & 4)
Generated with DALL·E 2
jovial merlin
Editorial photography often needs to “sell” the identity of a portrait subject by including background or context elements. This is particularly true for most stock photos. This example shows how the overall composition of an editorial portrait may need to feature a critical prop or attire. Here, the prompt does not specify the gender or race of the “scientist”, which is an interesting way to test for bias in training data. The cue “dof” stands for “depth of field”, a trick to make the image seem more realistic, with focus on the main subject. Here are variations on our generic scientist varying the keywords "senior scientist" and "young scientist".
Prompt: “serene portrait photo of a senior scientist. lab coat. microscope nearby. dof.” Can you spot any bias issues?
unborn wadi
Portraiture tries to tell an immediate visual story about the subjects -- what is their mood, what is their social status, how do they live, how do they use their time. To capture these elements in AI-generated portraits, you need to intentionally incorporate them into your prompts.
Instead of a generic "portrait of a man" (image 1), try to envision what your story is, what is the context of this portrait. For instance, "a photographic portrait of an elderly man using a walking frame (walker) in his garden, an overgrown greenhouse in the background" (image 2). This prompt not only describes the man's appearance but also gives a glimpse into his life. (It's struggling with the walker, but I'm going to stick with it.)
Of course, we have to add descriptions of the man's face if we want it to generate properly, so I'm going to enhance the prompt like this: "a friendly elderly man with happy eyes and a white beard, using a walking frame (walker) in his garden, knitted sweater, an overgrown greenhouse in the background, award-winning portrait photography" (images 3 & 4).
Generated with Bing Image Generator
sudden locust
Portraits don't necessarily have to be realistic to capture a person's "essence." The use of seemingly unconventional/unusual colors or styles can sometimes be even more effective at accomplishing this task and conveying complex emotions. (created with v2.Exp)
🟠
Portrait of a woman with bright orange hair. In the style of an ink wash painting. Ink splatters, white space.
🔵Portrait of a man with short spiked blue vibrant hair. In the style of an ink wash painting. White space.
jovial merlin
In some cultures, portraits of historical figures were not always based mainly on representation of accurate physical features, but on important events or roles they held in society. This was a way of transmitting historical culture. In portraits of this type, overall composition was critical to suggest a historical event, showing the importance of the figure through composition in a mythic context. Here is an example based on a famous real work. Here, Emperor Sun Quan is emphasized by his size and placement among a group of scholars that he is directing. Prompt: "Emperor Sun Quan in the Thirteen Emperors Scroll and Northern Qi Scholars Collating Classic Texts, by Yan Liben (c. 600–673 AD), Chinese". Made with the Dall-E 2.0 base model.
unborn wadi
Generating portraits with DALL·E 2.0 can be tough, but a great workaround is to generate your face as a close-up portrait, and then use the Editor (Outpainting) to create the rest of the image separately.
Step 1. Generate a face! If you already know your subject, add any elements that should be visible on the face: specific lighting, a mood, medium, effects, etc. I chose this prompt: 📝"a photographic portrait of a young woman with cropped black hair and closed eyes, contented, morning sunlight through windows" (image 1)
Step 2. Open the editor (https://labs.openai.com/editor) and import (U) the generated face. Before clicking on the checkmark to confirm, resize the face to match the image you're trying to generate. Erase any background or elements that you don't want, and use the smallest eraser size to 'scallop' the edges so there are no straight lines that might remain visible in the final image. Don't forget to erase the watermark! (image 2)
Step 3. Create a prompt that describes the entire final image as a whole. I want my subject to be a relaxed woman sitting in a chair with a mug of coffee, so I'm going to describe her like this: 📝"a full-length portrait of a young woman wearing pajamas, sitting in a comfortable armchair, cup of coffee, morning sunlight streams through the window, cosy living room, award-winning photography" (image 3)
If you don't have a specific image in mind, you can use a generic prompt for more random results. I tried outpainting with this: 📝"a full-length portrait of a young woman, award-winning photography" (image 4)
I got a similar image because the eyes being closed suggests a relaxed body, and the curtains, light colour, and short hair were all present in the starting image.
Because we don't need to specify any facial details in the outpainting prompt, we can be more specific about clothing and setting and mood.
Images generated with DALL·E 2.0
snow arch
Sometimes outputs can improve when a prompt is simplified. For example, here are some examples of "Portrait of a (middle-aged, mature Caucasian, African-American, Muslim, etc.) woman with a kind expression." The model is able to isolate and hone in detail on just one concept: a face.
Add in many other details and things get to be a bit more convoluted. For example, attached is "Portrait of a tall woman with a kind expression strolling through the woods on an overcast day in England. It is raining softly and she is carrying a red umbrella."
The model now has to guess which areas of the portrait are the most critical, and in this case, the face appears to lose importance while the woods are rendered quite well.
You can use this knowledge to your advantage by rendering out different areas of an image separately and stitching them together in the end with inpainting /outpainting. In this case, I might work on the face separately as well as the umbrella, while ensuring the lighting still matches the final scene (overcast, rainy day) as I work.
DALL•E v2 Experimental for all
sour wren
To control the distance from the object you can use the photography terms with whatever art medium:
- extreme close up
- close up
- medium close up
- medium shot
Promt: "[extreme close up portrait], by William-Adolphe Bouguereau and frazetta, a young elfen woman, an arcane mage, epic fantasy, professional watercolor, intricate, look at detail, beautiful, highly detailed"
Made with DALL-E 2.exp but works with any DALL-E version.
jovial merlin
Portraiture as Storytelling — Capturing personality
A great portrait does more than record a face. It should say something about who the person is. The best portraits make us feel that we are meeting someone. This kind of portraiture tries to suggest a person's character, mood, or social identity. In art history, this approach is sometimes called psychological portraiture or character portraiture. Personality refers to the inner qualities that give a person their distinctive character, apart from their physical appearance.
Artists use visual clues to help tell that story. These clues are called attributes (objects that help identify a person). A botanist might appear with a rare plant. A musician might be shown with an instrument. A writer might sit in a room lined with books or papers. Clothing, background, pose, and lighting can all suggest something about personality. These details are part of the portrait’s iconography (the visual symbols and meanings carried by an image). They help turn a likeness into a story about a life.
The sample image is inspired by a real portrait, “Rubens Peale with a Geranium” (1801), by Rembrandt Peale. Rubens was 17 at the time, and Rembrandt was 23. The geranium may have been the first specimen grown in the United States, making the plant almost a co-subject of the portrait. Rembrandt echoed the plant's rising stalks in the V-shape of Rubens' lapels, and the curves of the branches in the curls of his hair. The painting shows Rubens both wearing and holding a pair of glasses — an unusual choice for portraiture of the era, and a direct reference to his poor eyesight.
Alt text: A 19th century academic American style oil painting of a thoughtful seventeen-year-old botanist seated at a table, holding a large terracotta pot with a flowering geranium. The young man has softly curled dark hair, round wire-rim glasses, and a calm, introspective expression that suggests intelligence, sensitivity, and patient attention. He wears a brown coat over a high white collar, and the warm brown background and careful modeling of light give the portrait a formal studio quality typical of academic oil painting. An open botanical book, a magnifying glass, and extra spectacles on the table reinforce his identity as a serious young student of plants, while his gentle touch on the pot and steady gaze convey a quiet, observant personality.
Portraiture as Storytelling
Portraiture as Storytelling - Capturing personality
fluid rain
** Portraiture Isn’t One Style**
Portraiture is often thought of as simply painting or photographing a face, but artists have approached portraits in very different ways depending on what they wanted the image to communicate.
Some portraits aim for dignity and timeless presence. Others define a subject through their surroundings. Others use stylization to build a strong sense of identity without relying on realism. Thinking in terms of portrait approaches can open up much more variety than defaulting to a single “serious face” formula.
A classical portrait usually emphasizes formality, balance, and presence. Controlled lighting, composed pose, restrained gesture, and carefully chosen clothing or props help the subject feel dignified, authoritative, or enduring.
An environmental portrait uses setting as part of the portrait itself. A workshop, kitchen, greenhouse, stage, study, or street can reveal just as much about the subject as expression or pose. In this kind of portrait, place becomes part of identity.
A stylized portrait is less concerned with realism and more interested in interpretation. Exaggerated shapes, limited palettes, fantasy costume, graphic design influences, or unusual textures can all create a vivid sense of character. The subject may not look realistic, but they still feel recognizable as an individual presence.
Sample Prompt 1 — Classical Portraiture
A formal oil portrait of an elderly woman seated in three-quarter view, dark velvet dress, subtle gold jewelry, hand resting on a carved wooden chair, warm chiaroscuro lighting, muted brown background, refined brushwork, composed expression, old master portrait painting, timeless dignity
Sample Prompt 2 — Environmental Portraiture
A portrait of a florist standing in a narrow flower shop at dawn, apron dusted with pollen, buckets of roses and ranunculus surrounding her, soft window light mixing with warm interior glow, cinematic realism, intimate portrait shaped by place, natural posture, quiet pride
Sample Prompt 3 — Stylized Portraiture
A stylized portrait of a young man in a deep crimson coat with geometric gold embroidery, simplified facial planes, limited palette of cream, black, and red, decorative background patterns, bold silhouette, painterly graphic design influences, striking character-focused composition
stoic jacinth
Capturing Personality (Street Photography)
Expression reveals personality
Look for subtle emotions—eyes, tension, small gestures. That’s where the real story lives.
Environment adds context
The background should say something about the person, not just fill space.
Light creates mood
Use shadows or soft light to shape how the subject feels, not just how they look.
Connection matters
Eye contact invites the viewer in; looking away creates mystery or distance.
Embrace imperfection
Blur, grain, or awkward framing can make the moment feel more honest and alive.
A middle-aged man with weathered features, salt-and-pepper stubble, and tired eyes stands beneath a black umbrella, waiting at a rainy bus stop in the fading twilight, his gaze fixed on the wet street ahead. Dressed in a dark green jacket and wrinkled shirt, the shallow-focus image captures the moody, rain-soaked cityscape, where raindrops blur the lights and background details.
lost tusk
The ‘mood’ factor in portraits.
Much of the personality in portraiture comes from the mood we attribute to the image. In the following example, I composed a neutral prompt of a 35-year-old woman. Here was the original prompt. a 35-year-old redheaded woman, medium-length natural hair, vivid green eyes, statuesque nose, balanced facial symmetry, neutral expression, relaxed posture, facing forward, minimal makeup, natural skin texture, soft diffused lighting, clean soft-focus background, no emotional bias, observational realism, cinematic clarity, sharp focus, subtle depth of field, lifelike detail.
The images read 1-4, top to bottom, left to right.
The initial neutral image became image 1. I then only asked GPT 5.4, in the same chat, to change the mood. To moderate worry (image 2), excitement and anticipation (image 3) and anger bubbling out from under the surface (image 4).
When mood shifts, five levers reshape the portrait: micro-expressions in brows, eyes, and mouth; eye behavior including focus and moisture; facial tension patterns across jaw and cheeks; lighting and color temperature guiding emotional tone; and posture or head angle, subtly altering presence. Together, these changes convert the same face into entirely different emotional narratives. And it’s as simple as telling the engine about the personality you’d like to convey in the image.
fluid rain
**Animal Portraiture
**
Animal portraiture treats an animal as an individual subject rather than a generic example of a species. Like human portraiture, it focuses on presence, character, expression, and identity. A strong animal portrait is not just about showing fur, feathers, or anatomy clearly — it is about making the viewer feel they are encountering a distinct being.
That is what separates animal portraiture from wildlife imagery. Wildlife art often emphasizes habitat, movement, behavior, or the relationship between an animal and its environment. Animal portraiture shifts the focus to the who of the creature rather than the broader context around it. Close framing, controlled lighting, and attention to the eyes, posture, and facial structure can make an image feel intimate, dignified, or monumental.
Prompt: A hyper-realistic portrait of a majestic male lion against a deep black background, close framing, direct gaze, richly detailed mane, subtle rim light outlining the fur, dramatic low-key studio lighting, sharp focus on the eyes, dark negative space, powerful and dignified presence, fine fur texture, museum-quality animal portrait photography
spark bison
Dignity of the Ordinary: Anders Zorn vs. Walker Evans
Anders Zorn is often celebrated for his dazzling portraits of society figures, but his rural portrait paintings reveal another side of his art: quieter, tougher, and more deeply rooted in lived experience. In these works, Zorn turns his attention to working people, especially in the Swedish countryside, and paints them with extraordinary dignity. He does not sentimentalize rural life or turn it into charming folklore. Instead, he observes it closely.
What makes these paintings so striking is their balance of immediacy and restraint. Zorn’s figures feel alive because their personalities emerge through posture, gesture, and presence rather than dramatic storytelling. A farmer may lean forward with physical readiness; a woman may sit with calm endurance and self-command. These are not decorative “types,” but individuals shaped by labor, weather, and habit.
His brushwork plays a major role in this effect. Zorn often uses broad, economical strokes, allowing faces and hands to carry the greatest focus while clothing and background dissolve into looser passages. Light also matters: warm illumination across flesh and fabric is often set against deep, cool shadow, giving the paintings strong tonal unity.
In Zorn’s rural portraits, everyday people are neither idealized nor diminished. They are simply seen clearly, and that clarity gives the work its lasting power.
(Image was generated following a long discussion about Zorn's rural portraiture. I provided the personality descriotion of a rural family and GPT5.4 Thinking did the rest to prepare a production-ready prompt. )
Walker Evans worked in very different times and mediums than Zorn, but they share a striking quality: both could portray ordinary people with dignity without slipping into sentimentality. Zorn did it through paint; Evans did it through the camera. In both cases, the result can feel plainspoken, observant, and psychologically exact.
Zorn’s portraits are warmer, richer, and more sensuous. Even when he painted working people or rural subjects, he brought to them luminous flesh tones, atmospheric light, and fluid brushwork. His figures often feel alive in a physical, bodily way, as if they have just shifted their weight or turned toward the viewer. The paint itself carries warmth, air, and human presence.
Evans, by contrast, is cooler and more austere. His rural portraits, especially those associated with Depression-era documentary photography, strip away theatricality and visual embellishment. He relies on frontal clarity, stillness, and descriptive precision. Clothing, posture, facial expression, and setting quietly reveal class, labor, hardship, and self-possession. The emotional force comes from restraint.
The difference, then, is not simply painting versus photography. It is also a difference in register. Zorn gives us human presence through the sensual power of paint, while Evans offers a harder, more forensic clarity. Yet both refuse prettification. Both show that unsentimental observation can be one of the deepest forms of respect.
(I discussed the difference between Zorn and Evans, then, asked to generate a prompt for Evans-inspired portrait using the same subjects.)
leaden lodge
Portraiture as Storytelling — Capturing Personality
Core Idea
Portrait = Subject + Context + Emotion + Intent
If one is missing → the story thins out.
If all four align → the portrait speaks without permission.
🧩 The Story Equation
- Subject — Who is this?
Not just a person. A specific version of them.
- “a student” ❌
- “a student running on 3 hours of sleep before finals” ✅
👉 Specificity = instant depth
- Context — Where are they, really?
Environment = receipts.
- physical space (room, street, nature)
- time of day (morning calm vs 2am spiral)
- surrounding objects
👉 Context = lived reality
- Emotion — What’s leaking out?
Avoid obvious emotions. Go for layered ones.
- not “happy” → “relieved but unsure”
- not “sad” → “quiet disappointment”
👉 Emotion = entry point for the viewer
- Intent — Why this moment?
This is the director brain.
- what just happened?
- what’s about to happen?
👉 Intent = the invisible narrative thread
🎬 3 Portrait Lenses (pick your weapon)
-
Environmental
Story lives in the space
→ grounded, believable, observational -
Conceptual
Story lives in symbolism
→ abstract, internal, metaphor-heavy -
Documentary
Story lives in timing
→ raw, unposed, “caught in the act”
⚠️ Common Misses
- “person looking at camera” → no story
- overly generic emotion → no hook
- clean empty background with no reason → lost context
- no implied moment → static image
Final Anchor
A strong storytelling portrait makes the viewer think:
“I walked in at the exact wrong (or right) time.”
Prompt: "Portrait of an elderly man in a cluttered workshop, surrounded by tools and half-built objects, showing calm focus and nostalgia, at the moment he pauses to remember something. Natural light through window, environmental portrait style."
