What to Type for AI

Why Your AI Headshots Look Different Every Time—and How to Fix It

You've nailed the lighting setup in your studio. You know exactly how you want your client's headshot to look. But when you generate it in Midjourney, every single attempt looks slightly different—different skin tone, different eye direction, different vibe entirely.

This isn't a tool problem. It's a communication problem. You're describing the same headshot in different words each time, and AI responds to every tiny variation. The solution isn't to become a prompt engineer. It's to isolate the 6 micro-decisions that actually control consistency, lock them into reusable language, and use that language identically across every client session.

Headshot Prompt Anatomy gives you exactly those 6 variables—skin diffusion, eye catchlight, jaw angle, smile warmth, background blur, and crop ratio—plus a full spectrum of tested prompt phrases for each one. Once you pick your phrases, you paste the same ones. Same input. Same output. Every time.

The Real Reason Your Headshots Don't Match

Most photographers who generate AI headshots treat the prompt like a loose description. They write something different each time—tweaking adjectives, rewording lighting notes, changing how they describe the client's expression. Each tiny variation signals to the AI that something should change. After five generations, you've got five completely different looks, and you're not sure which one to show your client or how to recreate it next week. The professionals who get consistent results do the opposite. They write the same prompt structure every single time, changing only the variables they actually want to change. The 6 Micro-Decision Quick-Reference Card in Headshot Prompt Anatomy maps those exact six touch points, with 4–6 specific phrase options for each one. Pick your phrases once. Reuse them. Watch consistency appear immediately.

How the 6-Variable System Locks Your Style

Instead of describing a headshot in prose, think of it as six independent controls on a mixing console: **Skin Diffusion** controls whether the skin looks studio-lit and sculpted or soft and diffused. "Soft skin, subtle texture" vs. "sculpted cheekbones, sharp definition." Pick one. Use it every session. **Eye Catchlight** is the highlight in the eye—it signals where the light came from and defines how alive the face looks. "Double catchlights" vs. "single side catchlight" vs. "subtle catchlight." One phrase per client type. Done. **Jaw Angle** controls how the face is positioned relative to camera—affects whether the client looks strong, vulnerable, or neutral. "Slight over-the-shoulder angle" vs. "straight-on" vs. "chin slightly tilted down." Lock it. Repeat it. **Smile Warmth** ranges from neutral to genuine to practiced-corporate. "Neutral expression" vs. "warm genuine smile" vs. "professional polite smile." Same phrase, same result. **Background Blur** is your depth control—how out-of-focus the background appears. "Soft bokeh blur" vs. "slight blur" vs. "sharp clean background." Write it once per client, use it every time. **Crop Ratio** determines whether it's a tight face crop, a chest-up professional headshot, or a wider frame. "Tight face crop" vs. "standard headshot crop to mid-chest" vs. "environmental headshot." One per shoot type. Once you've written those six phrases, your prompt becomes a fill-in-the-blank formula. Same structure. Identical results. Every client. Every session.

Real Workflow: From Inconsistent to Locked in 2–3 Rounds

The 5-Iteration Workflow System in Headshot Prompt Anatomy shows exactly how to move from "let me try this" to "this is locked." Round 1: Generate five variations using your baseline prompt. Pick the one closest to your vision. Round 2: Edit only the 6 variables you want to change, keeping everything else identical. Generate five more. Round 3: Lock the language. Those exact phrases become your signature formula for this client type or shoot style. Most photographers get to "locked" in two rounds. Corporate neutral headshots lock fastest—the category has the tightest definition. LinkedIn vibrant takes slightly longer because you're balancing warmth with professionalism. Creative editorial can take three rounds because there's more room for interpretation. But once it's locked, generating a new client's headshot becomes copy-paste-generate. Thirty seconds of prompt work instead of thirty minutes of trial and error.

Translating Your Studio Setup Into Prompt Language

You know exactly how you light a headshot in your studio. Soft key light from 45 degrees, fill light at half power, rim light behind. But if you describe that in technical studio terms, the AI doesn't always translate it correctly. The Photographer-to-AI Translation Glossary in this product converts your real studio setup—your actual lighting rig, your camera, your posing directions, your color grading style—into the exact prompt phrases that produce results matching your on-camera work. Instead of writing "Hasselblad, Zeiss 80mm lens, 1/160th at f/5.6," you'd write "tight professional headshot with subtle depth of field." Instead of "catch light in both eyes from the key light," you'd write "double catchlights from 45-degree key light." These aren't random translations—they're tested across dozens of client headshots and verified to produce consistent, studio-matched results.

When Consistency Breaks: The 8 Failure Patterns and Exact Fixes

Even with locked variables, certain AI outputs fail predictably. Eyes looking in different directions. Skin that's too shiny or too dry. Background color that shifts. The 8 Common AI Headshot Failures section in Headshot Prompt Anatomy shows each failure, explains why it happens, and gives you the exact diagnostic phrase to add to your prompt to fix it—no guessing, no regenerating ten times hoping it works. Example: If your client's eyes are looking off-camera instead of at lens, the fix is specific phrasing about "direct eye contact" or "looking straight into the camera." If skin looks plastic-smooth when you want texture, you add "skin texture visible" or "natural skin pores." If the background keeps shifting color, you specify the exact color alongside the blur depth. These aren't prompt engineering tricks—they're photographer-to-AI translations of retouching notes you'd give your studio assistant.

FAQ

Do I need to learn Midjourney inside and out to use this?

No. This teaches you the six variables that actually matter for headshot consistency, gives you pre-written phrases to paste, and shows you exactly which ones to change for different client types. You're not learning prompt engineering—you're learning the photographer's framework for consistency.

Will the same prompts work for every AI tool, or just Midjourney?

The 6-variable structure and the translation glossary work across Midjourney, DALL-E, Stable Diffusion, and other image generators. The exact tested prompts are calibrated for Midjourney, but the framework transfers.

How long does it take to go from 'inconsistent results' to a locked style?

Most photographers lock their first signature style in one or two sessions using the 5-iteration workflow. Once you've locked one style (like corporate neutral), locking a second or third takes even less time because you understand the variable adjustments.

Can I use this to show clients preview variations before booking a studio session?

Yes. The three fill-in-the-blank master templates (Corporate Neutral, LinkedIn Vibrant, Creative Editorial) are designed exactly for this—generate three distinct looks in under 30 minutes, show them to your prospect, lock the winner before the shoot.

What if I already use Midjourney but my results are all over the place?

The Quick-Reference Card shows which of the 6 variables you're probably changing accidentally between sessions. Most photographers lock consistency within one complete shoot after seeing which variables were drifting.

Related

AI Design Prompt Formulas →Prompt formulas for brand identity, client briefs, design revisions, and AI-assisted creative workflows. How to Brief AI Image Generators So You Get Usable Product Photos on the First Try →Stop re-prompting endlessly. Learn the exact brief structure that gets AI product renders right the first time. How to Keep Brand Colors Consistent When Using AI Design Tools →Stop AI from changing your approved brand colors mid-project. Learn the exact prompt structure that keeps palettes locked across all generations. How to Prevent Your Approved Logo from Morphing During Brand Guideline Updates →Keep your approved logo stable while iterating other brand elements. The revision-control prompt structure that stops unwanted AI drift in logo design. Brand Identity Prompt Formulas →Turn any 3–4 sentence client brief into an on-brand logo concept, color palette, and brand guideline visuals in 2–4 AI generations — not 15+.... Prompt Formulas for Client Revisions →Most revision cycles aren't caused by bad prompts — they're caused by a missing translation layer between what clients feel and what prompts can specify....