Verified against Midjourney · 2026-08-08
Set an exact lighting ratio for a moody, cinematic low-key headshot
A short, direct portrait brief built around naming an exact key-to-fill lighting ratio and a specific film-stock reference, so a request for a 'cinematic headshot' stops meaning generic dark-and-moody and starts meaning a controllable, repeatable look.
The prompt
Ready to copy — highlighted parts are example details you can swap.
SUBJECT a middle-aged jazz trumpeter with close-cropped grey hair and reading glasses pushed up on his forehead LIGHTING RATIO 4:1 key-to-fill. State the number, not just "dramatic" or "moody" — a ratio like 4:1 tells the model exactly how deep the shadow side of the face should fall relative to the lit side, while a mood adjective alone leaves that decision to Midjourney's own default, which trends toward flatter, more evenly lit faces than most cinematic references actually use. FILM EMULATION REFERENCE shot on Kodak Vision3 500T tungsten-balanced stock, slightly warm highlights against cool shadow — name a specific stock or digital-cinema look rather than the word "cinematic" on its own, since that word alone has no fixed visual meaning to the model and gets interpreted differently prompt to prompt. FRAMING AND CROP tight head-and-shoulders crop, three-quarter angle MOOD quiet intensity, mid-performance focus rather than a posed studio smile WHAT NOT TO DO Do not add a heavy vignette on top of the lighting ratio above — a strong key-to-fill ratio already darkens the frame's edges naturally as a byproduct of the light falloff, and stacking an explicit vignette instruction on top of that tends to crush the shadow side into pure black instead of the graded, detailed dark the ratio was meant to produce. PARAMETERS --ar 2:3 --stylize 150 --v 7 OUTPUT a middle-aged jazz trumpeter with close-cropped grey hair and reading glasses pushed up on his forehead, lit with a 4:1 key-to-fill ratio, shot on Kodak Vision3 500T tungsten-balanced stock, slightly warm highlights against cool shadow, tight head-and-shoulders crop, three-quarter angle, quiet intensity, mid-performance focus rather than a posed studio smile --ar 2:3 --stylize 150 --v 7
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Naming an explicit key-to-fill ratio gives Midjourney a quantifiable instruction instead of a mood word with no fixed meaning — "4:1" specifies exactly how much darker the shadow side of the face should read relative to the lit side, whereas "moody lighting" or "dramatic lighting" is a phrase the model has seen attached to an enormous range of actual contrast levels in training, so it produces an inconsistent, often too-flat result because it has no single number to converge on. V7's default portrait behavior leans toward even, flattering, front-heavy fill light — a holdover from the aesthetic bias baked into what the model considers a "good" portrait — so an explicit ratio is doing real corrective work against that default rather than just adding flavor text on top of it.
Naming a specific film stock instead of the word "cinematic" works for the same reason the ratio does: "cinematic" is a category label with no fixed visual content, while "Kodak Vision3 500T, tungsten-balanced" names an actual color response — warm highlights, cooler shadows, a specific grain structure — that the model has strong, consistent associations for because that stock's look is well-represented and consistently labeled in its training data. Two different film-stock names will reliably produce two visibly different color grades from the same lighting ratio, which is the whole point of naming one specifically rather than reaching for a generic mood tag.
The warning against stacking a vignette on top of the ratio targets a specific, avoidable failure: a strong lighting ratio already produces natural edge falloff as a physical consequence of the light source's falloff curve, and adding an explicit vignette instruction on top of that double-darkens the frame edges, frequently crushing the graded, still-detailed shadow side the ratio was built to produce into flat, textureless black — losing exactly the shadow detail a genuine low-key cinematic look depends on to read as lit rather than simply underexposed.
What you get back
A tightly cropped three-quarter portrait with a clearly deeper shadow side of the face than lit side, warm tungsten-leaning highlights and a cooler shadow tone consistent with the named film stock, and visible shadow detail rather than crushed black at the frame edges.
Verified against
Midjourney v7 · 2026-08-08
Changelog
- 2026-08-08 — Initial publish, verified against Midjourney v7 for explicit lighting-ratio cinematic portraits.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
