How to structure a prompt for a geometric ruby-skinned figure, keeping facet edges sharp and material reflections distinct by limiting competing surface tokens.
A hyper-realistic portrait of a humanoid figure with skin composed of glossy faceted ruby-red geometric polygonal panels, each facet reflecting light with mirror-like precision, wearing angular black crystalline sunglasses that form a sharp geometric star pattern across the eyes, large ornate earrings covered in clustered red gemstones and crystals, a single large teardrop ruby pendant at the throat, seamless panel transitions with subtle gold metallic edging between facets, smooth glossy finish with high specular highlights, symmetrical frontal composition, dramatic studio lighting from above creating defined reflections on each facet, deep charcoal gray background, fashion photography aesthetic, ultra-detailed 8K render, octane render quality, luxury jewelry campaign style --ar 9:16 --style raw --s 750
Stacking multiple gemstone names in a single prompt—ruby, diamond, sapphire, obsidian—usually degrades the output into a uniform warm-red plastic sheen. The model averages overlapping material tokens into a midpoint, softening facet edges and flattening the surface. A portrait built from one dominant material specification, like faceted ruby, with subordinate elements kept brief, forces sharper separation because there is only one primary surface physics instruction to resolve.
Defining the Panel Geometry
The words "faceted" and "geometric polygonal panels" serve different functions. "Faceted" establishes discrete flat planes instead of a continuous curve. "Polygonal panels" dictates that those planes have hard, straight edges rather than organic or rounded shapes. Dropping either term changes the result: without "faceted," the skin renders as a smooth solid; without "polygonal," the planes tend to soften into overlapping scales.
"Seamless panel transitions" instructs the model to place these polygons edge-to-edge without gaps. This is what separates a polished gemstone surface from a shattered-glass effect. The "subtle gold metallic edging" then reads as an intentional accent line rather than a structural repair. If the panel boundaries blur into gradients during generation, the geometry tokens are likely being outweighted by realism descriptors later in the prompt. Moving "hyper-realistic" earlier in the text, or reducing competing surface adjectives, can restore edge definition.
Gloss, Specular Highlights, and Reflections
"Glossy finish" describes a reflectance profile where light bounces off the surface rather than absorbing into it. "High specular highlights" describes the visible outcome: tight, bright concentrations of light where a facet meets the source at the correct angle. Both are needed because models often treat outcome words as compositional instructions. "Mirror-like precision" adds a third constraint, pushing the reflections toward sharp, undistorted returns rather than diffuse scattering. This is what moves the visual read from polished resin toward crystalline stone.
If reflections appear smeared, check for nearby softness words like "velvet" or "matte" and remove them. Increasing a stylization parameter rarely fixes this; removing contradictory tokens does. The phrase "octane render quality" pulls the aesthetic toward a path-traced appearance where reflections behave consistently across adjacent planes. Look for variation in how each facet catches the light source—if every panel reflects identically, the three-dimensional geometry is not being resolved.
Lighting Direction and Background Value
A faceted surface is only legible because different planes catch light differently. "Dramatic studio lighting from above" provides a single dominant source, producing one consistent set of highlight angles across the face. Multi-source lighting scatters reflections and reduces contrast between adjacent panels, flattening the geometry.
The "deep charcoal gray background" provides a neutral field that keeps both the red facets and their specular hits readable. A white background washes out the highlights; a pure black background dissolves the outer silhouette of the figure. Charcoal sits in the narrow value band that preserves both. "Symmetrical frontal composition" locks the geometry into a mirror axis, which is necessary for the viewer to read the left and right panels as deliberate reflections of each other.
Supporting Elements and Detail Density
The "angular black crystalline sunglasses" act as a contrast anchor. Black and ruby sit at opposite ends of the value scale, creating a hard visual break that prevents the portrait from reading as a single red mass. Keeping the glasses described as "crystalline" maintains the material family without introducing transparency that would compete with the skin's facet reflections.
The earrings ("clustered red gemstones and crystals") and the "single large teardrop ruby pendant" introduce high-frequency detail against the large flat panels of the face. Without them, the composition lacks the texture variation expected in a luxury jewelry campaign aesthetic. However, if these small details merge into blobs, the model is prioritizing overall composition over micro-detail. Describing supporting elements in fewer words often resolves cleaner than layering additional adjectives onto each item.
Prompt Parameters and Platform Limits
The supplied prompt includes specific flags: --ar 9:16, --style raw, and --s 750. These are Midjourney parameters. --ar 9:16 sets a vertical aspect ratio suited for mobile viewing or poster formats. --style raw reduces the platform's default aesthetic filtering, giving the prompt text more direct control over the output. --s 750 sets the stylize value, balancing adherence to the prompt against artistic interpretation. These flags will not function on platforms like Stable Diffusion or DALL-E. If porting this prompt elsewhere, strip the double-dash parameters and rely entirely on the natural language descriptions.
Where This Technique Stops Working
This approach relies on a single dominant material. If the subject is meant to combine ruby facets with matte fabric, soft organic skin, or translucent glass, the logic breaks down. The model struggles to hold two contradictory surface physics in the same latent space without one bleeding into the other. The visible signal is the non-faceted areas picking up unnecessary gloss, or the facets softening into painted textures.
Density limits also apply. At a 9:16 aspect ratio with panel geometry, eyewear, earrings, and a pendant all specified, the model has limited resolution to dedicate to each element. The smallest details degrade first. When the earrings turn into indistinct red noise and the panel edges lose their gold lines, the prompt has exceeded what the framing can carry. Removing one supporting element usually restores clarity to the remaining surfaces.
This is a simulation of physical optics, not a photograph. The model approximates how light interacts with flat planes based on its training data; it does not compute actual ray tracing. Expect fine edge details to soften depending on the generation platform and resolution.
Comments
Post a Comment