pages.practiceQuickWins.hero.title

Create [number] concept art variations of [subject], in the style of [aesthetic reference]. Include variations in lighting (golden hour, overcast, dramatic chiaroscuro), color palette (warm earth tones, cool pastels, high-contrast neon), and composition (wide establishing shot, intimate close-up, bird's-eye view). Maintain consistent character/object design across all variations. --ar 16:9 --style raw --v 6

Tips

• Start with a broad description and narrow down after seeing initial results
• Use the /describe command on existing artwork to reverse-engineer effective prompts
• Save your best prompts in a personal library for future projects
• Combine multiple style references using image weights for unique hybrid aesthetics

Cautions

• AI-generated concept art is a starting point, not a finished product — always refine by hand
• Be cautious about referencing specific living artists by name in your prompts
• Verify that generated compositions don't inadvertently replicate existing copyrighted works

Analyze this artwork and extract a professional color palette. For each color, provide: (1) the hex code, (2) a descriptive name, (3) the percentage it occupies in the composition, (4) its emotional/psychological association, and (5) suggested use cases in design (background, accent, text, CTA). Also identify the overall color harmony type (complementary, analogous, triadic, etc.) and suggest two alternative palettes inspired by the same mood but with different hue families.

Tips

• Upload the actual image to a multimodal AI for most accurate color extraction
• Ask the AI to suggest accessible color combinations that meet WCAG 2.1 AA contrast ratios
• Request palettes in multiple formats: hex, RGB, HSL, and Tailwind CSS class names
• Use this technique to study the palettes of master painters — analyze Vermeer, Hopper, or Klimt

Cautions

• AI color extraction is approximate — verify hex codes with a proper eyedropper tool
• Screen calibration affects how you perceive the extracted colors

Create a 60-second instrumental track based on this melody [hum/upload reference]. Render it in [target genre] style with authentic instrumentation and production techniques typical of the genre. Maintain the original melodic contour and harmonic progression, but adapt the rhythm, tempo, arrangement, and timbre to fit the new style. Include an intro and a natural ending.

Tips

• Start with a simple melody — complex sources produce muddy results across genre transfers
• Generate three to five genre variations of the same melody to hear which direction resonates
• Use the Custom mode in Suno to specify exact instrumentation and song structure
• Layer AI-generated variations with your own live recordings for hybrid textures

Cautions

• AI music generators may produce melodies that closely resemble existing copyrighted songs — always check
• The output quality varies significantly across genres; electronic styles tend to be more convincing than acoustic
• Rights and licensing for AI-generated music remain legally ambiguous in many jurisdictions

Transform this photograph into a Studio Ghibli-inspired illustration. Maintain the original composition and key subjects, but reimagine the scene with: (1) soft, watercolor-like lighting with warm golden tones and gentle gradients, (2) lush, hand-painted environmental details — every leaf, cloud, and surface should feel lovingly rendered, (3) slightly simplified but expressive facial features if people are present, (4) a sense of quiet wonder and emotional warmth characteristic of films like Spirited Away and My Neighbor Totoro, (5) visible brushwork texture that evokes cel animation painted on backgrounds. Keep the color palette naturalistic but slightly more saturated than reality, with particular attention to sky gradients and natural light.

Tips

• Photos with natural landscapes, small-town streets, or cozy interiors translate best to the Ghibli aesthetic
• If the result looks too digital, add 'hand-painted texture with visible brushstrokes on watercolor paper' to your prompt
• Try specifying a particular Ghibli film's mood — 'the melancholy of Grave of the Fireflies' produces very different results from 'the whimsy of Kiki's Delivery Service'
• Upload the highest resolution version of your photo for the most detailed transformation

Cautions

• Studio Ghibli is a trademarked studio — use 'Ghibli-inspired' or 'Ghibli-style' language and understand that outputs are derivative, not officially affiliated
• AI transformations may alter faces significantly — do not use this for portraits where likeness accuracy matters without consent
• The viral Ghibli filter trend of 2025 raised legitimate concerns from animators about stylistic appropriation — be thoughtful about context and credit

I am writing a song and I have created a mood board to capture the feeling I want. Here is what is on the mood board: [describe images, colors, textures, words, and emotional associations]. The genre is [genre]. The intended listener should feel [target emotion]. Write lyrics for a complete song (verse 1, chorus, verse 2, chorus, bridge, final chorus) that translates the visual and emotional language of this mood board into words. Use concrete sensory imagery rather than abstract statements. Match the syllable patterns and internal rhyme density to the conventions of [genre]. Include a suggested song title. After the lyrics, explain which mood board elements influenced each section.

Tips

• Photograph your physical mood board or screenshot your digital one and upload it directly — multimodal AI can read the visual language
• Include specific sensory details in your mood board description: the smell of rain on concrete, the texture of frayed denim, the sound of a distant train
• Ask for three versions with different emotional angles on the same mood board — one hopeful, one melancholy, one angry — to explore the range
• Specify the vocal delivery style you imagine (whispered, belted, spoken word) to influence word choice and line length

Cautions

• AI-generated lyrics often default to cliches — push past the first draft by asking for less obvious metaphors and more specific imagery
• Always check generated lyrics against existing songs to avoid unintentional similarity to copyrighted material
• The emotional authenticity of lyrics depends on your personal revision — AI provides raw material, not finished art

Storyboard frame [frame number] of [total frames]: [scene description from screenplay]. Camera angle: [wide/medium/close-up/over-the-shoulder/bird's-eye/low-angle]. Lens: [wide-angle/standard/telephoto]. Lighting: [key lighting description]. Character(s): [brief character description for visual consistency]. The frame should read as a professional cinematographic storyboard panel in graphite pencil on white paper, with clear composition and implied motion. Include directional arrows for camera or character movement if applicable. --ar 16:9 --style raw --v 6

Tips

• Number each frame sequentially and include brief action notes below each generated image to create a readable storyboard document
• Use Midjourney's character reference feature (--cref) with a previously generated character image to maintain visual consistency across frames
• Generate each key story beat as its own frame — aim for 6 to 12 frames per scene, covering establishing shot, key actions, reactions, and transitions
• Add '--no color' to your prompt if you want traditional black-and-white storyboard panels

Cautions

• AI-generated storyboards lack the deliberate cinematographic intent of a human storyboard artist — use them as conversation starters with your DP, not as final shot lists
• Character consistency across multiple frames remains imperfect even with reference features — expect to do some visual editing for presentations
• Storyboard frames generated by AI may default to conventional compositions — push for unusual angles and dynamic framing in your prompts

Design a typography-based logo for the brand name '[BRAND NAME]'. Style: [style description]. The logo should be a clean, professional typographic treatment on a plain white background. Focus on letterform design, spacing, and visual hierarchy. The text must be clearly legible and spelled correctly. No icons, no symbols, no imagery — pure typography. Render as a vector-style graphic with crisp edges. --ar 3:1 --style raw --v 6

Tips

• Run five separate prompts with different style descriptions: serif classic, geometric sans-serif, hand-lettered script, brutalist/experimental, and retro vintage
• Use short brand names (1-2 words) for best results — AI struggles with long text and multi-word spelling
• Always double-check the spelling in generated logos — AI frequently misspells words in images and you may need to regenerate
• Use the outputs as a style direction presentation for clients, not as final logo files — redraw the selected direction in Illustrator for production use

Cautions

• AI-generated type is not vector art — it cannot be scaled infinitely or edited as outlines without being redrawn in a vector tool
• Text generation in AI images remains imperfect — expect letter-level errors that require manual correction
• AI may inadvertently reproduce elements of existing logos — always run a visual similarity check before developing any direction further

Restore this vintage photograph by: (1) removing visible scratches, dust spots, and emulsion damage, (2) correcting faded contrast and restoring a full tonal range from deep blacks to clean whites, (3) repairing any torn or missing areas using contextually appropriate fill, (4) sharpening soft details while preserving the natural grain structure of the original film stock. Then colorize the restored image with historically accurate colors based on the estimated era [decade], geographical region [region], and subject matter. Apply color with naturalistic subtlety — skin tones should be warm and varied, fabrics should reflect period-appropriate dyes and materials, and environmental colors should match the region and season.

Tips

• Scan the original photograph at 600 DPI or higher before processing — the more detail the AI has to work with, the better the restoration
• For historically accurate colorization, provide context: 'This is a 1940s photograph taken in rural Ireland' produces much better color choices than uploading without context
• Process restoration and colorization as separate steps for more control — restore first, then colorize the restored version
• Compare your colorized result against known color photographs from the same era and region to check plausibility

Cautions

• AI colorization is an educated guess, not historical truth — the actual colors of clothing, paint, and objects may have been completely different
• Restored and colorized photos can feel deceptively 'real' — always label them as AI-enhanced when sharing publicly or in archival contexts
• Heavy-handed restoration can erase the character and grain of vintage photographs — preserve the image's age gracefully rather than making it look like a modern photo

Create a [duration]-second instrumental background track for a [video type] video. Genre: [genre/mood]. Tempo: [BPM or descriptor]. The track should have: (1) a gentle intro that works under an opening title or voiceover, (2) a subtle build through the middle section that supports visual storytelling without competing for attention, (3) a natural resolution or fade that suits an ending. Instrumentation: [specific instruments]. The mix should be balanced for use under narration — no dominant melodic lines that clash with speech frequencies, emphasis on texture, rhythm, and atmosphere. Avoid sudden dynamic changes that would distract from visual content.

Tips

• Specify the exact duration you need — generating a 45-second track is more useful than trimming a 3-minute song
• Use descriptive mood language rather than technical music jargon: 'hopeful but not cheesy' or 'tense without being horror-movie dramatic' communicates clearly to AI models
• Generate three to four versions and audition each one against your video timeline before committing — what sounds great in isolation may not work under your footage
• If the track will play under narration or dialogue, specify 'leave frequency space for human voice in the 300Hz-3kHz range' for a cleaner mix

Cautions

• Licensing terms for AI-generated music vary by platform — read Suno's and Udio's current terms carefully, especially for commercial use
• AI music generators occasionally produce passages that closely resemble existing copyrighted songs — listen critically and compare against known works before publishing
• Music generated at the free tier of most platforms may have usage restrictions that prohibit commercial distribution