Deep dives into modern screenwriting, visual storytelling, and how AI is reshaping the creative process.
Categories
Dialogue scene prompts that hold together: plan the coverage first, write screen sides and eyelines into every prompt, label each speaker and keep lines short enough for the clip.
Dialogue Scene Prompts: Two-Shots, Coverage and Eyelines for AI Video
Kling prompts that land: one visible camera move per shot, motion that agrees with the subject, beats spread across 3 to 15 seconds, and what Kling 4.0's 30-second clips change.
Kling Prompts: Directing Motion in Kling 3.0, and What Kling 4.0 Changes
Veo 3.1 prompts that hold up: pick settings first (1080p and 4K mean 8 seconds), write one moment per clip, give sound its own sentences, and keep dialogue off the screen.
Veo Prompts: The Cinematic Quality Settings That Matter in Veo 3.1
Seedance 2.5 prompts that work: bind every asset by number, write one summary line, time beats in whole seconds, and keep negatives for subtitles and audio only.
Seedance Prompts: What Seedance 2.5 Actually Rewards
Which video models have a negative prompt field, how to name the defect you see, why you write nouns instead of orders, and where the fix goes when the field is missing.
Negative Prompts for AI Video: Kill the AI Look Without Fighting the Model
How to write the character part of an AI video prompt: bind the reference image, describe only what it cannot show, freeze one identity line and cut the words that move the face.
Character Consistency Prompts: What the Text Should Say When the Image Holds the Face
Why a stock name or grade label rarely sets the look of an AI video shot, and what to write instead: the light sources, color balance, black level, saturation and grain.
Film Stock and Grade Prompts for AI Video: Kodak, Noir, Teal-Orange
Why a focal length alone rarely changes an AI video shot, and what to write instead: camera distance, background size, depth of field, a table from 14mm to anamorphic.
Lens Prompts for AI Video: 24mm vs 50mm vs 85mm Looks