Drama Color Pattern Analysis

Color grading is a defining visual element of Korean dramas, yet previous research had two blind spots: studies either compared color differences between dramas without connecting them to viewer preferences, or measured preferences using artificially manipulated colors detached from real content.

This study bridges both approaches. By cross-applying the actual color properties from the Spanish original Money Heist (La Casa de Papel) and its Korean remake Money Heist: Joint Economic Area, we tested whether viewers prefer Korean color grading — and whether that preference stems from the colors themselves or from personal attributes like nationality.

The core challenge was creating a fair comparison: how do you isolate the effect of color grading from everything else in a scene? We designed a six-stage pipeline that extracts, measures, and cross-applies color properties between the two versions of the same drama.

This study quantified something that content platforms and media companies have long relied on intuition for: how color composition affects viewer engagement. The finding that saturation harmony — not contrast or brightness — drives preference has direct applications in automated color grading pipelines, content recommendation systems, and A/B testing visual assets.

The methodology itself — combining deep learning-based scene clustering with color space analysis and statistical validation — demonstrates a reusable pipeline for any domain that needs to extract and evaluate visual features from large media datasets. This is the kind of computational analysis that turns subjective creative decisions into data-driven optimization.