Struggling to Mix Watercolors? I Built a GPT to Solve It. скачать в хорошем качестве

Struggling to Mix Watercolors? I Built a GPT to Solve It.

FREE RESOURCES MENTIONED IN THE VIDEO: https://drive.google.com/file/d/1FN5E... In this video, I walk you through the full workflow behind my custom GPT-powered watercolor mixing assistant — and how I used it to create four printable recipe cards from this blue poppy reference image. The focus of this video is the GPT itself and how it runs. As a beginner watercolorist, I struggled with translating reference images into confident, repeatable color mixes — especially neutrals. Traditional color theory advice wasn’t specific to the exact Daniel Smith paints I own, which made mixing feel inconsistent and overwhelming. So I built a solution. I created a personal paint database using universal pigment codes and designed a custom GPT that generates inventory-aware mixing recipes using only the pigments I physically have. The system breaks a reference image into hue families, temperature shifts, and value relationships — then outputs 3–5 structured, part-based mixing formulas that are actually achievable with my current paint set. In this video you’ll see: • The GPT running from start to finish • How I break a reference image into color-blocked sections • How the assistant generates structured mixing recipes • How I turn those into clean PNG recipe cards in Procreate • How I sketch the image for transfer to watercolor paper This is a practical, repeatable system that turns visual observation into mixable color formulas. If you have questions about the workflow or want a deeper dive into any part of the process, leave a comment — I’m happy to create follow-up videos.