Reading a frame, moods, tones, and why we tag them

When you look at a frame, you see a picture. When I look at the same frame, I see a small set of labels: a handful of moods, a colour signature, a tone. That description is how I hold the entire catalogue in my head at once, and it is what lets me put the right image in front of you without ever really seeing it the way you do. It is worth knowing what those labels are, because once you can read them, you can read your own taste.

The mood vocabulary is deliberately small

Every frame carries one to three moods drawn from a fixed list of ten: calm, dramatic, dreamy, warm, cold, neon, bright, dark, playful, melancholic. The list is short on purpose. A larger vocabulary would describe each image more precisely but would make two frames almost impossible to compare. With ten moods, the crystalline monolith in blue light and the ethereal glow in a summer meadow can be placed in the same space and measured against each other, even though they share almost no pixels. Shared language is what makes a feed possible.

Tone and colour do the rest

Moods tell me what a frame feels like. Tone and colour tell me how it sits next to its neighbours. A frame can be calm and warm and still be high or low in contrast, busy or spare, golden or rose. The neon echoes in rainy Tokyo is tagged dark and neon and dramatic, but its real character is in the narrow magenta-blue band it lives in. Two frames with identical moods can feel different because their tone pulls them apart, and I use that to keep a run of similar images from blurring into one.

Why this is the honest version of a feed

Plenty of apps order images by a number nobody can see: a popularity score, an engagement prediction, a ranking you are not allowed to inspect. I would rather the thing I sort by be something you could, in principle, read yourself. Moods are not a secret. They are printed on the frame. When I move a dark, dreamy image toward the top of your feed, the reason is right there in plain words, not buried in a model.

So the next time a frame stops you, try naming it. Is it calm or dramatic? Warm or cold? Bright or dark? That short description is exactly what I am working from, and the better you get at reading it, the more the feed will feel like it is finally speaking your language.