Seeds of Renaissance · YouTube · podcast thumbnails · backgrounds

Thumbnail backgrounds

Our first thumbnails sit on flat white, yellow or black: one value, one colour, nothing to lead the eye. These sixteen backgrounds try the other way. Each is built in the steps below and grouped by its main idea: light and depth, one big shape, landscape and drawing, light through things. Use the switch to see each one alone, with a face and words on top, or blurred.

What makes a good background

  1. A value plan. Light and dark placed on purpose: a glow behind the head, calmer tone under the words. Squinting, it reads as two or three big shapes.
  2. Depth. A gradient, a vignette, a glow or layered hills fading into distance turn a colour into a place.
  3. Two zones. Rich behind the face, quiet behind the words. Never texture under text.
  4. Big shapes, not fine detail. At 168 px fine lines turn to mud. One large disc, ridge, ring or shadow survives.
  5. A light source. A sun, a dawn or light through leaves gives the glow a reason and lifts the head off the ground.
  6. Surface. Grain, paper or ink makes it feel made, not flat digital colour.
  7. One recurring motif. So the series is recognised before it is read.
  8. Contrast with the subject. The face is greyscale, so a warm ground makes it stand out; keep the darkest and lightest areas away from its edges.

Steps

  1. Zones. On 1280 × 720: words left, face right, logo top left, duration badge bottom right.
  2. Value sketch. Where it is light, where it is dark. Squint-test it.
  3. Colour family. Sunlight, night, paper, plus the drawings' cobalt and gold.
  4. Depth layer. A gradient, glow or vignette.
  5. One motif. Large, low contrast, behind the face, cropped by the frame.
  6. Surface. Grain, paper or print texture.
  7. Test. With face and words on top, blurred, and at 168 px in light and dark feeds.
  8. Systematise. Decide what varies per episode (colour, motif, glow) and what stays fixed.

A first finding. A face as big as version A's covers most of any background, which then only shows at the edges. These tests use a slightly smaller face (about 70% of the height) and put each motif around the head, not under it: above it, beside it, or as a ring or horizon that the head breaks.

Tags: PALETTE stays within colour.md; STRETCH uses tints or colours beyond it (deeper golds, cobalt washes, a faint red) and would need a palette decision. The drawings and the dawn photo are drafts whose rights are not yet cleared.

View
Combination · 2 + 10

Dawn ridge

The dawn gradient and the band of light across the middle from 2, with the brush ridge from 10 rising out of it. Behind the head the ridge was hidden, so here its main peak stands in the gap between the words and the face, rising from the horizon below the words. Four variations: the plain combination, the ridge emerging from mist, a warmer gold dawn, and a second, paler ridge behind for more distance.

1 · Light and depth

Tone alone, no motif

The cheapest step away from flat: a glow behind the head, a vignette, a horizon. Works for every episode and never fights the words.

2 · One big shape

A disc, rays, a ring

A single large, simple shape behind the head. It survives at feed size and can become the series signature.

3 · Landscape and drawing

Mountains, drawn and photographed

The mountain drawings and the dawn photo as scenery, faded out before they reach the words.

4 · Light through things

Shadows, ink and paper

Softer, more handmade: dappled shadow, ink wash, watercolour, torn paper.

Feed check

All of them at 168 px, with face and words

Sidebar size, the hardest test. Which still read as a place, not a flat colour, and which turn to noise?