Drop the shadow value slowly, with no visible line
Built the modeling out of up to roughly thirty microscopic translucent oil glazes and scumbles laid over dried underlayers, with a high oil-to-pigment ratio, lowering the shadow value by degrees.
Why it matters · A hard line between light and shadow reads as an edge, not a turn in space. When the value drops across many thin layers instead of one opaque pass, the form turns without ever announcing where the shadow begins. The patience is the whole effect. One layer cannot do it.
Gemini Deep Research, 5 famous painters, 2026 — Compiled from the Materials and technique section.
Prepare exhaustively, never start blank
Worked out anatomy and composition in sketches on paper first, then transferred them to the panel by spolvero, pricking holes along the lines and pouncing charcoal through them.
Why it matters · The drawing problem and the painting problem are different problems. Solving placement on paper means the panel is never a guessing field. Infrared imaging found this pounced underdrawing beneath the Mona Lisa's forehead and hands, so even the most famous "spontaneous" face sits on a transferred plan.
Gemini Deep Research, 5 famous painters, 2026 — Process from blank canvas section.
Refuse the medium that forbids revision
Rejected buon fresco for The Last Supper because its wet plaster allows no slow reworking, and invented an experimental dry-wall method on a gesso, pitch, and mastic ground instead.
Why it matters · The material has to match how you actually work. Leonardo reworked endlessly, and fresco demands a section be finished in a day. So he chose a method that let him revise. The method failed chemically and began flaking in his own lifetime, which is the honest cost of choosing revision over permanence.
Gemini Deep Research, 5 famous painters, 2026 — Materials and technique section.
Paint when the light is soft, not bright
Diffused the sun with canvas awnings stretched over courtyards and favored dusk or overcast weather for portraits, because the softest shadows gave the most grace.
Why it matters · Hard direct light cuts the face into flat planes of black and white. Soft light keeps the half-tones, and the half-tone is where the modeling lives. If you control the light at the source, the painting is easier before you have made a single mark.
Leonardo, Trattato della pittura (compiled by Melzi), 1651
Work in long bursts, then stop cold
On The Last Supper, Matteo Bandello saw him climb the scaffold at sunrise and paint until dusk forgetting to eat, then on other days stand and stare for hours without a mark, or leave it for three or four days.
Why it matters · Output is not steady, and pretending it should be produces forced work. The staring is part of the painting. The decision of where the next stroke goes can take longer than the stroke. A reader who measures progress only in marks made will quit on the staring days.
Matteo Bandello, Novelle, 1554
Find landscapes in stains on a wall
For invented backgrounds, stared at walls splashed with stains, at ashes and puddles, until he saw mountains, ruins, rocks, and woods in them, then reduced the hallucination to drawn form.
Why it matters · Invention does not have to start from nothing. A random texture gives the eye something to push against, and the mind supplies the rest. The skill is the second half: taking the thing you imagined and pinning it down as actual form. The prompt is free. The reduction is the work.
Leonardo, Trattato della pittura (compiled by Melzi), 1651
Let a picture run for years rather than declare it done
Rarely treated a painting as finished; worked the Mona Lisa over the course of a decade.
Why it matters · Some paintings are not a sprint. Holding a picture open for years means returning to it with a changed eye and finding what the last pass missed. The risk is the opposite of overwork, which is never letting go. Leonardo lived on that edge on purpose.
Gemini Deep Research, 5 famous painters, 2026 — Process from blank canvas section.