Colour studies of films, advertising and other moving images, shown once.
Browse by title, type, cinematographer or colorist. Combine filters to
find a credited work across all its studies. The complete research index still holds everything.
82 reports
No colour studies match these filters. Clear a filter to see more.
Thirty curated frames from Ffern's Autumn 26 film Hara & the Hare, each read against scope baselines built from all 220 photographed samples — a corpus taken every 0.48 seconds on a clock, because the film's 38 cuts are too few to make a baseline of.
The two Joker films, read frame for frame from an 8-bit web copy and a 10-bit master. The colour of the dark survives. The web copy forgets its grain, the fine steps between one shadow and the next, and all light above an ordinary screen's white; for Folie à Deux it also carries a second grade that yellows the dark.
Four films graded by Jill Bogdanowicz, read against every re-measured shot of 51 films. Her darks are not more colourful than cinema's. They are coloured by subtraction: one channel leaves the toe before the others, so black turns cyan — or, in one film, magenta — and reads saturated without being chromatic.
Forty-three curated active-picture midpoint frames from Joker (2019), each read against scope baselines built from all 1,302 photographed shots of the film.
Thirty-six curated active-picture midpoint frames from The Bride! (2026), each read against scope baselines built from all 2,324 photographed shots of the film.
Eight frames from Joker, where a city is made of sodium amber, institutional green-teal fluorescent and overcast olive daylight, where the clown's white greasepaint and red suit have to survive every one of those lights, and where the grade's hardest work is keeping skin human in rooms that are designed to make it sick.
Eight frames from The Grand Budapest Hotel, where pink stucco, elevator red, lobby-boy purple and 1968 tangerine are painted, dyed and built rather than graded in, and the grade's job is to carry saturations up to C* 48 without clipping, keep faces alive inside coloured rooms, and let miniatures, night and snow sit in the same storybook as the sets.
Eight frames from the dark passages of The Bride!, a film whose median frame sits at L* 6.7, where most of the picture is held at black on purpose and every face, lamp, spotlight and neon sign has to be kept alive in a small pool above a floor that an 8-bit delivery can barely describe.
Forty-four curated active-picture midpoint frames from The Grand Budapest Hotel (2014), each cropped to its own era's geometry and read against scope baselines built from all 1,024 photographed shots of the film.
Eight frames from The Aviator's first fifty-three minutes, where every colour in the picture is made to lie on one line through grey — skin and red paint at one end, sky, grass and peas at the other — and the grade's work is to place those two poles, protect the faces on the warm one and let everything that should have been green go cyan.
Grain in 41 delivered films, calibrated against known clips and compared across photographic origin, tonal placement, colour and simulated Resolve workflows.
Thirty-one curated active-picture midpoint frames from Joker: Folie à Deux (2024), each read against scope baselines built from all 1,320 photographed shots of the film.
Thirty curated active-picture midpoint frames from Nosferatu (2024), each read against scope baselines built from all 500 photographed shots of the film.
Forty curated active-picture midpoint frames from Hunt for the Wilderpeople (2016), each read against scope baselines built from all 1,512 photographed shots of the film.
Sixty-four curated active-picture midpoint frames from The Aviator (2004), each read against scope baselines built from all 3,456 photographed shots of the film.
Thirty-nine curated active-picture midpoint frames from Pan's Labyrinth (2006), each read against scope baselines built from all 1,123 photographed shots of the film.
Sixty curated active-picture midpoint frames from The Shape of Water (2017), each read against scope baselines built from all 1,459 photographed shots of the film.
Ninety-five shots of one man crossing a desert to find the trucks, a dead man and the money. What does the scene hold constant as the light moves, and what does it let change?
Fifty-nine shots of two men across a counter, and a coin. What does the scene hold constant as the camera moves between them, and what does it let change?
Eight frames in which desert daylight is built as a warm, hard-shouldered field, black costume is held above crushed black, faces keep the side light that models them, and night, fluorescent tube and muzzle flash each ask for a different separation.
Twenty-eight shots of two children and a Rubik's cube in the snow. What does the scene hold constant as the camera moves between them, and what does it let change?
Eight frames from the only corpus in this repository with no colour in it at all — a measured mean chroma of 0.0000 across all 303 shots — and the only one graded from a PQ master, where the top of the scale is not white and every separation the film makes has to be built out of value, texture and edge.
A practical reconstruction study of Ffern: sixteen image-construction problems, twelve documented nodes, and a different set of decisions for each film.
Cinema does not have a highlight roll-off. Measured over 52,527 shots and 9,856 re-measured frames, 95% of the variation in how much a film clips is between shots of the same film — but the colour of a film's whites, and the number of distinguishable steps it spends inside them, do belong to the film.
Two thirds of cinema sits below L* 20. The shape of a film's toe is not a property of the film — its intraclass correlation is 0.003 — while whether a film lets pixels reach absolute zero has 0.23, four and a half times the equivalent number at the top of the range on the same films.
First Cow is said to have been modelled on Frederic Remington's nocturnes. Measured on colour against 36 other films it is one of the least Remington-like in the corpus; measured on seventeen axes that are not colour it is middling. Its firelight and Remington's are the same to the limit of what a reproduction can resolve.
Eight frames from the darkest and least chromatic corpus in this repository, where daylight stops at a third of the scale, shadows are almost neutral, small warm sources carry all the chroma, and nothing clips.
Seven classical hue arrangements fitted to every photographed shot in forty-two films. The fourth edition adds seven films, reports what their arrival did to the corpus, and keeps between one and three reviewed specimens per shape instead of filling a quota.
Eight frames in which snow becomes a luminous environment, small colours acquire narrative force, and pale skin, dark rooms, red-green interiors and underwater cyan each demand a different separation strategy.
Eight depictive frames showing how practical signage, low exposure, restricted colour families, contaminated optics and selective density make Tokyo feel luminous without turning every surface into an effect.
Seven classical hue arrangements fitted to the full photographed corpus. The third edition separates source rows, photographed shots and eligible candidates; shows the purest, louder and third-film examples; and measures how often each film departs from one hue family.
Eight frames in which near-white costume, dense period interiors, mauve spectacle and enormous reconstructed spaces are made to coexist through value design, palette subtraction, protected texture and careful foreground-background integration.
Eight frames in which colour is not a wash laid over the picture but a negotiation between painted rooms, narrow-spectrum lamps, costume, exposure, optical contamination and an unusually fearless finish.
Thirty-seven curated active-picture midpoint frames from Point Blank (1967), each read against scope baselines built from all 693 photographed shots of the film.
Thirty-three curated active-picture midpoint frames from Leviathan (2014), each read against scope baselines built from all 490 photographed shots of the film.
Thirty-four curated active-picture midpoint frames from Another Round (2020), each read against scope baselines built from all 452 photographed shots of the film.
Thirty-five curated active-picture midpoint frames from Let the Right One In (2008), each read against scope baselines built from all 703 photographed shots of the film.
Twenty-eight curated active-picture midpoint frames from Enter the Void (2009), each read against scope baselines built from all 2,391 photographed shots of the film.
Thirty curated active-picture midpoint frames from The Straight Story (1999), each read against scope baselines built from all 794 photographed shots of the film.
Thirty-six curated active-picture midpoint frames from Gangubai Kathiawadi (2022), each read against scope baselines built from all 1,558 photographed shots of the film.
Forty curated active-picture midpoint frames from The Banshees of Inisherin (2022), each read against scope baselines built from all 1,342 photographed shots of the film.
Forty-one curated active-picture midpoint frames from Super Deluxe (2019), each read against scope baselines built from all 1,390 photographed shots of the film.
Thirty-six curated active-picture midpoint frames from Drive (2011), each read against scope baselines built from all 877 photographed shots of the film.
Thirty curated active-picture midpoint frames from No Country for Old Men (2007), each read against scope baselines built from all 1,382 photographed shots of the film.
Thirty curated active-picture midpoint frames from Drive My Car (2021), each read against scope baselines built from all 875 photographed shots of the film.
Thirty curated active-picture midpoint frames from Portrait of a Lady on Fire (2019), each read against scope baselines built from all 423 photographed shots of the film.
Thirty curated active-picture midpoint frames from O Brother, Where Art Thou? (2000), each read against scope baselines built from all 1,125 photographed shots of the film.
Thirty curated active-picture midpoint frames from Moonlight (2016), each read against scope baselines built from all 712 photographed shots of the film.
Thirty curated active-picture midpoint frames from Kwaidan (1964), each read against scope baselines built from all 894 photographed shots of the film.
Thirty curated active-picture midpoint frames from Suspiria (1977), each read against scope baselines built from all 814 photographed shots of the film.
Thirty curated active-picture midpoint frames from Bugonia (2025), each read against scope baselines built from all 1,178 photographed shots of the film.
Thirty curated active-picture midpoint frames from The Conformist (1970), each read against scope baselines built from all 713 photographed shots of the film.
Thirty curated active-picture midpoint frames from A Short Film About Killing (1988), each read against scope baselines built from all 462 photographed shots of the film.
Thirty curated active-picture midpoint frames from Skyfall (2012), each read against scope baselines built from all 2,619 photographed shots of the film.
Thirty curated active-picture midpoint frames from In the Mood for Love (2000), each read against scope baselines built from all 438 photographed shots of the film.
Thirty curated active-picture midpoint frames from The Handmaiden (2016, Extended Cut), each read against scope baselines built from all 1,373 photographed shots of the film.
Classical colour-harmony templates fitted again to every measured shot, now across eleven films instead of six. Each shape is shown by the frame that fits it best, the corpus's most forceful legible example of it, and a third example from a film neither of those used.
Blade Runner 2049, Sicario, Skyfall and 1917 share a cinematographer and a show LUT and almost nothing you can see. Forty frames of how unalike they are, and the single measurement — agreement at the top of the tone scale — on which all four behave the same and the rest of the corpus does not.
Thirty curated active-picture midpoint frames from Oppenheimer (2023), each read against scope baselines built from all 3,460 photographed shots of the film — a corpus that spans both of the film's photographic strands.
Thirty curated active-picture midpoint frames from Blade Runner 2049 (2017), each read against scope baselines built from all 1,901 photographed shots of the film.
Blade Runner 2049, Sicario, Skyfall and 1917 pooled into a single 8,975-frame corpus, and thirty frames climbing from the darkest of the four to the brightest — every scope read against the body of work rather than against the film the frame came from.
Thirty curated active-picture midpoint frames from Sicario (2015), each read against scope baselines built from all 1,205 photographed shots of the film.
Thirty curated active-picture frames from 1917 (2019), each read against scope baselines built from all 3,250 photographed samples of the film — a corpus taken on a clock, because the film barely cuts.
Complementary, split-complementary, analogous, triadic, monochromatic and achromatic — each illustrated by the shot that fits it best out of six films, shown as frame, hue fit and vectorscope, and set beside the corpus's most forceful legible example of the same shape.
What luma, L*, chroma, saturation and warm balance actually mean; how to read a waveform, a parade, a vectorscope, a hue compass and a grade fingerprint; and where each of them stops being evidence.
Thirty curated active-picture midpoint frames from Zatoichi (2003), each read against scope baselines built from all 945 photographed shots of the film.
Thirty curated active-picture midpoint frames from The City of Lost Children, each read against scope baselines built from all 1,252 shots of the film.
May I count your visit? Google Analytics would tell me which pages get read.
Nothing loads from Google unless you say yes, and you can change your mind
from the footer.