Short answer
Ask which of three things the claim is. A production record says what was done and can be checked against the people who did it. A visual description says how the result looks to someone, on some copy, on some screen. An interpretation says what the colour means. All three are legitimate and they fail in different ways, so Colourwise tags every statement about a named production as one of them and gives its source. Almost nothing about film palettes has been measured.
The firmest kind of claim records an action: the stock, the laboratory process, the gel, the scan resolution, the paint on the set. Its usual sources are trade interviews with the cinematographer, designer or colourist, and case studies from manufacturers and laboratories. Its weaknesses are specific. An interview during production states intent before a print exists. A retrospective given decades later is memory. A manufacturer's case study is accurate about its own product and silent on the rest. And a record of what was done is not a record of how much: no source here gives a degree of desaturation.
A description such as 'brown and black', 'like mercury' or 'more neutral' is a judgement made by looking. It can come from the maker, which carries weight, but it is still not a measurement. The harder problem is the object being looked at. A film exists as original prints, later prints, restorations and several generations of video master, graded by different people for different displays. A frame grab found online adds a capture, a compression and the reader's own screen. Nothing on this site is a calibrated reference. A web page is shown by a browser that may or may not colour-manage, on a screen of unknown white point, brightness and response, so no still or swatch here can reproduce a graded image.
Statements that green is unappealing, that draining colour stands for a loss of meaning, or that an autumnal palette suggests decline are interpretations. When the maker says so, it is good evidence of intention and no evidence that audiences read it that way. When a critic says so, it is a reading. Meanings attached to colours also differ between cultures and periods, so an interpretation offered as universal deserves the most caution. Colourwise records who is speaking for every interpretation and does not restate one as fact.
Very little. The one peer-reviewed measurement Colourwise holds for film is a study of 160 English-language films released between 1935 and 2010, which found mean frame luminance falling over the period. That is brightness, not colourfulness, and the sample spans the change from black-and-white to colour. No published study that Colourwise could find measures saturation or hue across a representative sample of films, so widely repeated statements that modern films are more muted, or more orange and teal, rest on impression and on selected examples.
The table holds 18 productions and 59 statements. A statement is included only if a source that was read directly makes it; where sources disagreed or could not be read, the point was left out, and several well-known titles are absent for that reason. 40 statements are production record, 8 are visual description and 11 are interpretation. Every production rests on at least one production-record statement. The titles are those with a good paper trail, which favours large productions, English-language trade coverage and unusual methods. It is a set of case records, not a sample.
| Production | Route | Production record | Visual description | Interpretation | Confidence |
|---|---|---|---|---|---|
| The Wizard of Oz (1939) | Three-strip Technicolor; Kansas scenes black-and-white, printed sepia | 3 | 1 | 0 | medium |
| The Red Shoes (1948) | Technicolor, photographed by Jack Cardiff | 1 | 1 | 1 | medium |
| The Godfather (1972) | Colour negative, deliberately underexposed and push-processed; photochemical timing | 2 | 1 | 0 | medium |
| Suspiria (1977) | Colour negative; dye-transfer prints by Technicolor Rome | 3 | 0 | 1 | medium |
| Se7en (1995) | Colour negative, pushed and flashed; silver-retention show prints | 2 | 1 | 1 | high |
| Pleasantville (1998) | Colour negative; scanned at 2K; colour removed and restored digitally; film-out | 3 | 0 | 0 | high |
| Saving Private Ryan (1998) | Colour negative, pushed and flashed; ENR prints; altered lenses and shutter | 2 | 1 | 1 | high |
| The Matrix (1999) | Colour negative; filtration, set colour and photochemical timing; some dye-transfer prints | 3 | 0 | 1 | high |
| O Brother, Where Art Thou? (2000) | Colour negative; scanned at 2K; digital grade; film-out | 2 | 0 | 0 | high |
| The Lord of the Rings: The Fellowship of the Ring (2001) | Super 35 colour negative; digitally graded throughout; film prints | 3 | 0 | 1 | high |
| The Aviator (2004) | Film origination; digital intermediate emulating two-colour and three-strip Technicolor | 2 | 0 | 1 | medium |
| Blade Runner 2049 (2017) | Digital capture, open gate; on-set lookup table; digital grade | 2 | 1 | 0 | high |
| The Vault (2017) | Digital capture; ACES colour management through post-production | 2 | 0 | 0 | medium |
| Roma (2018) | Large-format digital colour capture; monochrome made in the grade; SDR and HDR masters | 3 | 0 | 1 | high |
| The Irishman (2019) | Mixed film and digital capture; one emulation lookup table per period | 1 | 1 | 1 | high |
| House of the Dragon (season one) (2022) | Large-format digital capture; single show lookup table; HDR-first grade | 2 | 1 | 0 | high |
| Killers of the Flower Moon (2023) | 35 mm and digital capture; emulation lookup tables by story phase; printer-light dailies | 2 | 0 | 1 | high |
| Oppenheimer (2023) | 65 mm colour and black-and-white negative; photochemical prints | 2 | 0 | 1 | medium |
| Production | Kind of statement | Statement | Voice |
|---|---|---|---|
| The Godfather (1972) | Production record | Gordon Willis rated the colour negative faster than normal and had it pushed one stop, leaving it about half a stop underexposed. He said this left the laboratory one printing range and no room to change the look afterwards. | Production record |
| The Godfather (1972) | Production record | The Sicilian exteriors were shot through heavy chocolate-brown filters, with lighter ones indoors to keep the relationship; the Hollywood scenes were given a brighter, harder, cleaner image. | Production record |
| The Godfather (1972) | Visual description | Willis described the look he wanted as brown and black, like a newspaper photograph with a little colour let in, with an occasional blue-black, and the underexposed negative as foggy and not quite there. | Gordon Willis, interviewed during production |
| Saving Private Ryan (1998) | Production record | Janusz Kamiński used ENR at 70 per cent, pushed the negative a stop, flashed it by about 15 per cent, and replaced the usual 85 correction filter with a half Coral to leave a slight blue cast. | Production record |
| Saving Private Ryan (1998) | Production record | Panavision stripped the coatings from a set of older lenses, and the cameras ran with 45- and 90-degree shutters, one of them deliberately mistimed to streak highlights. | Production record |
| Saving Private Ryan (1998) | Visual description | Kamiński described flatter contrast that was foggy yet still sharp, desaturated colour, and metal and water reflections that looked like mercury under ENR. | Janusz Kamiński |
| Saving Private Ryan (1998) | Interpretation | The muted, slightly blue image was meant to suit overcast weather and the subject of war. | Janusz Kamiński |
| The Irishman (2019) | Production record | Rodrigo Prieto gave each period its own lookup table: a Kodachrome emulation for the 1950s, Ektachrome for the 1960s, a light ENR emulation for the early 1970s and a heavier one after. They were developed with an image scientist and a colour scientist and finished with the colourists so that film and digital footage matched. | Production record |
| The Irishman (2019) | Visual description | Prieto characterised the Kodachrome look as colourful with saturated reds, the Ektachrome look as slightly desaturated with a cyan-green bias in the shadows, and the ENR looks as progressively drained of colour with added contrast. | Rodrigo Prieto |
| The Irishman (2019) | Interpretation | The decades were separated to give the central character an emotional arc, with the draining of colour standing for his loss of meaning late in life. | Rodrigo Prieto |
| Period | Finding | What it does not show |
|---|---|---|
| 1935–2010 | Across 160 films released between 1935 and 2010, mean frame luminance falls over time: films have become darker, alongside shorter shots and more on-screen motion. | Luminance is brightness, not colourfulness, and the sample spans the switch from black-and-white to colour film. It shows darker films, not greyer ones; no comparable peer-reviewed study of film saturation over time was found. |
| 2000s–2020s | The widely discussed 'orange and teal' look and the impression that recent films and series are more muted are not backed by any published measurement of saturation across a representative sample of titles that Colourwise could find. | An absence of published data, not evidence that no change happened. |
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
FactModerate evidence
A study of 160 English-language films released from 1935 to 2010 measured each frame and found mean luminance decreasing over time, alongside shorter shots and more motion.
Caveat: Luminance is brightness, not saturation or hue, and the sample includes the transition from black-and-white to colour.
Source: Quicker, faster, darker: Changes in Hollywood film over 75 years
FactModerate evidence
The same production can be recorded differently at different times: Gordon Willis's account of The Godfather dates from the shoot and describes intent, while Robert Richardson's account of The Aviator was given two decades after release.
Source: On Location with The Godfather: A Discussion with Gordon Willis (American Cinematographer, June 1971); Robert Richardson, ASC: On Color (Iain Marcks, American Cinematographer, 18 November 2024)
FactStrong evidence
A look applied at the print stage may exist on only some copies: on Se7en the silver-retention process was applied to show prints, and on The Matrix 150 prints were dye-transfer while the remainder were on ordinary print stock.
Source: Deluxe Color Contrast Enhancement (1995–c.2012) (Film Atlas entry by Beverly Wood); Flashback: The Matrix (Christopher Probst, American Cinematographer, April 1999)
Colourwise analysisModerate evidence
In Colourwise's productions table, production-record statements outnumber visual descriptions and interpretations combined, and no statement is a measurement.
Based on: Counted from the productions dataset: 40 production-record statements against 8 visual descriptions and 11 interpretations, across 18 productions.
Caveat: A count of Colourwise's own records; it says nothing about writing on film colour at large.
Source: Ring Bearers: The Lord of the Rings: The Fellowship of the Ring (Simon Gray, American Cinematographer, December 2001); Roma: Memories of Mexico (Mark Dillon, American Cinematographer, 7 January 2019)
Colourwise interpretationModerate evidence
A colour palette extracted from frame grabs describes those image files on the software used, not the film: the grabs come from one master among several and have been through capture and compression.
Based on: Follows from the documented existence of differing prints and masters for the same title, and from the absence of controlled viewing conditions for online stills.
Source: Flashback: The Matrix (Christopher Probst, American Cinematographer, April 1999)
Reviewed 6 October 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.