Why Log: the science of color and sound in a promotional filmWhy Log: the science of color and sound in a promotional film

Cinema shows 24 frames per second, and the number has nothing to do with the eye. In the silent era cameras were cranked by hand, and 16 to 18 frames were considered enough. When Warner's Vitaphone system began locking sound to film in 1926, sound became the problem: if the film ran slow the sound sagged, and if it ran fast it turned shrill. The lowest speed at which the optical soundtrack could be read at acceptable quality was 24 frames; it became the standard in 1929 and has not changed in a century. Sound had set the tempo of the image. Every technical decision in making a promotional film is a chain of standards of this kind; whoever does not know the links of the chain may make the film "beautiful" but cannot make it "right."

In the nineteenth century Ernst Weber and Gustav Fechner showed by measurement that the senses respond to a stimulus not linearly but logarithmically; in 1957 Stanley Stevens refined this with a power law: perceived brightness grows roughly with the cube root of the light. Twice the light does not look "twice as bright"; it looks "one step brighter." A digital camera sensor, on the other hand, is linear: it gives twice the number for twice the photons, and it spends most of those numbers on the bright region where the eye cannot tell the difference.

The log curve corrects this waste. In 1993 Kodak encoded film scans logarithmically with its Cineon system; Sony introduced S-Log in 2007, and Apple brought Apple Log with the iPhone 15 Pro in 2023. Log distributes the sensor's 12 to 15 stops of dynamic range according to the steps the eye can distinguish; that is why the recording looks washed out and gray on screen. The washed-out image is not an error but a deliberate way of storing the picture; at the color grading stage it is converted to the display's own curve with a lookup table (LUT). Whoever shoots in log and does not grade delivers a film half finished.

In a film camera the shutter is a rotating disk; if the open part of the disk is 180 degrees, each frame is exposed for half the frame's duration: at 24 frames, a 48th of a second. This ratio matches the natural blur the human eye produces as it perceives motion; that is why the "180-degree rule" has been cinema's motion signature for a century. In 1998 Steven Spielberg dropped the angle to 45 and 90 degrees for the landing sequence of Saving Private Ryan; the sharpened, jittery motion threw the viewer into the battle. The rule, once broken, proved what it was for.

Frame rate is a perceptual decision as well. When Peter Jackson screened The Hobbit at 48 frames per second in 2012, a sizable share of the audience found the image "too real," "like a television series"; the smooth motion was breaking cinema's century-old dream texture. In a promotional film, 24 or 25 frames and 180 degrees are not a technical habit but the source of the viewer's sense that "I am watching a film."

The color space of television and web video was defined in 1990 by the International Telecommunication Union's BT.709 recommendation; the display curve was fixed at gamma 2.4 in 2011 with BT.1886. Movie theaters use the wider DCI-P3, and ultra-high-definition broadcasting uses BT.2020, defined in 2012. Three spaces give three different reds; a brand's corporate color is graded in video not merely so that it "looks the same" but by calculating which value it will land on in which space, and it is checked with ΔE.

Color grading's best-known habit, the "teal and orange" palette, also comes from physiology: human skin carries a tone close to orange across every ethnic background, and on the color wheel the opposite of orange is teal; cool the background and the skin comes forward. On a factory shoot, steel and concrete provide that cool background on their own; the team's faces and the color of the product sit on the orange side.

In the early 1920s the Soviet director Lev Kuleshov cut the expressionless face of the actor Ivan Mozzhukhin after three different images: a bowl of soup, a child in a coffin, a woman on a divan. Viewers saw hunger, grief, and desire in the same face and praised the actor's mastery. The meaning was built not in the face but in the preceding shot. In a promotional film, the face that follows the machine reads as "trust," and the machine that follows the face reads as "the human hand"; the order is the message.

The tempo of editing has been measured too. In 2010, in Psychological Science, James Cutting, Jordan DeLong, and Christine Nothelfer examined 150 Hollywood films from 1935–2005: the average shot length had fallen from ten seconds to under four, and the distribution of shot lengths increasingly resembled the natural fluctuation pattern of human attention (1/f). Films had not merely sped up; they had been tuned to the rhythm of attention. That is why the shot lengths of a three-minute promotional film are cut not to a uniform beat but to a rhythm that breathes.

In 1999 John Beerends and Frank de Caluwe published a crossover experiment in the Journal of the Audio Engineering Society: when the same video was shown with better sound, viewers rated the image as higher quality too; poor sound lowered the image score. Sound is the invisible coefficient of the image's perceived quality. Even so, in promotional films the sound budget is usually left for last.

The most repeated figure about communication is also about sound, and it is wrong: the line "93 percent of communication is nonverbal" was derived from two narrow experiments Albert Mehrabian ran in 1967. Mehrabian had measured emotional inconsistency in single-word utterances; the result belonged only to the question "when word and tone contradict each other, which one is the emotion read from?" and had nothing to do with communication in general; Mehrabian wrote this himself many times over. The correct inference is more modest: when tone and word conflict, the viewer believes the tone. The tone of the voice-over can undo what the script says. There is also a fact known since 2016: on social platforms the great majority of videos are watched without sound; publisher data reported rates of up to 85 percent. Subtitles are not optional; they are a second sound channel.

Commercials coming out louder than the program was a complaint for years; the solution was a unit of measurement. In 2006 the ITU defined, with BS.1770, the LUFS unit for measuring perceived loudness. In 2010 the European Broadcasting Union fixed the broadcast target at −23 LUFS with R128; in America ATSC A/85 said −24 LKFS, and the CALM Act passed in 2010 made it mandatory for commercials. Streaming platforms use higher targets: Spotify and YouTube at −14, Apple Music at around −16. The same film is delivered to television at −23 and to Instagram at −14; a film at the wrong level is either pushed down by the platform or left thin, and the viewer calls it "cheap." Factory noise is not the enemy of this measure but its raw material: machine sound recorded separately and set at the right level is the film's most convincing layer; the tradition of effects recording to which Jack Foley gave his name in the 1920s is exactly this.

In 1982, in the Journal of Marketing, Ronald Milliman varied the tempo of the background music in a supermarket for nine weeks: when slow-tempo music played, shoppers walked the aisles more slowly, and daily sales were 38 percent higher than on fast-tempo days. The music was not being listened to; the body was keeping time with it. Nor is a promotional film's music chosen because it is "pleasant": the beat count sets the shot length of the edit, and the shot length sets the viewer's breathing. The heavy tempo of an industrial film at 80 beats per minute says trust, a fashion brand's 128 beats say energy, and neither can take the other's place.

Our protocol is a chain of standards: we shoot with a log curve (S-Log3 or Apple Log) and with a color chart under the light of the set; the shutter is 180 degrees and the frame rate is 25; color grading is finished at a Rec. 709 / BT.1886 target, with the brand color checked by ΔE; machine and ambient sound are recorded separately, and the music tempo is chosen according to the rhythm of the edit; sound is delivered at −23 LUFS for broadcast and −14 for social; every film ships with a subtitled version and a silent loop for the trade-fair screen. For a company manufacturing in Denizli, the set is already built; our job is to turn that set into a film with a century of standards.

Fechner, G. T. (1860). Elemente der Psychophysik; Stevens, S. S. (1957). On the psychophysical law. Psychological Review, 64(3), 153–181.

Kodak Cineon system (1993); Sony S-Log (2007); Apple Log (2023, iPhone 15 Pro).

ITU-R BT.709 (1990), BT.1886 (2011), BT.2020 (2012); DCI Digital Cinema System Specification (2005).

Kuleshov, L. (1929). Iskusstvo kino [The Art of Cinema]; Kuleshov experiments (1918–1921).

Cutting, J. E., DeLong, J. E. & Nothelfer, C. E. (2010). Attention and the evolution of Hollywood film. Psychological Science, 21(3), 432–439.

Beerends, J. G. & de Caluwe, F. E. (1999). The influence of video quality on perceived audio quality and vice versa. Journal of the Audio Engineering Society, 47(5), 355–362.

Mehrabian, A. & Wiener, M. (1967). Decoding of inconsistent communications. Journal of Personality and Social Psychology, 6(1), 109–114; Mehrabian, A. & Ferris, S. R. (1967). Inference of attitudes from nonverbal communication in two channels. Journal of Consulting Psychology, 31(3), 248–252.

ITU-R BS.1770 (2006); EBU R128 (2010); ATSC A/85 (2009) and CALM Act (2010).

Milliman, R. E. (1982). Using background music to affect the behavior of supermarket shoppers. Journal of Marketing, 46(3), 86–91.

Vitaphone (1926) and the adoption of the 24-frame standard (1929); Saving Private Ryan (1998) shutter angle; The Hobbit (2012) 48-frame screening.