Why Film Shadows Go Cyan — And Why It Took Me a Long Time to Understand
There’s a quality of film that I have spent years trying to understand. There’s an element of beauty and naturalism inherent to the photochemical process: the way grain is organically embedded within the silver halides rather than overlaid as a digital texture, or how dye density builds subtractively rather than adding light onto the screen.
Over the years, I continued trying to replicate looks from various films, either for commercial projects or personal practice. However, there was a ceiling I kept reaching. I couldn’t understand how to truly emulate these remarkable, elusive qualities of film.
The Orange Mask and the Mystery of the Print
Every colour negative film ever made for cinema has a built-in orange cast. Hold a roll of unexposed, developed 35mm colour negative up to a light and you'll see it immediately - that characteristic amber-orange base.
The cyan, magenta, and yellow dye layers in a colour negative have what chemists call unwanted secondary absorptions. They don't only absorb their target wavelengths. The cyan dye, which is supposed to absorb only red light, accidentally absorbs some green and blue. To compensate for these chemical imperfections, Kodak and Fujifilm built coloured couplers into the negative emulsion itself. This is the orange mask: a physical, analog look-up table designed to pre-correct for dye imperfections before the negative ever hits a printer.
But when that negative is projected onto a positive print stock like Kodak 2383, a second chemical reaction occurs. Print stock doesn't possess a neutral, linear black. To make blacks appear deeper and richer to the human eye, film manufacturers engineered the print stock's D-Max to be non-uniform.
In the deepest shadows of a Kodak print, the red-sensitive layer holds onto its density, meaning a rich amount of cyan dye remains firmly on the cell. It isn't a stylistic choice made in a grading suite; it is a physical consequence of print stock chemistry. Every time you see that characteristic deep teal in the shadows of a film print — in Interstellar (2014) or The Dark Knight (2008) - you're witnessing the deliberate limits of analog chemistry.
Warm Highlights Are Equally Mechanical
The warm, creamy quality of film highlights is equally mechanical in its genesis, born from a phenomenon called differential dye layer behavior.
A positive print stock has three distinct chemical layers. As exposure increases from the negative, the dyes in the print stock begin to clear out to let the projector light through. The blue-sensitive layer — the one responsible for forming yellow dye — behaves differently than the red and green layers. In the highlight zones just below clipping, the yellow dye thins out at a different rate, allowing a beautiful, golden-amber warmth to bleed through the mid-highs before the image reaches total white.
This is why film highlights have that particular quality that's so hard to replicate digitally. It's not a uniform warmth applied via primary or HDR adjustments. It is warmth that emerges dynamically from a physical process — layers compressing and colour crosstalk that are out of sync with one another.
Reimagining the Negative: Moving to Linear
Understanding these mechanisms changed my entire approach to look development. Most digital film emulations are built aesthetically. A colourist looks at a frame, decides it looks nice, and maneuvers their wheels until it slightly resembles the reference. While this can yield some good results, it breaks down instantly when the input camera changes, or when a scene transitions from a tungsten interior to a daylight exterior. An aesthetic approach can't tell you why the math is changing.
To solve this, I had to stop treating film emulation as an artistic grading exercise and start treating it as a digital lab.
I set about completely reimagining how I handle the digital negative. I realised that copying raw characteristic curves into standard DaVinci Resolve curves failed because Resolve’s native tools use human-centric Bézier math, not chemical math. To truly simulate how light hits a negative, I had to move my pipeline into a Scene-Referred, Linear Light workspace and order my nodes methodically so as to emulate the entire negative process. By operating in a Scene-Referred Linear space, the digital data mimics raw photons hitting a physical sensor. From there, I bypassed Resolve's standard sliders entirely and began writing a custom DCTL which now slots into my negative pipeline. With the DCTL, I’ve combined colour cross-talk and even added the naturally occurring D-min Base + Fog for hazy shadows. If the digital "green" layer receives a certain amount of light, it mathematically suppresses the "red" layer, perfectly replicating the secondary dye absorptions of stocks like the Eastman 5254 used on one of my favourite films, Chinatown (1974).
Below is a visual representation of how exactly the dye layers interact with the negative even before a positive print stock.
In an additive digital pipeline, creating vibrant colour often means adding light, which can make a frame feel artificially thin. By contrast, true film emulation relies on subtractive density. As shown in the pipeline tests, my custom DCTL doesn't boost RGB values to find saturation. Instead, the subtractive density calculations simulate physical dye layers weaving together - increasing density exactly where colours overlap to compress the channels organically. As an example, notice how in the top right of the waveform in the second image, the blue/cyan information in the highlights maintain a strong, solid line while the red and green channels are carefully maneuvered into place. It’s not adding light, it’s weaving density into specific zones so that the image looks structured.
Now, below are two additional stills of the same image. Applied is a Kodak2383 positive print - the first image has no dye layer cross-talk whereas the second image has.
Notice the depth of colour that emerges through subtractive density alone. No exposure shifts, primaries, or secondary grading adjustments.The print stock isn't just grading the image; it is reacting to the chemistry underneath in the negative.
By separating the pipeline into a mathematically pure Linear negative zone, followed by a non-linear positive print profile, I finally broke through that digital ceiling. I can now offer clients an emulation that doesn't just look like film, but behaves like chemistry. It has elevated each frame and the pipeline is instantly adaptable to any camera manufacturer.