The digital Mexican sepia filter is a long-sought-after visual effect in American cinematography. The effect is simple to imitate, but creating an authentic Mexican refraction index is more complex. Due to a growing demand for more authentic Mexican visual filters by modern audiences, studios have spent additional millions paying actors to film in the middle of the desert of the Chihuahua and Coahuila regions, where the natural Mexican sepia light filter occurs at its maximum state. Using basic digital image processing techniques, this paper will create an authentic Mexican atmosphere in any environment by adjusting the refractive index, dust, and cumin levels to reach a South-of-the-border flavor.

Keywords: Authentic Mexican Sepia Filter, Sonoran Refractive Index, Cumin Dust Spread, Yellow Heat Index, Too Much Spice, Carranzan Mandate of 1918, Image Processing, OpenCV, Image Blurring, Dust Cloud Computational Fluid Dynamics

Without a serious regressive change in automotive emission standards, there is little chance an authentic Mexican sepia filter can be achieved in Hollywood without the assistance of digital image processing. Even then, the oppressive heat index, the difference in the refractive index of light, and the cumin dust cloud make the Mexican atmosphere one of the most unique and iconic in the world [1]. Despite a similar effect when filming in the oppressive heat of yellow-tinted Southern Asia, a Graham-Masala finishing cloud creates an entirely different color tint effect, which savvy viewers will pick out instantly [2]. It is time to develop a more robust and researched method for generating an authentic Mexican color filter using basic functions shamelessly stolen from Stack Overflow!

After the passing of the Mexican constitution in 1917, the new government’s president, Venustiano Carranza, passed a widely beloved law in 1918 to change the refractive index of the air in Mexico in order to instill a sense of national pride. With the new refractive index, longer wavelengths “…would scatter to create a perpetual sunset to signify the diminishing government of Diaz and a sunrise for Mexico’s future…” according to Carranza. This law was intended to consolidate the remaining revolutionary factions under the newly formed government and make it even more difficult for rebel leaders such as Pancho Villa and Zapata to maintain local support for their violent causes [3]. The plan worked, and Mexico soon developed an identity around a yellow-orange tinted color palette that transformed the sky into the Mexico we know today.

Figure 1.1: US-Mexican color filter seen along the Mexican-US border at Big Bend National Park

Not all Mexican yellow filters are the same. Known for its sea-misted ceviche tone, the Baja region tends to diffract more blues into the atmosphere than other regions, in stark contrast to the rest of the North’s predominant reddish-orange desert color tone, reaching as far down as Mexico City. As one travels south out of the desert and into the jungles of Southern Mexico or the Yucatán Peninsula, greener, more colorful light travels further as the increased flora filters the government-mandated haze, adjusting the atmosphere to Carranza’s 1918 law.

Countless Hollywood films and TV classics, such as Breaking Bad, Traffic, and Extraction, aimed to imitate this color filter in film to denote a change of location. However, many critics have criticized, as critics often do, the inauthentic visual effect, leading studios to fork over the funding to refilm on location. Breaking Bad, in particular, required the orangest of colors set in the cartel-ridden Northern desert regions of Mexico close to the New Mexico-Texas border. With TV fans growing more annoying than ever, studios quickly became hostage to their demands.

The transformation of a normal-colored image to the synthetic Mexican atmosphere will be implemented using a three-stage filter. First, a yellow tint filter will be applied, then a yellower dust filter, and then, finally, a cumin dust cloud. Each filter will be applied by transforming the BGR (I don’t know why OpenCV does it that way and not RGB) elements using a 3x3 transformation matrix, averaged with a variable intensity using the Python package OpenCV. Each filter will be demonstrated on the following test image, taken from a normal, non-yellow-tinted city. Additionally, the blur operation allows each image to be synthetically blurred as if there were a dust or cumin cloud.

Figure 1.2: A normal city test image (Mexico City-Carlos Valenzuela, CC BY-SA 4.0, via Wikimedia Commons)

3.1 Heat

Through the combination of the intense heat of the Mexican sun, significantly stronger than anywhere else on the continent, and Carranza’s famously mandated refraction index, even on a clear day, the sky will be noticeably yellower.

Using the preceding transformation matrix along with an intensity of 0.6 to be averaged with the original image, any image may be made authentically more yellow, as shown in Figure 1.3. Comparative physics-based simulation models guided by the 1918 Carranzan Mandate have been verified in [4], showing that this transformation matrix is good enough.

Figure 1.3: Yellow-filtered city

3.2 Dust

Next, to transform each image with a generous dust cloud, but at a level that is still visible, the image color palette is washed out using the following color transformation dust matrix:

With the dust layer applied, the original image is further blurred by a 5x5 convolution matrix in OpenCV and averaged at an intensity of 0.4. We believe these are the correct values because it’s starting to look about right. The new dusty image tricks the viewer into believing that they are now in a more arid environment, reducing the need for further establishing shots.

The results of this filter are shown in Figure 1.4.

Figure 1.4: Dust-filtered city

3.3 Cumin

Next, a fine layer of brown from a special concoction of coriander, cinnamon, and primarily cumin, creates the final transformation matrix, as follows:

As the cloud generated using the computational fluid dynamics (CFD) models from the methodologies in [5] is much more exact, the numerical certainty of this special spice blend color palette transformation is paramount! The image is then further blurred with a 2x2 convolution matrix and then averaged in with an intensity of 0.4. The results are shown in Figure 1.5.

Figure 1.5: Cumin dust cloud-filtered city

For the final test, a new test image needed to be generated using generative AI. IEEE recently banned the use of the famous Lena Forsén test image [7] in their research papers in an attempt to modernize [6]. This means we need a new test image. As we are always looking for opportunities to provide new positive male role models for boys and maintain a connection to our shared image-processing heritage, Lena was fused with John Cena through MidJourney. This generated a new standard test image, The John-Lena, shown in Figure 1.6.

Figure 1.6: John-Lena test image

Through the recent advancements in AI, we could expand the image, something many engineers have always wished they could do. We then finalized the new and improved test image to be adopted by computer science and image-processing journals everywhere. Gaze into the future in Figure 1.7, a bold, refreshing new look for the field. It is John-Lena.

Figure 1.7: Expanded John-Lena

Each successive filter was applied to the John-Lena test image, as shown in Figure 1.8. As shown, through only the yellow refractive filter and the dust filter, John-Lena became far too yellow, which was nothing like an authentic Mexican atmosphere. By applying the final seasoning blend filter at the end, John-Lena now appears as if he were on some adventure south of the border in a hit new TV show or movie.

Figure 1.8: Clockwise from top left: Lena-Cena filtered successively through the yellow refractive filter, then the dust blur filter, then the cumin cloud

Thanks to the new filter, 98% of our test audiences identified this image as having been shot in Mexico without any additional shots for context. The color filter was found to be authentic enough to get the most annoying fans off our backs once again. There appeared to be no need to shoot the film on location in the Sonoran Desert at twice the cost. Now, shoots may occur at Vasquez Rock, as in the Hollywood of old.

Many different filter iterations were generated by experimentally dialing in the filters based on feels and vibes. It was important to create a variety of color palettes for moods and differing atmospheric locations according to heat index, wind conditions, and spiciness.

Those experiments revealed a clear line a filmmaker should not cross while applying the spice filter, as shown in Figure 1.9.

Figure 1.9: Too much spice! HE CONSUMED THE SPICE. THEY DESIRE THE SPICE

If too much spice is applied in the successive filters, images will ingest too much spice, which will cause the Eyes of Ibad phenomenon [8]. By exposing a human to a heavy dose of the spice Melange, even just synthetically, the subject appears to pass the Arrakis threshold, coloring their eyes blue on blue. It is not recommended unless you prefer to symbolize your character’s connection to the land and their enlightenment by communing with their ancestral memories through the Water of Life. Lisan al Gaib.

Finally, through the magic of matrix transformations, OpenCV, and fiddling around with numbers, an effect that normally requires a passport may now be accomplished using a simple Python script. In this paper, not only have we saved Hollywood millions of dollars per year, but we introduced a fantastic new test image for image processors everywhere. It’s impossible to hate John Cena, so we suspect John-Lena will catch on like wildfire.

1. Gilligan Vincent, et al., 2008. The Mexican Yellow Tint Color Palette :: Journal of Modern Western Cinematography
2. Sterling “Silver Screen” Stone, 2014. Adventures in Mood Colors: Synthesizing Color Palettes Around the World :: Journal of Post-Production Techniques, Tips, and Tools of the Trade
3. Veronica VonDutch, 2016. A Brief History of the Mexican Atmospheric Mandates: Revolution Gone too Far :: Annals of Alternative Mexican History
4. Winston “The Wizard” Whitaker, 2020. Computational Models of the Mexican Atmospheric Refractive Index in a Post-Carranzan World :: Journal of Fiesta-Physics
5. Melba McCormick, 2023. Pink and Fire: Computational Fluid Dynamics Simulation Analysis of My Recent Gender Reveal Party Disaster :: Journal of Astrological Big Data Ecology
6. https://www.theguardian.com/technology/2024/mar/31/tech-publisher-bans-playboy-centrefold-test-image-from-its-journals
7. https://en.wikipedia.org/wiki/Lenna: Oh man, if you haven’t read up on this, you gotta: back in the 1970s, to get a test image, an engineer scanned this image from a Playboy magazine. Then it somehow became the standard test image for image processing for decades. It was only banned by IEEE in 2024.
8. https://dune.fandom.com/wiki/Eyes_of_Ibad