Notes on Notes

This pieces is part of Scent Access Memory, our editorial series with Dirt. Maya’s generative art collection Notes was first released with Bright Moments in Paris in February 2024.

Cinnamon, Dreamwood, Rice Water, Orange Soda, Sweet Pea, Sour Cherry, Dew Drop, Lemon Pie. All of these words, selected at random by an algorithm, feature prominently in my generative art collection Notes, which dives into the vivid language of perfume descriptions. Pulling from lists of fragrance “notes,” product descriptions, and personal user reviews on Fragrantica, an online perfume encyclopedia dedicated to fragrance enthusiasts, Notes remixes found language into a unique visual output—a collaged, fictional fragrance. Each edition in the collection is generated by an algorithm that selects from a variety of color palettes, gradients, typefaces, words, and phrases each time it executes. The resulting outputs are evocative, detailed, absurd descriptions of perfumes that do not actually exist. 

To build the algorithm’s language library, I first web-scraped Fragrantica’s list of notes. I made a few manual cuts here and there, instinctually removing the strangest, least romantic sounding ones like “Aldambre” and “Guaiacol,” but the final array still contains a weighty 1,216 note possibilities. The notes range from familiar vocabulary, such as Cotton Candy and Lilac, to more obscure terms, at least to me, such as Lamprocapnos, Yohimbe, and Chayote. The longer phrases in the collection are sentences remixed through a fill-in-the-blank style algorithm. Using descriptions from Lucky Scent as well as user-commented reviews on Fragrantica as source material (this instagram account collects some of the wildest ones), I broke down phrases such as “literally spring in a bottle” or “smells like fresh laundry on me” into modular pieces that could be mixed and matched to create new, unexpected pairings. That’s how a descriptor like “smells like girly concrete on me” comes about. 

Each color palette featured in Notes is derived from a single perfume advertisement. Famed scent ads analyzed for their dominant set of hex codes include Miss Dior, Victoria Secret’s Bombshell, and Chance by Chanel. I wanted the visual world of Notes to evoke the colorful fantasy of print perfume ads in the glossy magazines I used to flip through as a teenager. Bright hues and slinky gradients. Big glamor and editorial energy. But with the sweetness and innocence of serif, sans-serif, and cursive — three web-safe fonts I love for their overtly internet-native feel. 

For all types of products, online shopping is an exercise in make-believe. We browse through items on our screens while working to contort them into physical objects in our mind. We picture what it might be like to call a collection of pixels ours if we dare “add to cart” and “check out.” But shopping for a fragrance online forces our minds to work overtime. Given the inability to sample scents through our phones or computers, perfumes rely on detailed descriptions to best communicate their energy and persuade potential customers to “blind buy.” Notes recognizes perfume product copy as a unique genre of poetry. Through algorithmic mimicry, the collection requests you to use your imagination. Could this be your new signature scent? Sample it on your skin. My algorithm feels confident that it can deliver you the computer conjured scent of your dreams.