Loading ...

Yolanda Zappaterra previously covered the generative approach to design— and the expanding definition of what it means to be a designer—in an article titled “Generative design: Redefining the designer,” published in Communication Arts’ 2017 July/August issue. “Unlike machines, humans have preconceived notions about how things should look and perform,” she wrote. “Yet more designers are realizing thousands of design directions and solutions that we would almost certainly never envision on our own.” What does the work behind such solutions look like? To find out, Zappaterra explores three recent projects that have been enriched with data, algorithms or code, revealing how designers are continuing to experiment with generative design tools and redefine the creative process.

London-based studio MuirMcNeil devised a variable data system to create a unique cover design for each of the 68,000 publications in the 2019 London College of Music Examinations guitar handbook series. Readers can easily tell that the three covers shown above are for the same instrument and grade level—electric guitar, grade 4—through the consistent color palette and the degree of complexity and detail of the geometric design.

London College of Music Examinations 2019 guitar handbook covers, by MuirMcNeil
In August 2017, London-based graphic design studio MuirMcNeil caused quite a stir when it created 8,000 unique covers for Eye magazine’s annual typography special issue. Last year, the studio, helmed by Hamish Muir and typographic designer Paul McNeil, decided to go a step further. A very large step further. “We were approached by the London College of Music (LCM) in late 2017/early 2018. They had already embarked on a major revamp of their music examination handbook covers, and were keen to make use of the type of variable data printing we’d used to create the Eye covers for a coordinated instrument series,” says Muir.

For those 8,000 covers, MuirMcNeil had combined different fonts in three layers to generate around 8,500 letter combinations across ten visual elds, or “seed” patterns. For the LCM guitar handbook covers, it was dealing with a much more complex project: six instruments, from bass guitar to ukulele, with up to ten levels of difficulty for each instrument—so, 54 titles in total, and 68,000 publications in the series. “There had to be a clearly visible differentiation across six instruments—six discrete sets of covers that were ‘all the same, but all different’ within the series,” says Muir.

Using “FontLab + InDesign + a calculator + trial + error,” MuirMcNeil set about devising a variable data system to realize its idea, using typefaces from its TwoPoint, TwoPlus and TwoBit type systems to create a series of six multilayered visual fields, or “seed” compositions, sections of which were then zoomed, cropped and color adjusted. The type composition for each instrument uses a consistent color system, modified at the output stage with a color filter overlay unique to each instrument group. To represent the different skill levels in the series, the seed compositions were scaled in calibrated increments, zooming deeper into the artworks for the early levels and revealing more of the textual basis of each composition for the higher levels. Within the parameters of that hierarchical structure, the covers were generated randomly. The final result is a coherent yet completely unique cover design for each of the 68,000 handbooks in the series.

“For us, there are many benefits to working in this way,” says Muir. “There is the thrill of seeing unexpected features that challenge one’s aesthetic and compositional sensibilities. There is the possibility of avoiding ‘ideas.’ There’s the joy of planting small seeds that grow unpredictably into myriad forms. There’s the fascination of seeing the ‘rules of the game’ being played out autonomously and automatically. There’s a refreshing sense of detachment from the results—you have to let go of individual intentions, which is very good for the ego.”

The letters SSE, part of Sydney-based design agency For The People’s branding for the Sydney School of Entrepreneurship, are shaped automatically by different data streams.

Sydney School of Entrepreneurship identity, by For The People
When Sydney design agency For The People was commissioned to create the branding for the Sydney School of Entrepreneurship (SSE), a new collaboration bringing together the best student entrepreneurs from twelve New South Wales (NSW) educational institutions, it took up the opportunity with enthusiasm.

Charged with reflecting the diversity, energy and purpose of SSE, For The People began exploring how different data streams could inform the outcome of the identity. The three streams that the agency developed center around the school, its students and the economy; for example, “school” draws on the physicality and day-to-day activities of the school, such as enrollment, the number of courses and social media mentions, and “economy” draws on publicly available data on the state of the NSW economy. These data streams inform how the letters SSE, which make up the school logo, will look by tweaking things like the thickness of the letters and the graphic elements making up the letters. “So, in essence, the logo is a record of how the economy is tracking over time and how the school’s impact might directly affect that. It’s a 20-plus-year experiment, as the whole point of the school is to build up Australia’s entrepreneurial future and help generate future successful businesses that help build a stronger economy,” says Jason Little, executive creative director and cofounder of the agency.

The result is a changing SSE identity, which, on any given day, will have slightly different characteristics shaped automatically by fluctuations in economic activity and the impact of the school. The process of developing the system constantly shifted and was affected by the process itself, Little recalls. “The build and construction were throwing out some unexpected outcomes, which then played back into the development. So the process was informed by the outputs we were getting, and we adjusted and evolved the system accordingly,” he says. “As for the data feeds themselves, we were essentially building a program that could take in a variety of data and generate a very wide number of outputs using a number of constraints within minimal and maximal values. Then, tweaking as we went to make it a feasible identity that could also be usable.”

The resulting brand and experience, pushed out across digital, communications and environmental media, is a vibrant, dynamic piece of generative design that constantly reflects the school, its students and its evolution. Using generative design on such projects, Little enthuses, “is certainly helping us get to more potential outcomes faster and then choose from those. It’s just a question of [how much] time [will pass] before much of what we do is replaced. Algorithms are now—and will continue to become—essential tools for us to do our jobs more efficiently and more effectively. Time is our most valuable asset, and the ability to off-load certain tasks is a natural progression.”

Three posters for the 2018 edition of the Festival Archipel, part of the visual identity by WePlayDesign. The Swiss design studio developed an algorithm that created myriad poster variations. Each of the 500 posters that WePlayDesign ultimately chose is completely unique. Cédric Rossel/Sophie Rubin, art directors; Cédric Rossel, type designer.

Festival Archipel posters, by WePlayDesign
Swiss design studio WePlayDesign has a strong track record of creating award-winning identities and posters for music festivals. One such recent project was for the 2018 edition of the Festival Archipel, a series of events and concerts in Geneva, Switzerland. “At the heart of this edition, called Ecce Robo, was the questioning of the relationship between the human and the machine in musical creation,” say the agency’s cofounders and art directors, Sophie Rubin and Cédric Rossel.

With an earlier poster series (for Les Concerts du Mercredi 2017–18), WePlayDesign had developed a generative design approach, which culminated in a series of unique posters and programs. With the Festival Archipel project, the agency wanted to continue working in this generative way, reflecting the theme of the festival by literally questioning the process of creation between man and machine in graphic design. Having learned during the Mercredi project that well-framed randomness can give good results and meet the needs of a communication project in unique ways, Rubin and Rossel felt confident in going further with the Festival Archipel work, delegating more tasks—including the creation of images and more freedom around the layout—to the computer.

They began by imagining and defining the initial functions of the algorithm, and then developed it as a parameterizable tool, using a text and code editor for script writing and code development. “It was a more complex project than anything we’d done before because the number of possible combinations, both in the creation of images and in the layout, was far greater than anything we had created before; more permutations than we ever imagined. We were surprised by the posters composed by the algorithm, and our role became that of a curator who has to make choices to arrive at the 500 posters that make up the series,” Rubin and Rossel say.

For them, the balance of man and machine is a fascinating one that intrinsically changes the role of the designer. “As designers, we are used to answering a problem and quite quickly considering the visual form that answer will take. But because the machine performs some of the tasks usually determined and controlled by the designer, writing an algorithm requires thinking and organizing the creative process differently,” they say. “This amounts to defining at the outset the creation process and accepting that you will lose some control over the randomness with which the machine will execute your orders.”

The problems inherent in such a way of designing make the process a new and exciting one. “Too precise a script will result in a surprise-free result and automatically undermine any creative interest,” Rubin and Rossel say. “Conversely, a script that is too inaccurate will generate visuals that will surely be surprising, but will probably not meet the needs of a communication project. The difficulty lies in the balance of the space left to the machine and that left to the designer.” ca

Yolanda Zappaterra (yolandazappaterra.wordpress.com) is a London-based writer and blogger. She writes about architecture, design, fine art, photography, food and travel for a range of European publications, including Time Out and Blueprint. She has written five books on editorial design and illustration, and is currently writing an architecture book, Skylines, about the world's 50 greatest city skylines.


With a free Commarts account, you can enjoy 50% more free content
Create an Account
Get a subscription and have unlimited access
Already a subscriber or have a Commarts account?
Sign In

Get a subscription and have unlimited access
Already a subscriber?
Sign In