Michele Fumagalli, 24, “originally from a small town between Lodi and Crema” and a student at the master’s degree in Design for the Fashion System, declares himself “passionate about knitwear and storytelling” where, as an attentive storyteller, “in short, fashion is studied in its entirety as an industrial and productive system inserted in the territory and its relationship with topics such as sustainability and digitalization”. A choice to undertake this path that is a happy mix of passion and awareness. “After my first degree, I undertook a work experience in production, which made my goals and ambitions much clearer. I realized that only this faculty and the Polytechnic of Milan could give me all the contacts, skills and means necessary to achieve what I set out to do. After graduation, I would like to enter the world of industrial knitwear research, in particular the streamlining of processes in terms of sustainability”. Michele does not only show clear ideas about his future, but also in the projects he is involved in, such as, for example, the creation of the tapestries that will decorate the Tollegno 1900 stand at Pitti Filati.
In practice, how did you proceed to create them?
It is not an easy question to answer, but I will try. First, I created “packages” containing a series of complex instructions compressed into a palette of codes to be used for rapid programming. For example, I created the package that corresponds to all the commands that the machine must execute to make a stitch of a pre-established color while the other colors are effectively blocked on the back. Once everything was working, together with some members of the Tollegno 1900 staff and my classmate Federica Coti Zelati, we chose the combinations from the vast Harmony range and then enhanced it with 5 vibrant palettes at the base of the tapestries. Federica then created a range of evocative images, which were passed through my exclusive color compression algorithm. This algorithm ensures that, from the millions of possible colors that a pixel can have, 5 are synthesized, thus making the image perfectly convertible to a knitting program.
Did this algorithm have other merits?
Combined with the “yarnbank” database, it proved extremely useful for providing small real-time renderings of how the palettes would look. Once the images, palettes and yarns that would be used had been decided, I created a knitting rendering test with the Tollegno 1900 material to establish the correct number of needles and rows to obtain a product with the expected measurements.
These tests are also useful for other things.
Exactly. These types of tests are also done to check the graphic rendering and consistency of the yarn. The Harmony yarn immediately proved to be extraordinary, not only in color rendering, but also in resistance and workability, guaranteeing a perfect result and above all almost eliminating unexpected events.
Did the last phase of the process require further “optimization”?
Yes, I proceeded to give the finishing touches to the images so that they were the correct size and took into account the natural deformation of a jacquard. The goal is for them to be proportionate once stained. Finally, I converted them into a knitting program using the packages created previously.
Once you have the programs, what happens?
We proceed with the most satisfying part of the whole project: putting the threads in the machine and giving life to all the work we have done, finally creating our tapestries.
How does the program you created differ from the “classic” Shima one?
I am not an expert programmer and therefore I do not think I have done anything particularly new. The real challenge was to vary the way in which the threads are worked on the back of the cloth, minimizing the stretching of the image without losing the consistency of the final cloth. Rethinking the arrangement of the thread guides was certainly valuable and impactful. The solution adopted practically allowed us to eliminate the empty runs that the machine usually makes in the default programs. This allowed us to cut the time needed for knitting by at least 3/4 compared to the traditional technique. Even in this case, however, I would say, it is a basic improvement that has always been made.
An important aspect also concerns the colors: how were those used on the tapestries obtained?
The color compression is undoubtedly the most interesting part. Knitting and digital images are perfectly superimposable, as both speak in “boxes” that can assume certain values, whether they are pixels with colors or meshes with stitches. The algorithm I created, in addition to allowing me to make adjustments “on the fly” without having to redo everything, was based on some principles borrowed from pixel art. In short: thanks to a particular type of pointillism developed by me, it is possible to create the illusion of colors and details much more complex than traditional compression, when in reality we are talking about 5 colors and a definition worthy of an old Atari. This technique only works if certain conditions of the yarn are respected and Harmony has proven to be perfect for the purpose.
In terms of the future, does the program developed have further room for improvement in your opinion?
Definitely, as beautiful as it is to enjoy the spectacle of vibrant images, the entire system I created has enormous defects. The first is that, at the moment, this system can only be used for blankets, tapestries or carpets. The next step is to actually create garments or other dropped products, which is not at all a given with the jacquard system. The second problem is that my algorithm does not differentiate colors but only grays. While traditional applications are incredible at distinguishing colors from each other, my algorithm is only capable of understanding how dark or light they are, therefore having enormous difficulties with images with little contrast.
Including the margins for improvement, are you satisfied with this experience?
Let’s say that these tapestries represent for me a snapshot of my current results in my journey into knitwear, a good result but not a finish line, there is so much to explore and learn. This system that I have developed can only be improved and I can’t wait to do so.
