COMPUTATIONAL FABRICATION

Jyrus CadmanFinal Project Proposal

Final Project Proposal – Jyrus Cadman

For my final project, I aim to create a tangible data visualization that reflects the most frequently played keys on a piano. Specifically, I will produce individual keys with varying depths based on how often each key is played. The depth of each key will resemble a “crater” effect, symbolizing frequency intensity—keys that are more frequently played will have deeper craters, while less frequently played keys will be shallower.

Once each key is printed, I will assemble them into a unified piano model, creating a tactile and visually striking representation of note usage. This project will combine my skills in Rhino, Grasshopper, and Python to accomplish a meaningful data physicalization, transforming abstract musical data into an engaging and interpretable 3D object.

Deliverables

  1. Data-Driven 3D-Printed Piano Keys: Each key will have a unique depth based on play frequency, reflecting the data-driven concept.
  2. Assembled Piano Model: The keys will be assembled into a single, cohesive model, offering a holistic view of the frequency distribution.
  3. Digital Documentation: A supplementary digital model in Grasshopper or Rhino showcasing the frequency dataset and how it correlates to each key’s depth, for reference and comparison.

Project Timeline and Milestones (Tentative)

Week 1: Data Collection and Preparation

  • Goal: Obtain relevant data on piano key frequencies, focusing on MIDI files from popular music genres and classical compositions.
  • Tasks:
    • Download the MAESTRO dataset from Google’s Magenta project, focusing on a subset of compositions for efficiency.
    • Process the data in Python using pretty_midi to calculate the frequency of each note across selected compositions.
  • Milestone: Generate a note frequency dataset that can be mapped to individual piano keys.

Week 2: Modeling and Test Prints

  • Goal: Begin translating data into 3D form.
  • Tasks:
    • Use Rhino and Grasshopper to create parametric designs for each piano key, assigning depth values to each based on the data from Week 1.
    • Run a series of test prints to evaluate the depth representation, ensuring that the cratering effect accurately reflects frequency differences and remains structurally sound.
  • Milestone: Complete and approve test prints with desired depth precision.

Week 3: Final Prints, Assembly, and Documentation

  • Goal: Print, assemble, and document the final model.
  • Tasks:
    • Print all keys, ensuring consistency in depth variations.
    • Assemble the printed keys into a single model and document the outcome.
    • Prepare digital documentation that illustrates the frequency dataset and maps it to the physical keys.
  • Milestone: Submit the assembled model and digital documentation for grading and display.

Related Work

  1. Data Physicalization: A New Frontier for Data Science by Yvonne Jansen, Pierre Dragicevic, and Sheelagh Carpendale
    This paper explores methods for transforming abstract data into tangible, physical forms, offering insights into the field of data physicalization. Jansen’s work emphasizes how physical properties, like depth, can encode information, which is directly applicable to my project. This concept will inform how I use “cratering” to signify frequency, making the piano keys not just functional parts but interpretable data artifacts.
  2. Tangible Bits: Towards Seamless Interfaces between People, Bits and Atoms by Hiroshi Ishii and Brygg Ullmer
    Ishii’s work on “tangible bits” emphasizes creating seamless, interactive interfaces between the digital and physical. By designing objects that communicate data through touch and sight, I hope to create a similar interaction in my final project, where each piano key’s depth will visually and tactilely express musical frequency. This approach directly inspires the project’s interactive design, adding depth both conceptually and literally to the visualization.
    Link to paper on ACM Digital Library

Summary

My goal with this project is to blend music, data science, and computational fabrication to create an insightful and interactive representation of piano key frequency. By focusing on creating a model with meaningful tactile qualities, I hope to convey musical data in an innovative, physically engaging way. This project will certainly challenge my skills in parametric design, data handling, and 3D printing, ultimately producing a unique visualization of sound.

8 thoughts on “Final Project Proposal – Jyrus Cadman”

  1. Hi Jyrus,
    I cannot wait to see your completed project, it sounds like its going to be very unique! Do you plan on just using PLA filament or do you have any ideas on another type of material? Also, out of curiosity do you have any specific genres you are going to focus more on?

    1. Hi Andrea,

      For now, I plan on using the same PLA filament I’ve been using all semester. Although, I may switch to using white instead of red to match an actual piano. Also, your second question has me thinking if I should focus on a specific genre or song(s), and if I should create models of my favorites instead of a full blown piano model. I’ve been listening to a ton of Fred again.. and my favorite songs of his have some of the most refreshing piano chords I’ve heard in the EDM genre.

  2. Hi Jyrus，
    Your project looks very interesting. Will your final print resemble a piano? Where the depth of each piano grid depends on how often it is struck?

    1. Hi Qinghong,

      Thanks! For now, I want the final print to fit like a puzzle that resembles a nicely scaled piano. Recently, I’ve realized that this could cause issues since I have a small printer. So I may just focus on printing a smaller model, or octaves of a large piano. And yes, I want each key to have a depth sort of feel that you could imagine would form if you struck a malleable piano key over hundreds of reps.

  3. Hello Jyrus,

    I really liked how thorough your planning is. I think the idea is awesome and will look super cool once completed. Are you planning on designing a piano for the keys to fit into or are you going to assemble them in a different way?

    1. Hello Andrew,

      Thanks! I’m hoping my plan has all the right pieces and I don’t have to pivot late in the process. With that said, I’m still thinking of designing a full-scale piano piece-by-piece, octave-by-octave, and then assembling it all together once all the pieces are printed. If that doesn’t pan out how I want to with the first iterations, then I will just proceed with downscaling the final model or printing octaves of keys from my favorite songs. There is much to think about. Thanks for keeping me on my toes!

  4. Hi Jyrus,

    This is a neat idea, and I think you’ve put together a very nicely detailed schedule for it. Are you planning on counting note frequency by the time played (such that an 1/4 note would be more strongly weighted than an 1/8th note), or by the times pressed?

    Looking forward to seeing your final project!

    1. Hi Louis,

      Thanks! You and Andrew both seem to find my schedule nicely laid out. Like I said, I’m hoping it can be followed as closely as possible. Also, great question! I did not consider that when planning. And now that I think about it, I would probably weight a 1/8th note more heavily and have it result in a greater depth than a 1/4 note. But my thought process is that I will keep some sort of sum value and use it to determine the final depth for a key.

Leave a Reply