# Ingredients expansion · 2026-09-27

The core recipe index grew from 100 to 105 distinct contributors. All 40 recipes and all 108 original contribution records are preserved; eight checked contribution records were added. The broader directory now has 120 people. The 18 animated artist studies are unchanged.

| Person | Actual recipe placement | Credited role and source |
| --- | --- | --- |
| Golan Levin | Moiré; Dithering as art | Co-author and educator, [Code as Creative Medium examples](https://github.com/golanlevin/exercises). |
| Tega Brain | Moiré; Dithering as art | Co-author and educator, same shared example collection. |
| Gene Kogan | Moiré; Vector fields from math | Artist, [Interference](https://genekogan.com/works/interference/) and [Vertex rivers in his work index](https://genekogan.com/work/). |
| Zach Lieberman | Harmonographs / Lissajous | Educator, [ofBook animation chapter](https://openframeworks.cc/ofBook/chapters/animation.html#lissajous-figures). |
| Lauren Lee McCarthy | fBm + ridged / turbulent | Creator of p5.js, an enabling platform with [octave/falloff controls](https://p5js.org/reference/p5/noiseDetail/); [creator attribution](https://eyebeam.org/artists/lauren-lee-mccarthy/). |

## Credit boundaries

These additions identify education, toolmaking, or an artistic application, not invention of the underlying mathematics. The Code as Creative Medium repository credits assistance from Lingdong Huang, Tatyana Mustakos, Cassie Scheirer, and Xinyi (Joyce) Wang; individual ports must not be attributed solely to the book authors. The recipe descriptions preserve these names. Some repository exercises adapt earlier work, as recorded in individual examples. The two examples linked here are the moiré and random-dot dithering exercises.

McCarthy's credit concerns creating p5.js. It does not claim that she invented fBm, wrote every noise implementation, or implemented all variants in the recipe title. Kogan says the Interference code was lost. His visual work record remains evidence of the artistic application, not a reproducible code archive.

All five have Eyebeam affiliation sources in the linked dataset. Kogan's [2015 record](https://eyebeam.org/artists/lisa-kori-gene-kogan/) is a joint project residency with Lisa Kori, retained explicitly. Shared residency does not establish collaboration among these five, nor Eyebeam sponsorship of this project.

## Portraits and biographies

Five concise bios link to Eyebeam profiles. Two newly included photographs—Lauren Lee McCarthy and Tega Brain—come from images published with their Eyebeam profiles. Original asset URLs and retrieval dates are preserved in `portrait-map.json`. The retrieved profiles did not identify photographers; the site links the source rather than inventing a credit. Third-party image rights remain with their holders; no blanket reuse license is granted.

Golan Levin and Zach Lieberman reuse the collection's existing illustrated portraits, labeled as portrait credits. At this expansion’s initial release, Gene Kogan had a labeled missing portrait. The September 28 portrait pass adds a photograph-based illustration; see PORTRAITS.md for the current assets and sources.

## Language evidence and remaining research

New scoped links cover Kogan's Processing/GLSL shader examples and Brain's shared Processing/JavaScript educational examples. Lieberman's C++ association is upgraded from toolkit-based inference to documented evidence from his authored animation chapter. Four inferred links remain elsewhere; the meter retains that distinction.

Kyle McDonald and Taeyoon Choi remain in the visible placement queue. Their practices and affiliations are documented, but this release does not force a recipe connection. In particular, the existence of a repository fork is not evidence that its owner authored the original algorithm. McDonald's multiscale Turing work is a useful future lead but must not be silently labeled as the Gray–Scott model.

The Eyebeam overview now derives its people and recipe placements from the same datasets as Ingredients. Future additions cannot become stranded in a hand-maintained overview with stale “to be assigned” labels. Data checks preserve the original credits and require sources, bios, checked status, and exact recipe placement for this expansion.