Olfactory family
“Woody” or “floral” describes a perceived character, not a chemically defined ingredient or guaranteed outcome.
AN ATELIER FOR TESTABLE CURIOSITY · 2026
What happens between an idea and an object?
A question becomes a sketch. A sketch becomes a prototype. The prototype resists. The maker listens. Explore three workshops where beauty meets constraints.
The visual sculpture is decorative; every activity below works without it.
01 / THE METHOD
Five stations. One honest loop.
Choose a station to see what a maker can actually do. This is an illustrative method, not a documented chronology of any particular commercial product.
Evidence: illustrative process. Public portfolio context: Sahil Tayal’s atlas (link provided; content not independently verified for this build).
02 / OLFACTORY STUDIO
Compose a scent idea.
Pick an aromatic family for each layer. The result is a moodboard and discussion aid, not a formulation. Colours, note pyramids and descriptions cannot reproduce or reliably predict a smell.
Illustrative simulation · no ingredient amounts, safety clearance or performance prediction.
Materials, claims and limits
IFRA standards
IFRA Standards address restrictions, prohibitions and specifications for certain fragrance materials in specified uses. Actual compliance requires material identity, category, concentration and current standards.
Primary source: IFRA Standards ↗Safety is separate from art
Allergen labelling, local cosmetic rules, stability, quality and safety evaluation are distinct from making an attractive scent concept. This page does not provide a safety assessment.
FDA fragrance overview ↗03 / INTERFACE PHYSICS
Resize the constraints.
A working layout sandbox. Move the width slider and compare fluid grid behaviour; switch motion to see a reduced-motion alternative. This simulates CSS behaviour rather than measuring a real device.
Responsive canvas Wide
Content firstReadable hierarchy
Flexible gridNo fixed columns
Input parityTouch and keyboard
What changes?
CSS container queries switch from three columns to one as the preview narrows. The surrounding page remains stable.
What shaders cost
GPU fragment effects can animate millions of pixels, but require fallback, motion controls and careful battery budgeting. This page uses a small, capped 2D canvas instead.
04 / EDITORIAL ENGINE
Turn an argument into a journey.
Reorder cards with Up/Down controls or drag them. The live essay assembles a coherent structure from your chosen sequence. These are illustrative claims to examine, not research findings.
Your reader’s journey
05 / THE CABINET
Tools are decisions made visible.
⌁
Paper & pencil
Low-fidelity sketches make changing direction cheap. Draw the user’s question before decorating its answer.
CRAFT NOTE · ARGUMENT
◌
Scent blotter
Smell evolves with time and context. A descriptive note is useful, but never a substitute for controlled smelling and safety work.
CRAFT NOTE · GENERAL PRACTICE
⌘
Browser inspector
Inspect layouts at narrow widths, disable JavaScript, navigate by keyboard and test whether the page still makes sense.
CRAFT NOTE · TESTABLE
Three small essays on making
The dignity of the draft
An unfinished draft is not an embarrassment to hide. It is a record of what the maker has learned to notice. The question is whether revision changes the work rather than merely polishing its surface.
Constraint as collaborator
Budget, screen width, materials and regulation are not enemies of creativity. They help distinguish an attractive fantasy from something another person can use, wear or understand.
What iteration proves
A sequence of versions proves effort and changing decisions. It does not, by itself, prove customer benefit, product safety or scientific validity. Those require different evidence.
06 / ITERATION GALLERY
Purposeful revisions, not invented milestones.
Make the idea legible
One clear statement of what a visitor should gain. Success test: can a stranger explain the purpose?
Make it usable
Implement real controls and fallback paths. Success test: can keyboard and touch users complete the same task?
Make it accountable
Separate simulation from evidence, verify links and record known limitations. Success test: what would falsify the claim?
These are illustrative design iterations, not claims about a specific released product.
Made as an independent Sahil Tayal experiment. The central atlas is the master index of projects and writing.