Sense to Sense Publication Website

Sense to Sense was a writing class about the five senses, taught at Ultralight School by Laurel Schwulst and Meg Miller.

This website complements the printed book, layering sound, image, and text. In the editors' words, the layering "added tactility to our digital class ('touch from a distance')." I worked on the interaction and sound design.

Check out the live website at: https://s2s.ultralight.school/#

Role

Interaction Design

Sound Design

Timeline

Summer 2026

4 weeks

4 weeks

Tools

Figma

RX IZotope

Logic Pro

Team

2 Designers

3 Developers

1 Art Director

Challenge

Translating art direction to a scalable system of interactions

Our publication had 15 pieces, each with its own audio and print layout, and we had 4 weeks to ship. I owned layout design and prototypes for interaction with 3 main goals.

1

Keep the art direction's tactile, playful feel

We wanted written pieces layering dynamically, inspired by texture & wonkiness of real paper.

2

Easy for readers to navigate

Usability and reading experience was priority for visitors to read pieces as a stand alone website.

3

Buildable for developers in a tight deadline

Hand-coding with only 4 weeks to ship. Communicate with developers to keep scope technically feasible.

Interaction Design

How might we let readers layer pieces into a soundscape, without losing track of which sound belongs to which piece?

Each piece came with its own audio, and we wanted readers to build their own soundscape by layering pieces together. Clicking a vessel (an illustrated icon for each piece) opened its text and added its sound to the mix. The challenge was making this feel intuitive.

Iteration 1 — Pieces slide in from right

Following the wireframe handoff from the art director, I prototyped an interaction where clicking a vessel slid its piece in from the right and added its audio to the mix.

Where it broke

As more pieces were open, the stacked pages covered the vessels, which were the controls users needed to keep exploring. This interaction worked well with 5 or fewer pieces, but not for all 15 pieces.

Iteration 2 — Pieces slide in from bottom

What if pieces slid in from the bottom instead? That would keep the vessels uncovered.

Where it broke

Our developers flagged that scroll behavior would be unpredictable with scrollable pages stacked on top of each other. The reading space was also compromised.

Final Iteration — Split screen, separating listening from reading

Iterations 1 and 2 both failed because each piece opened its text and its audio together.

|

|

Key Insight

We realized that listening and reading don't work the same way.

You can hear several sounds at once, but you can only read one.

Solution

We split them into two sides. Linked hover states connect the two sides, so readers always know which sound belongs to which piece

Left — vessel illustrations act as audio mixer

Right — select title to read one piece at a time

Why it worked

Instead of keeping every piece open, the modal puts one piece in the spotlight at a time. Readers still layer sounds freely, the vessels are never covered, and only one page scrolls at a time, which also fixed the scroll issue.

Why it worked

Instead of keeping every piece open, the modal puts one piece in the spotlight at a time. Readers still layer sounds freely, the vessels are never covered, and only one page scrolls at a time, which also fixed the scroll issue.

Design System

One template instead of 15 custom layouts

The print team designed a unique spread for each of the 15 pieces. With only 4 weeks to build the site, I designed one template that developers could reuse across all 15, dropping in each piece's text, image, and audio. The random tilt and paper colors keep the handmade feel of the print spreads.

Our book publication design team created beautiful spreads for each piece with their own unique layouts and visual assets. With a tight deadline, however, we needed something more efficient for the web version. I designed a scalable template that would be used across all 15 pieces to make it easier for developers to build.

Adapting for mobile

Removed layering for audio

Layering depended on the split screen and on hover states, and phones have neither. Instead of squeezing the desktop layout onto a small screen, we simplified it into a single list with each vessel next to its title, and one piece's audio playing at a time.

Sound

Layering 15 DIY field recordings with varying quality

Each author recorded an ambient field recording of the place where they'd want their piece read, all on iPhone Voice Memos. Each one had its own background noise, clippings, and volume, and stacking them without any processing sounded harsh to the ears.

Testing how different combinations layer in Logic Pro

Readers could mix any combination of the 15 sounds, including all of them at once with "play all." Balancing EQ and loudness across every track was important in making sure that no single sound jumps out.

Denoising each field recording using RX IZotope

I removed the hiss and hum from each recording so that, when layered together, they wouldn't build into static.

Takeaways

Managing scope with a tight deadline

This project taught me a lot about managing scope when working with and communicating effectively with a diverse team. Working with a tight publication deadline, I learned to iterate fast alongside developers, as well as being mindful of what features are technically feasible.

Scaling takes planning

Our initial art direction worked effectively with fewer moving parts. Scaling it to 15 pieces required more iterations and testing for edge cases before making sure it was ready for development.

Get in touch ? (๑'ᵕ'๑)⸝*

©2026 April Jung

last updated 09.30.26

Get in touch ? (๑'ᵕ'๑)⸝*

©2026 April Jung

last updated 09.30.26

Get in touch ? (๑'ᵕ'๑)⸝*

©2026 April Jung

last updated 09.30.26