Back to browse
I built an audiobook player that syncs with your physical book

I built an audiobook player that syncs with your physical book

by arcadianalpaca·Mar 18, 2026·1 point·2 comments

AI Analysis

●●●BangerCozyBig Brain

Page Sync uses on-device OCR to jump audio to your physical book's page.

Strengths
  • On-device speech-to-text and OCR means zero privacy leaks or cloud costs.
  • Fuzzy matching handles OCR and transcription errors gracefully with Levenshtein distance.
  • Nested collections and per-listen tracking solve real library organization pain points.
Weaknesses
  • Android-only limits reach; iOS users are out of luck.
  • Transcription delay for long books might frustrate users wanting instant playback.
Category
Target Audience

Android users who listen to audiobooks and read physical books

Similar To

Smart Audiobook Player · Spotify · Audible

Post Description

Hey HN! I recently switched to Android after 15 years as an iPhone user. What I missed the most after switching was my old audiobook player, and I couldn't find one I liked. So I decided to build one.

It's called Earleaf, and lets you play your local audibook files.

The feature I'm most excited about is Page Sync. You take a photo of a page in your physical book (or e-book) and the app finds that position in the audiobook and jumps to it. It works by transcribing the book on-device, and when you take a photo, it uses ML Kit to do the OCR scanning, and matches the text against the transcription. The initial transcription takes a while, but that's a one-time thing, and when it's in place, the actual search takes about two seconds on my device. And everything runs locally.

Aside from Page Sync, it's a full-featured player with nested collections, independent listen-through tracking, and a pretty good statistics feature. There are no accounts, no internet required, no ads.

Happy to answer questions about the app and its features!

Similar Projects

Security●●Solid

SentryRF – A private, local-first Android app to detect hidden trackers

Uses RSSI time-series and sensor fusion — BLE plus magnetometer and ambient-light hints — to separate a 'passing' AirTag from a likely 'following' device, and then guides you to it with an audio 'Geiger counter' beacon. Smart, pragmatic feature choices (timeline, background protection, strictly local processing) make this actually useful in tense situations; caveat is hardware limits — claims about finding truly 'dead' GPS units or infrared cameras will depend heavily on your phone's sensors and will produce edge cases and false positives.

WizardrySolve My Problem
vidoluc
253mo ago