Maker compresses a 2.9MB song by 1000x and prints it on paper as eight QR codes — 21KB song is two minutes long, requires a neural network for playback
- ID
- 14599
- Status
- summarized
- Published
- 16 Aug 2026, 10:22 PM
- Fetched
- 17 Aug 2026, 12:10 AM
- Provider
- Tom's Hardware
- Category
- technology
- Original URL
- https://www.tomshardware.com/tech-industry/maker-compresses-a-2-9mb-song-1000-times-with-metas-ai-codec-and-prints-it-on-paper-as-eight-qr-codes
- Source URL
- https://www.tomshardware.com/feeds/all
Summary
- Score
- 3.5
- Created
- 17 Aug 2026, 12:10 AM
- Tags
- Audience
- developersai_ml_learners
What happened
A maker used Meta's AI audio codec to compress a 2.9MB song down to 21KB—a 1000x reduction—then printed the result as eight QR codes on paper. Playback requires a neural network to decode the compressed representation back into audio.
Why it matters
Demonstrates the extreme compression ratios possible with neural codecs like Meta's EnCodec, but the approach is impractical for production: you need a trained neural network on the decoding side, and the quality tradeoffs at 1000x compression are severe. Useful as a conversation starter about where neural compression fits versus traditional codecs in real shipping applications.
Discussion angle
Where does neural audio compression actually make sense in production—edge devices, low-bandwidth scenarios—and where is the decoder dependency a dealbreaker compared to Opus or AAC?