Elsehow is a browser-based modality-translation instrument. Give it an image and it turns horizontal position, light, color, and the vertical center of brightness into a 16-part spatial sound-map.
No account. No upload. No setup.
The translation runs locally in the browser. Elsehow keeps three layers visible:
- what crossed into sound;
- how the source was transformed; and
- what could not cross.
It does not claim the sound-map is equivalent to the image. It is another route into the same source.
Open dist/index.html, choose an image, and select Hear the map. Use Left/Right Arrow to inspect individual slices or tap the image to hear one region.
The included calibration specimen makes the full loop demonstrable without uploading anything.
| Image measurement | Sound property |
|---|---|
| Horizontal position | Stereo pan |
| Vertical brightness center | Pitch |
| Average lightness | Loudness |
| Average color family | Timbre |
Pixels remain in the browser. There is no upload endpoint, account, analytics SDK, or durable storage in this prototype.
Created by Morrow with Miranda Price as Starhands for Hackyard, September 2026. It carries forward the central discipline of The Listening Glass: measurements, mappings, inference, and provenance should remain distinguishable.