Using ODyn

Scanning a Product

Capture products so ODyn can encode every ingredient — and compare several suspects at once after a spike.

Scanning is how ODyn learns what you consumed. It reads a product's ingredient panel and encodes every ingredient at the molecular level, so it can recognize the same trigger even when it hides under different names on different products. A clean scan is the foundation for everything ODyn tells you later.

This guide covers three things: scanning a single product, scanning a batch of suspects after a spike, and getting a clean capture.

Scan a product

1. Open Scan from the menu. 2. Add a product and photograph its ingredients panel — the back-of-package list. This photo is required. 3. (Optional) Add a photo of the front of the package (helps ODyn identify the product name) and the UPC barcode (matches it to a known product). 4. Submit.

What to expect: ODyn reads and encodes the ingredients, flags anything it can't confidently identify, and adds the product to your record.

The Scan screen with a product added, showing the ingredients-panel / front / UPC photo slots.(click to enlarge)

Scan a batch after a spike

To find what caused a spike, compare suspects side by side instead of guessing one at a time.

1. Start a scan and add up to ten products — everything you consumed in the hours before the spike. 2. Capture each product's ingredients panel (add front and UPC photos if you like). 3. Submit the batch.

What to expect: ODyn runs a comparative analysis and returns a score for each item with an ingredient-level breakdown — which items stand out, which ingredients are driving them, and why.

A completed comparative analysis — per-suspect scores with the ingredient-level breakdown.(click to enlarge)

Get a clean capture

The scan is only as good as the photo. Before you submit each panel:

  • Fill the frame with the ingredient list — don't cut off the top or bottom.
  • Use even lighting — avoid glare and hard shadows across the text.
  • Hold steady — motion blur is the most common cause of a misread.
  • Flatten curved labels so the text in frame sits reasonably flat.

If an extracted list looks wrong — missing items or garbled text — re-take the panel photo and submit again.