Run a large-area (stitched) scan
Draft outline
Scaffold. Replace the bullet prompts with your own text and delete this banner when done.
Learning-oriented, for users. A single camera frame only sees a tiny part of your sample. In this tutorial you'll define a region, let the FRAME image it tile by tile, and view the tiles stitched into one large overview image.
What stitching is (30-second version)
- The stage moves the sample under the fixed objective in a grid of positions.
- One frame is captured at each position; overlapping tiles are stitched together.
- Link to the concept page: How scanning works.
Before you start
- Sample inserted and roughly in focus (first sample).
- Objective chosen (4x is the easiest first scan — large field, forgiving focus).
- Illumination set.
Step 1 — Define the scan region
- How to set the scan area (corners / centre + size / well selection).
- How to set tile overlap.

TODO image
Screenshot of the scan/region setup panel. Notion source:
FRAME_DOKU_FROM_NOTION/FAT FRAME #0007 Korea - Part 5 ("First Stitching Test").
Step 2 — Choose single-Z (fixed focus) for now
- Explain single-Z vs. autofocus vs. focus map (forward-reference the other tutorials).
Step 3 — Run the scan
- Start the scan; what the progress display shows; roughly how long 4x takes.

TODO image
Scan-in-progress + resulting stitched overview. Notion source:
FAT FRAME #0007 Korea - Part 6 ("Stitching Tests 1-5") and Part 7/Part 8.
Step 4 — View and save the stitched result
- Where the stitched image appears; how to save/export it (OME-TIFF).
What "good" looks like vs. artefacts
- Seams / brightness steps between tiles → illumination flatness (diffusor), overlap.
- Notion source for the shadow/diffusor investigation:
FAT FRAME #0007 Korea - part 7("Imaging test to find bug with shadow", "w/o diffusor").
Try this
- Increase tile count / area and compare time and file size.
- Switch to 20x and notice the smaller field and tighter focus tolerance.
Related
- Acquire a Z-stack
- Keep a big scan in focus with a focus map
- How-to (after you've learned it): pixel-size calibration