Don't calculate it from lens specs — measure it once. The
magnification printed on clip-on and USB microscope lenses is a marketing
number, and phone cameras crop and resample internally, so any spec-sheet
calculation can be out by 2–3×. A one-time measurement is accurate to ~1%.
What you need
A stage micrometer — a glass calibration slide with a 1 mm
scale in 10 µm divisions (≈$10–20 online). Don't use a printed target:
printer dots are 50–100 µm, far too coarse.
1 · Lock your camera setup
- Attach the microscope lens over one specific camera and
select that lens by tapping its preset (e.g. 3×) — never
pinch-zoom. Check you're on the right camera by covering each lens in turn.
- On iPhone, disable the auto-macro switch (it silently swaps
to the ultrawide camera for close subjects, which invalidates the calibration).
- Fix the photo resolution (12/24/48 MP) and always use
the same setting — resolution changes µm/px even with identical optics.
2 · Photograph the scale
Shoot the micrometer through the exact same stack, at the lens's natural
working distance, slide flat and perpendicular to the lens.
Orientation: lay the scale parallel to the long edge of
the image so you can read the pixel span straight across. Camera pixels are
square, so direction doesn't change the answer — but the real distance and the
pixel distance must be measured along the same line. If your scale ended
up at an angle, use the straight-line distance between the two ticks:
pixels = √(Δx² + Δy²), not the horizontal span.
3 · Compute
Count the pixels across as long a run of divisions as possible (ImageJ, or
any editor that shows a selection size in pixels), then:
µm per pixel = known distance (µm) ÷ pixels spanned
Example: 50 divisions = 500 µm spans 820 px → 0.61 µm/px.
(The field-of-view option is the same number expressed differently:
FOV width = image width in px × µm/px.)
4 · Sanity checks
- Repeat the measurement near a corner of the frame. If it
differs from the centre by more than a few percent, the lens distorts — keep
particles in the central region of your patch shots.
- Confirm the smallest divisions are crisply resolved. If
10 µm lines smear together, your true sizing floor is coarser than the
pixel maths suggests.
- The smallest reliably sized particle is ≈ 3 × your µm/px. To measure the
ISO 4 µm channel you want ≲1.3 µm/px.
5 · When to redo it
Different phone, lens remounted differently, resolution setting changed, or a
major camera software update. Otherwise the number is stable — write it down
with the setup it belongs to (e.g. "iPhone 17 Pro, 3× tele, 40× clip-on, 24 MP →
0.61 µm/px").
Shortcut: the Calibrate from a photo button below the
calibration inputs does steps 3–4 for you — load the micrometer photo, click two
tick marks, type the real distance, and save the result as a named profile.
