Of all the small disciplines in image-based inspection, this is the cheapest to adopt and the most often skipped. It costs one click, and it is the difference between an image that stays evidence and an image that becomes decoration.
The problem with quoting magnification
Write "200×" under an image and you have made a claim about a specific display at a specific size. Now follow that image through its actual life:
- It is placed in a report and scaled to fit the page width
- The report is exported to PDF
- Somebody views the PDF at 150% on a laptop
- They screenshot a region and email it
- It is pasted into a presentation and dragged to fit a slide
- It is printed on A4 by one person and A3 by another
At every step the magnification changed and the caption did not. By the third step the number is decorative; by the sixth it is misleading. And unlike a wrong value in a table, nobody notices, because a magnification figure carries no internal contradiction — it just quietly stops being true.
Cropping is worse still. Crop to a quarter of the frame and the field of view has changed, so even a correct original magnification is now wrong for the picture in front of you.
What a scale bar does instead
A scale bar is drawn into the image, in the image's own pixels. It says: this drawn length equals this real length.
Because it is made of the same pixels as the part, every transformation that affects the part affects the bar identically. Halve the image and both halve. Print it on A3 and both grow. Crop it, and as long as the bar is still in frame, the relationship holds. Screenshot it, paste it, resize it — the bar remains correct because it is not a claim about the display at all.
A magnification figure describes the viewing conditions. A scale bar describes the image. Only one of those survives being emailed.
There is a second benefit that matters in an audit. A scale bar lets a reader independently check your numbers. If the bar says 1 mm and the feature you reported as 4 mm is visibly about four bar-lengths across, your report is self-consistent and anyone can see it. If it is visibly about ten, you have a problem worth finding before your customer does — typically a calibration applied at the wrong magnification.
What makes a good scale bar
- A round number. 1 mm, 500 µm, 10 mm. Not 1.37 mm.
- A sensible fraction of the frame — roughly a tenth to a third of the image width. A bar spanning the whole image is hard to compare against a small feature; a tiny one cannot be measured against anything.
- Labelled with its units. An unlabelled bar is an unlabelled number.
- High contrast, on a quiet part of the frame. Usually a corner, clear of the part.
- Not over the feature being measured.
Keep the magnification or optical configuration in the report's information block as well — it is genuinely useful for repeating the setup. Just do not rely on it to tell a reader how big anything is.
How Measuret handles it
Measuret has a Scale control in the Display group of the Measurement tab that shows or hides a scale marker on the image. You choose whether the bar runs horizontally or vertically, the length it represents, and the unit label — so it can be set to a round number that suits the field of view rather than whatever the software would have picked.
Because the marker is part of the overlay, it is included in everything the overlay is included in: the measured image on the A4 report, and the PNG export of the canvas. The PNG export is deliberately what-you-see-is-what-is-saved, so an image sent in an email carries its scale with it.
The bar is drawn from the same calibration that produces the measured values, so the bar and the numbers cannot disagree. If the calibration is wrong they are wrong together — which is exactly the behaviour you want, because a reader comparing the two will not be falsely reassured.
A scale bar makes an image interpretable. It is not a substitute for the calibration record behind it — the bar is only as honest as the scale factor that drew it.