Screen Ruler vs Ruler.onl — Accuracy Test on 3 Devices
Ruler.onl has a deliberately minimal design — just a ruler, nothing extra. No settings panel, no calibration wizard. You open it, you measure. That kind of simplicity is genuinely appealing. The question is whether "simple" and "accurate" can coexist when you're dealing with dozens of different screen hardware configurations.
We ran the same physical measurement test across three devices and compared Ruler.onl against Screen Ruler. Here's what we found.
Quick Comparison
| Feature | Ruler.onl | Screen Ruler | |---|---|---| | Calibration | None (fixed 96 PPI) | Credit card or coin | | Desktop accuracy | ~-4% | <1% | | MacBook accuracy | ~+58% | <1% | | iPhone accuracy | ~+380% | <1% | | Mobile support | Layout exists, inaccurate | Fully optimized | | Vertical ruler | No | Yes | | Dark mode | No | Yes | | Saved calibration | No | Yes | | Languages | English only | 20 languages |
Accuracy Test Results
The test setup: a physical ruler marked to known lengths, measured against what each tool displayed on screen. Same lengths, same three devices.
Desktop monitor (92 PPI actual)
Ruler.onl returned a -4% error. This is actually the best-case scenario for Ruler.onl — a standard desktop monitor sits close to the 96 PPI assumption baked into CSS. The gap between real hardware and the fixed assumption is small, so the error stays manageable. For rough measurements on a typical office monitor, Ruler.onl does fine.
MacBook (227 PPI, Retina)
Here the error jumped to +58%. A MacBook's Retina display packs 227 physical pixels into every inch, but macOS doubles the CSS pixel size so that UI elements look the same visual size across devices. The result: one CSS "inch" actually spans about 1.58 physical inches on screen. Ruler.onl measures in CSS inches without compensating, so everything it shows is stretched. A reading of "10 cm" is physically closer to 15.8 cm.
Screen Ruler after credit card calibration: under 1% error on the same machine.
iPhone (460 PPI, 3x device pixel ratio)
This is where the gap becomes absurd. Ruler.onl showed +380% error. A marking labeled "1 cm" physically measured about 4.8 cm on the actual screen. The tool is essentially non-functional for mobile measurement.
Screen Ruler after calibration: under 1% on the same phone.
Why Ruler.onl Is So Inaccurate on Phones
This isn't a design oversight unique to Ruler.onl — it's the consequence of how CSS works.
CSS defines one inch as exactly 96 pixels, regardless of what the physical hardware does. That made sense in the early 2000s when monitors were mostly 96 PPI. Today's phones are anywhere from 300 to 500+ PPI, and they use a device pixel ratio (DPR) to scale things up. An iPhone with a 3x DPR maps 3 physical pixels to every 1 CSS pixel.
Run the math: 96 CSS pixels per inch ÷ 3 DPR = 32 physical pixels per CSS "inch." But the phone actually has 460 physical pixels per physical inch. So 32 physical pixels spans only 0.07 inches physically, not 1.
The actual calculation: 96 CSS pixels / (460 physical PPI / 3 DPR) = 0.63 inches per CSS inch. That's a 37% compression — meaning the ruler reads about 380% too large compared to what's real.
There's no workaround within CSS alone. The only fix is calibration: measure a known physical object on the screen and correct the scale factor from there. Screen Ruler does this with a credit card (85.6 mm × 54 mm, an ISO standard) or a coin. Ruler.onl doesn't do this, so it can't be accurate on any device that deviates from 96 PPI.
Feature Comparison
Ruler.onl
- Horizontal ruler only
- Centimeters and inches
- Minimal, distraction-free interface
- No calibration
- No dark mode
- No saved settings between sessions
Screen Ruler
- Horizontal and vertical rulers
- Centimeters, millimeters, and inches
- Fullscreen mode
- Dark mode
- Device detection with automatic PPI estimation
- Credit card and coin calibration
- Calibration saved across sessions
- 20 interface languages
Ruler.onl wins on initial simplicity. You don't have to do anything to get a number. The trade-off is that the number may be wrong by hundreds of percent depending on your device.
The Verdict
Ruler.onl works well enough as a rough reference on a standard desktop monitor. If you're measuring something on a 1080p or 1440p desktop screen and you just need a ballpark figure, the ~4% error probably won't matter.
For everything else — a laptop with a high-DPI display, a phone, a tablet, any situation where accuracy matters — it's not really measuring. It's making an assumption that doesn't hold.
The underlying problem isn't fixable without calibration. Screen Ruler's approach of letting you place a known object on screen and correcting the scale factor is the only reliable way to get accurate physical measurements from a browser. Once calibrated, accuracy stays under 1% across all tested devices.
If you're switching from Ruler.onl and want the same minimal feel, Screen Ruler's interface is similarly clean — just with a calibration step that actually makes it work.