What this experiment shows

In this model, the near-to-far width ratio is enlarged by 20% at 30 cm and 3% at 200 cm, relative to the underlying width ratio. The dimensions do not change; their projected sizes do.

Try the camera-distance simulation

Two fixed width segments, 6 cm apart in depth. Move the camera to see their relative apparent sizes change.

Projected near / far width

0.2880

Relative enlargement of the near segment

12%

Focal length cancels from the ratio at a fixed viewpoint. This model does not generate a face score.

Inputs and method

We project two horizontal segments through an ideal pinhole camera. The near segment is 3.6 cm wide; the far segment is 14 cm wide and 6 cm farther from the camera. We vary distance to the near plane from 30 to 200 cm. These dimensions are chosen for illustration, not measured from a person. Image width = focal length × segment width ÷ distance.

Calculation version: ARF calculation experiments 1.0. Engine: ARF Local Geometry v3.0.

Calculated results

Relative enlargement (%)012.525Camera distance (cm): 30; Relative enlargement (%): 2030Camera distance (cm): 50; Relative enlargement (%): 1250Camera distance (cm): 100; Relative enlargement (%): 6100Camera distance (cm): 150; Relative enlargement (%): 4150Camera distance (cm): 200; Relative enlargement (%): 3200Camera distance (cm)
Calculated from the published inputs. The full values and configuration details are in the table below.
5 calculated configurations. Scores use the site's 0–100 scale; match shares and enlargement are percentages.
Distance (cm)Projected width ratioRelative enlargement (%)
300.308620
500.28812
1000.27266
1500.26744
2000.26493

Download this table (CSV) · Reproduce the calculation

How to use this when reading your result

For a more comparable face photo, keep camera distance consistent. Changing focal length at a fixed viewpoint scales both segments equally, so their ratio stays the same. Moving the camera to keep the same framing changes perspective. Cropping a close selfie cannot undo that perspective.

Compare your own photos →

What this experiment cannot tell you

This is an ideal two-plane projection, not a 3D head, a camera calibration or a face-score prediction. It omits lens distortion, pose, detection error and real facial depth variation. The percentages describe these chosen dimensions only.

Sources and calculation choices

  • Nasal Distortion in Short-Distance Photographs: The Selfie Effect — Ward et al., JAMA Facial Plastic Surgery (2018). Background for why camera distance changes relative feature size in a perspective image. This paper does not test AI Rate Face. Our camera simulation uses its own stated dimensions and is not a reproduction of the study.

See the full source notes and methodology. To report an error, contact us with the page URL and the input or calculation that appears wrong.