Did your face change, or did the photo?
A symmetry reading is a measurement of a photo, not a direct measurement of your face in person. Camera position, expression, lighting and landmark placement can all change it. Use the free browser symmetry tool to keep a baseline reading and compare a retake privately. Differences between the two photos are not proof of a physical change or improvement.
Try a two-photo capture experiment
- Use the photo checklist. Take photo A at eye level with a neutral expression.
- Analyze A in the symmetry tool, then choose “Use this reading as baseline”.
- Take photo B without changing your face or setup. Choose B in the same tab and note the differences.
- For a separate experiment, change just one capture condition, such as camera height, and compare again. Do not deliberately make an uncomfortable pose.
- Record the setup and each feature reading. Repeat on another occasion before drawing any conclusion about the setup.
Read the units correctly
| Illustrative reading | Meaning |
|---|---|
| Photo A: 1.2%; photo B: 1.8% | A change of +0.6 percentage points of cheekbone width |
| What it establishes | The tool returned different measurements for these photos |
| What it does not establish | That your face changed, or that the difference is medically meaningful |
The numbers above explain the units; they are not results from participants or a validation study. The browser tool uses rough bands (under 1.5% is a close match; 4% or more is a clear difference), not a measured noise floor. More decimal places do not make it more accurate.
What is measured and what stays local
The browser tool fits a face midline from 32 paired landmarks in a model-estimated three-dimensional frame, compares matching features, and divides differences by cheekbone width. It rejects large estimated turns and tilts. This may reduce pose sensitivity but does not remove landmark or camera error. The photo stays on the device; baseline readings remain only in the current tab until cleared or you leave.
A protocol others can repeat
To evaluate repeatability, use consented adult images, fixed camera and lighting, multiple independent captures per person, and recorded capture conditions. Report per-feature variation, failed captures, pose rejections and the tool version. Keep participants and source images private unless they separately agree to publication. We have not published a controlled participant validation study for this browser tool.
The downloadable checklist is an original MaxxLook resource you can share with a stylist or use for your own capture notes. A future validation study should report its sample and limitations alongside the results.
Common questions
Why does my face look more asymmetrical in some photos?
Different camera position, expression, lighting and a mirrored image can change what you notice and what a landmark tool measures. Match the setup and compare more than one photo. A photo tool cannot determine a medical cause.