The most stable setup: daytime, near a window, with soft diffused light
Stand directly or slightly diagonally in front of a window so both cheeks receive similar brightness. Overcast light is often useful. On a bright day, use a sheer white curtain to prevent hard shadows.
Use daytime natural light. Avoid sunset, colored lamps, and screen fill light.
Use the standard camera with filters, beauty smoothing, and makeup effects off.
Choose white, gray, or low-chroma surroundings that do not reflect color onto the face.
What separates a useful photo from a problem photo
Even light and a straight-on angle
- Both cheeks have similar brightness
- Eyes retain detail without blown highlights
- Natural hair color is visible
- The background is neutral
Yellow light, backlight, heavy makeup, or filters
- One side of the face is much darker
- The window is behind you
- Foundation or lipstick changes natural coloring
- Beauty smoothing removes color detail
Take the photo in six steps
- Clean the lensOil creates haze and lowers contrast.
- Turn off filters and beauty effectsInclude warm camera filters, portrait relighting, and third-party makeup effects.
- Face the windowLet natural light travel from the camera direction toward the face.
- Keep a natural appearanceAvoid makeup and colored contacts; clear hair away from the cheek while leaving natural hair visible.
- Look straight aheadKeep the camera near eye level at roughly arm’s length.
- Check exposureSkin should not contain large pure-white areas; eyes, brows, and hair should retain detail.
If results change, troubleshoot in this order
If the entire photo looks yellow or blue, retake before interpreting a season change.
Avoid shine, blemishes, obvious redness, and jaw shadow. Choose an even cheek area.
Use the same device, window, and similar time when comparing multiple tests.
If controlled results still alternate between neighbors, focus on the four axes rather than one label.
Separate the browser first test from classic app analysis
The photo stays in this browser
The web tool reads local pixels for skin, eye, and hair samples and does not upload the photo.
Data handling is explained before analysis
Before classic analysis starts, the app explains the purpose and requires explicit agreement to upload and retain that photo privately; if you decline, analysis cannot start. Retained photos can be deleted in Settings.
Do not upload another person’s photo without permission, and do not treat a personal color result as identity or health information.
Sources and editorial note
The standards and technical report below were used to check variables in illumination, camera color capture, and skin-color measurement. They support controlling capture conditions; they do not show that one consumer-camera photo can produce a medical or absolute season result.
- ISO/CIE 11664-2:2022: CIE Standard IlluminantsUsed to check spectral differences between tungsten light and representative daylight, plus known daylight variation by time, season, and location.
- CIE 256:2025: Measurement of Human Skin ColourUsed to check how illumination geometry, sample location, instrument settings, and individual diversity affect comparable skin-color measurements.
- ISO 17321-1:2012: Colour characterization of digital still camerasUsed to check that digital-camera color capture requires defined stimuli, metrology, and test conditions; a camera output should not be assumed to equal scene colorimetry.
Sources checked August 25, 2026.