Set up Selfie Mirror

1. In the sidebar, navigate to Apps → Installations
2. Open your Selfie Mirror installation
3. Go to the Settings tab
4. Under Global settings, click Edit settings

The mirror shows and speaks its texts in the installation’s default language.

1. In Multi-language support (see Glossary), enable the languages you want to write texts for
2. Set the default language for the mirror
3. Fill in the texts in each screen’s settings for that language

Voice answers are understood in English and Arabic for “yes” and “no”. Cover numbers are understood in English, Arabic, Spanish, and French, and as digits.

The idle screen runs while nobody is at the mirror. It wakes up when a visitor stands in front of the camera, or when someone taps the screen.

1. Expand 1. Idle → Idle / Attract Screen
2. Upload a Background image or video. Videos play muted and loop
3. Enter a Brand Title, such as “Come closer to take a selfie!”
4. Optionally, add up to three Sample Selfies to show visitors what to expect
5. Set the Idle Timeout (seconds): how long the mirror waits without interaction before returning to the idle screen
6. Enter the Timeout Phrase spoken when it returns, such as “Goodbye”

The prompt asks the visitor whether they would like to take a selfie. They answer by voice or by tapping a button.

1. Expand 2. CTA
2. Enter the Title, such as “Would you like to take a selfie?”, and an optional Sub-title
3. Enter the Yes button label and No button label
4. Optionally, upload a Background (override). Without it, the screen uses the idle background

1. Expand 3. Capture → Reference Backgrounds
2. Add one reference background per cover, up to three
3. For each one, upload the Image (the scene) and, optionally, the Overlay (a transparent PNG drawn on top of the person)
4. Use Person Placement to position the person in that scene:
   • Horizontal position: 0 is far left, 0.5 is centred, 1 is far right
   • Vertical position: 0 is the top of the scene, 1 is standing on the bottom edge
   • Scale: 1 fits the person to the full height of the scene. Lower values show more of the scene, higher values enlarge the person
5. Take a test selfie and adjust the placement until the person sits naturally in each scene

See Prepare cover images for Selfie Mirror for image sizes and formats.

1. In 3. Capture, set the Countdown (seconds) the visitor gets to pose. Set it to 0 to take the photo immediately
2. Enter the Camera Indicator Label that points visitors to the camera, such as “THE CAMERA IS AT THE TOP”
3. Optionally, upload a Background shown around the camera preview, and Processing Media shown while the covers are being made
4. Adjust the Flash Duration (ms) of the on-screen flash when the photo is taken

Visitors pick their favourite cover by tapping it or saying its number.

1. Expand 4. Selection
2. Enter the Heading, such as “Select the selfie you like the most!”, and an optional Sub-heading
3. Optionally, upload a Background

The share screen shows the chosen cover with a QR code. Scanning it opens WhatsApp with a pre-filled message. When the visitor sends it, the cover arrives in the same chat.

1. Expand 5. Share
2. Enter the Title, such as “Great selfie!”, and the Hint, such as “Scan the QR code with your phone to download it.”
3. Enter the Send Message: the text pre-filled in the visitor’s WhatsApp message
4. Enter the Reply Message: the text sent back together with the cover
5. Optionally, upload a Logo for the short-link label, typically the WhatsApp logo, and a Background

1. Expand General
2. Upload your Logo, shown at the top of the idle and prompt screens. Use a PNG or SVG with transparency
3. Leave Theme (Primary Color and Secondary Color) as it is: the current app version doesn’t use these colours
4. In Fonts, optionally upload Light Font, Regular Font, and Bold Font files. Empty weights use the system font

The mirror can read its texts aloud: the prompt title, the countdown, the selection heading, the share title, and the timeout phrase.

1. Expand Speech
2. Turn Enable Audio Speech on or off. When off, the mirror stays silent and listens for voice answers right away
3. Choose a female or male Voice
4. Leave TTS Model on auto unless PhyGrid support advises otherwise