Whether an app can read your palm from a photo comes down to three things: brightness, camera shake, and how much of the frame your hand fills. Even light with no hard shadow across the palm, a hand held flat and still, and a hand covering at least a third of the frame along its own long axis will work on most phones. Orientation is not one of the three. Sideways, upright, or at an angle all read the same.
PalmMuse measures three properties on the phone before a photo is ever uploaded. These are the actual limits it applies.
Average brightness across the frame has to sit between 0.25 and 0.85 on a zero-to-one scale. An indoor evening shot with the light behind you falls under the floor; direct sun that blows out the highlights goes over the ceiling. Daylight from a window, hitting the hand from the side, is the hardest version to get wrong.
Shadow causes more failures than brightness does. A light source directly above the hand drops a band of shadow from the base of the fingers into the centre of the palm, and the heart and head lines can merge into one crease under it. Put the light in front of you or off to one side.
Sharpness is measured from edge contrast across the image. Palm creases are fine enough that a photo perfectly acceptable for a portrait can still be too soft to trace lines from. Rest your wrist on a table or your knee and let the arm settle before the shutter. Holding the phone in one hand while presenting the other is the position most likely to shake.
A palm too small in the frame does not have the resolution to hold creases. The floor is thirty percent of the frame measured along the hand's own long direction. You do not have to fill the frame edge to edge, but a photo with your forearm in it is usually short.
Hand size is measured along the hand's own longer axis, not the screen's. A palm photographed sideways scores exactly the same as one photographed upright, so there is no need to rotate your wrist into a pose. Use whichever angle you can hold still.
It did not always work this way. An earlier version measured spread only on the vertical, which meant a palm photographed sideways failed every time regardless of how good the photo was. Measuring on the hand's axis is the fix.
The capture screen draws a hand outline. Sitting slightly outside it, or slightly inside it, does not stop the capture. The on-device check exists only to reject a hand small enough that the creases will not resolve, and the real decision is made by the analysis afterwards.
If the phone cannot detect a hand pose at all, the photo still goes through. Hand-pose detection misses plenty of perfectly usable palm photos, and blocking an upload on it costs more than letting the analysis answer.
Two things can come back. The first is that the image could not be judged to clearly show the palm side of a human hand. The back of the hand, fingers only, or a frame with something else dominating it all produce this.
The second is that the major creases or the full outline of the hand could not be assessed reliably. Darkness, shake, a hard shadow, or fingers curled enough to foreshorten the palm are the usual causes.
In both cases the app says so rather than guessing at an answer. Retakes are free and unlimited, and a failed scan never costs anything.
Faint lines and sparse palms are ordinary. Plenty of hands have no clear fate line, and plenty that do will not show it at a given angle. The confidence figure attached to a detected line describes how plainly that crease could be traced in this photograph. It says nothing about the person holding the hand, and a low figure never changes what a line is read to mean.
If a line you can see was not detected, you can correct it before going any further. Drag it onto the crease you can see, or mark it as not clearly visible. The reading is written from what you confirmed.
Either works. PalmMuse reads each hand separately and only adds the section comparing them once it has both. A single hand is a complete reading on its own.
Yes. The checks look at the photograph rather than at how it was taken, so an older photo works as long as the brightness, sharpness, and framing hold up.
Nail polish has no bearing on the creases, which are on the other side of the hand. Hand cream can, if it is fresh enough to make the palm shine and reflect light back into the lens, which flattens the creases. Let it absorb first.
No. Capture and line detection are free, including the confidence figures, and you decide whether to buy a reading after seeing them. Failed scans are never charged.
Better not to. A flash at close range blows out the centre of the palm and leaves the edges dark. Room light or daylight is far more consistent.
PalmMuse scans free and shows you every line it found, each with its confidence figure, before you decide whether to buy a reading.
Get the launch email