Cropping a Photo to ID Photo Size: Framing, Head Height and Margins

Why Is Cropping an ID Photo Not Just Cutting?

Because the crop has three constraints, not one. First the aspect ratio: does the document want a portrait or a square. Second the head's share of the frame height: the head has to occupy a defined portion of the photo. Third the margins: a thin gap above the head and, below the chin, space down to the shoulder line. A crop with the right ratio but the wrong head share gets rejected — which is why cropping is a geometry calculation rather than a cut.

The tool runs that calculation for you. The face map locates the chin point and the top of the skull, the head height is divided by the share the document format asks for to give the crop height, and the crop is then positioned according to the top margin. Once the crop height is known the width follows from the format ratio — so the correct framing is built without you needing a millimetre figure. All millimetre and pixel equivalents live somewhere else: the complete document photo size table. This page does not repeat that data; it explains how the crop is built.

How Is Head Height Measured in the Crop?

Head height runs from the chin to the top of the skull, and hair is not part of the measurement — ICAO 9303 defines the crown explicitly as the top of the head ignoring any hair. Since the face map does not return the skull top directly, the point is extrapolated above the forehead by a fixed proportion of the face height. Hair plays a separate role: volume does not enlarge the measurement, it only shifts the crop upwards so the hair is not cut off. That distinction matters, because measuring with hair included makes the head look smaller than the standard and needlessly enlarges the crop.

The expected share varies by document family. Some specifications state it as a ratio outright; the US digital image specification defines head height as 50% to 69% of the image height (travel.state.gov). The international biometric band sits at the top end, while the Turkish print tradition leaves more room around the head. You are not calculating a ratio: pick the document in the wizard and its band is applied automatically. Every document-level rule is collected in the biometric photo guide.

Which Framing Errors Are Fixed and Which Need a Reshoot?

One logic runs through the table: missing information cannot be invented, existing information can be rearranged. If there is no margin above the head, the crop can be pushed beyond the image edge and that area filled with the backdrop colour — the backdrop is generated anyway, so the fill is invisible. But if the head itself is cut in the source, no fill can build a head; a clear cut means rejection and the only remedy is a reshoot. The bottom edge has no such flexibility: filling under the body would leave the person floating, so the crop never overflows downward. Sides are not filled either, because if a shoulder touches the source edge the fill would slice the body with a vertical stripe.

The practical conclusion is that margin saves the crop. Official definitions ask for the same thing; the UK passport photo guidance describes the frame as a close-up of your head and shoulders (GOV.UK) — neither a portrait with the head clipped nor a landscape where the person is tiny. Shooting from a distance is still a recoverable error: if the crop area falls below the target pixel value, the super-resolution model enlarges the frame and the target size is met without quality loss. Shooting too close cannot be recovered, because there are no pixels to add below the chin.

Does Shifting the Crop Upward Break the Head Ratio?

It does not, because position and size are two separate decisions. The crop height is calculated once from the head share; the vertical position is set separately. With voluminous hair the crop shifts upward so the hair is not cut, while the head share stays the same. The shift has a limit: the chin cannot be pushed closer than a set distance to the bottom edge, so the wish to save the hair never goes far enough to erase the space under the chin.

What genuinely changes the head share is the width constraint. The crop rectangle cannot exceed the width of the source; if the face sits very close to the frame or the source is a landscape shot, the calculated height shrinks with the ratio kept. As the height shrinks the head looks larger in the frame and the share walks toward the upper band; the composition check measures and reports that. This is why the safest crop is a frame that already leaves a thin gap above the head and comfortable margin at the sides — file size and resolution can be adjusted later, lost margin cannot.

Can I Crop the Same Frame to Different Sizes?

Yes, but the right way round. When you pick a different document format the finished output is not re-cut; the original frame is reprocessed for the new ratio and the crop is rebuilt with that format's head share. Cropping the finished file breaks two things at once: the head share leaves the format's band and the pixel count drops. So when a second size is needed the move is simple — run the same photo again with the new format selected.

Getting several documents out of one shoot works the same way. For the framing of Turkish print documents the 50×60 mm guide and for international applications the 35×45 mm guide carry the relevant data; how the square format differs is visible on the US visa photo page. For document-level examples, the ID card photo, passport photo and student ID photo pages explain which framing each application expects. You are not the one making the crop decision: choose the format, upload the frame, read the head share and margins in the free analysis, download the result.