Shape Distortion: Elongation & Foreshortening

Part inclined · 20°

Foreshortened — the image is −5.3% of the aligned length

The CR and IR are perpendicular to each other, but the part is inclined. The greater the incline, the more foreshortening. Measured against the aligned setup at the same SID and OID, so the number is shape distortion on its own — the magnification both setups share has been divided out.

The part is presenting less of itself to the beam

With the receptor square to the beam and the part tipped up off it, the part covers less of the receptor than its true length would, and records short.

And the joint closes as well — this case costs you both

Tipping the part also tips the joint, so the rays that used to run down the gap now cut across it. A finger that is not parallel to the receptor shows both losses at once: the interphalangeal joint spaces are no longer open, and the phalanges are distorted as well. It is why a support block goes under an oblique hand — keeping the digits parallel prevents foreshortening of the phalanges and stops the joints being obscured.

When the part cannot be laid flat

If the part physically will not go parallel with the receptor — a knee that will not extend, for instance — the law of isometry says to angle the central ray to half the angle between the part and the receptor, which gives the least foreshortening available. It does not restore the joint relationships a perpendicular central ray would, so which compromise to take depends on what the projection has to show.

The asymmetry worth memorising

Three of these four errors elongate and only one foreshortens. Elongation comes from the part being off-centre, the central ray being angled, or the receptor not being parallel with the part. Foreshortening has a single cause: the part itself is inclined. Length is not the whole picture, though — the receptor being out of parallel is the one error that leaves the joint alone, while the two that barely change the length are the ones that shut it.