Length contraction — derivation: Visual walkthrough

Step 1 — Two observers, two rulers

WHAT. We set up two points of view (called frames). A frame is just "someone with their own ruler and their own set of clocks, all agreed among themselves." One frame belongs to the ground/lab — we call it $S$ . The other belongs to the rod itself — we call it $S^{'}$ (say it "S-prime").

WHY. The whole effect is a disagreement between two observers. You cannot see a disagreement with only one observer, so we always need exactly two frames: the one where the rod sits still, and the one where it whizzes past.

PICTURE. In the figure, the rod (blue) sits still in its own frame $S^{'}$ . The ground observer (in frame $S$ ) watches it fly by to the right at speed $v$ . Speed $v$ is measured in metres per second; $c$ is the speed of light, the cosmic speed limit.

Figure — Length contraction — derivation

Step 2 — What "length" even means for a moving thing

WHAT. Length = (position of the front end) minus (position of the back end). We write positions along the direction of motion as $x$ . So a length is a difference of two $x$ values: $x_{front} - x_{back}$ .

WHY. For a rod at rest this is boring — the ends don't move, read them whenever. But for a moving rod you must catch both ends at the same instant. If you photograph the front now and the back a second later, the rod slid forward in between and your answer is garbage.

PICTURE. Two panels. Left: both ends snapped at the same time — an honest length (green bracket). Right: front snapped early, back snapped late — the rod moved, giving a fake length (red bracket). The word " $t$ " is a time reading; two readings at the same $t$ are called simultaneous.

Step 3 — The proper length, read off in $S^{'}$

WHAT. In the rod's own frame $S^{'}$ the two ends sit at fixed positions, which we name $x_{1}^{'}$ (back) and $x_{2}^{'}$ (front). The prime $^{'}$ just means "as measured in $S^{'}$ ." Their difference is the proper length: $L_{0} = front, in S^{'} x_{2}^{'} - back, in S^{'} x_{1}^{'} .$

WHY. In $S^{'}$ the rod is standing still, so its ends never move — we do not even need simultaneity here. This is the "true, at-rest" length, the easy one to define.

PICTURE. The blue rod parked in $S^{'}$ , back end labelled $x_{1}^{'}$ , front end $x_{2}^{'}$ , with a yellow bracket marking $L_{0}$ between them.

Step 4 — The bridge between the frames: the Lorentz transformation

WHAT. How do coordinates in $S$ turn into coordinates in $S^{'}$ ? Not by plain subtraction (that would be old Newtonian physics). The correct rule is the Lorentz transformation: $x^{'} = γ (position in S x - shift for motion v t), γ = \frac{1}{1 - v ^{2} / c ^{2}} .$

Reading it term by term:

$x$ — where an event is, on the ground ruler $S$ .
$v t$ — how far the $S^{'}$ origin has slid by time $t$ ; subtracting it re-centres onto the moving frame.
$γ$ (gamma) — the stretch factor that fixes up the units so light has speed $c$ in both frames.

WHY this tool and not simple $x^{'} = x - v t$ ? Because the naive Galilean rule secretly assumes both frames share one universal clock. Experiments say they don't — light travels at $c$ for everyone. The only rule consistent with that is the Lorentz one, which multiplies by $γ$ . We use it precisely because it is the transformation that keeps the speed of light fixed. See Lorentz transformation and Lorentz factor gamma.

PICTURE. The factor $γ$ as a curve: it equals $1$ when $v = 0$ (no relativity) and shoots up toward infinity as $v \to c$ . That rising curve is why fast things behave strangely.

Step 5 — The measurement in $S$ : both ends at one instant

WHAT. The ground observer marks the two ends at ground positions $x_{1}$ (back) and $x_{2}$ (front), at the same ground time $t$ : $t_{1} = t_{2} = t .$ Feed each end through the Lorentz rule: $x_{1}^{'} = γ (x_{1} - v t), x_{2}^{'} = γ (x_{2} - v t) .$

WHY. This single line, $t_{1} = t_{2} = t$ , is the entire physics of length contraction. It enforces the Step-2 rule (simultaneous readings) inside the transformation. Watch what it does next.

PICTURE. A spacetime sketch: the two end-worldlines (paths through space and time) of the rod, and a horizontal "same-time-in- $S$ " slice cutting both at one instant $t$ . The two crossing points are $x_{1}$ and $x_{2}$ .

Step 6 — Subtract: watch the shift cancel

WHAT. Subtract the two end-equations from Step 5: $x_{2}^{'} - x_{1}^{'} = γ [(x_{2} - v t) - (x_{1} - v t)] = γ (x_{2} - x_{1}) .$

WHY. The $v t$ term is identical in both equations — only because $t_{1} = t_{2}$ . So when we subtract, the $- v t$ and $- v t$ cancel exactly. If the two readings had been at different times, $- v t_{1}$ and $- v t_{2}$ would not cancel, and there would be no clean length at all. This cancellation is where simultaneity earns its keep.

PICTURE. The two equations stacked, with the $- v t$ terms highlighted and an arrow showing them annihilate, leaving $γ (x_{2} - x_{1})$ .

Now rename the differences using what they physically are:

$x_{2}^{'} - x_{1}^{'} = L_{0}$ — the proper length (Step 3).
$x_{2} - x_{1} = L$ — the length the ground actually measured.

So: $L_{0} = γ L .$

Step 7 — Solve, and read off the shrinking

WHAT. Rearrange $L_{0} = γ L$ for the measured length $L$ : $L = \frac{L _{0}}{γ} = L_{0} 1 - \frac{v ^{2}}{c ^{2}}$

Term by term:

$L_{0}$ — the biggest (rest) length, sitting on top because everything else is smaller.
$γ \geq 1$ — dividing by something $\geq 1$ can only shrink $L_{0}$ .
$1 - v^{2} / c^{2}$ — a number between $0$ and $1$ ; multiplying $L_{0}$ by it shrinks the rod.

WHY divide, not multiply? Because $L_{0}$ is defined as the maximum. Any moving observer must measure less, so $L_{0}$ must be on top of the fraction. (Contrast Time dilation, where proper time is the minimum, so it gets multiplied up.)

PICTURE. The measured length $L$ as a fraction of $L_{0}$ versus speed: flat near $L_{0}$ at low speed, plunging to $0$ as $v \to c$ .

Step 8 — Every case: check the limits

WHAT & WHY. A formula you trust must behave sensibly at its edges. We test three regimes.

$v = 0$ (rod at rest): $γ = 1$ , so $L = L_{0}$ . No motion, no contraction. ✔
Small $v$ (everyday speeds): $v / c \approx 0$ , so $1 - v^{2} / c^{2} \approx 1$ ; $L \approx L_{0}$ . That is why we never notice contraction in daily life — cars are far too slow.
$v \to c$ (light speed): $γ \to \infty$ , so $L \to 0$ . The rod would appear crushed to zero thickness. No massive object can reach $c$ , so this is a limit, not an event.
Perpendicular direction (degenerate case): the derivation only ever touched $x$ , the direction of motion. Heights and widths (the $y, z$ directions) never entered, so they do not contract. A ball moving fast is squashed into a pancake along its motion but keeps its full height.

PICTURE. Three rods at $v = 0$ , $v = 0.6 c$ , $v = 0.98 c$ shrinking along the motion while their heights stay identical — showing both the shrink and the untouched perpendicular size.

The one-picture summary

Everything above in a single spacetime diagram: the two end-worldlines of the rod, the ground's "same-time" slice (giving the short $L$ ), the rod-frame's "same-time" slice (giving the long $L_{0}$ ), and the tilt between the two slices — that tilt is the relativity of simultaneity, the true source of the shrinking.

Recall Feynman retelling of the whole walkthrough

Picture a train racing past you. To fairly say "the train is this long," you and I both agree you must catch the front and the back at the very same instant — a photo, not two photos. Fine. Now here's Einstein's twist: the train's passengers and you do not agree on what "the very same instant" means. When you slice the train's history into your idea of "now," you cut it at a slightly slanted angle compared to how the passengers slice their "now." That slant is the whole story. Because your slice and their slice cross the train's two ends at different pairs of moments, you end up bracketing a shorter piece of the train than the passengers do of their own train. Nothing squeezed it, no force pushed on it — you two just disagree about "now," and disagreeing about "now" for a moving object automatically means disagreeing about its length. Divide by gamma, and out pops the shorter number every time.

Active recall

Connections

Length contraction — derivation — the parent note with the plain algebra.
Lorentz transformation — the bridge equation used in Steps 4–6.
Lorentz factor gamma — the $γ$ curve of Step 4.
Relativity of simultaneity — the tilt in the summary figure, the true cause.
Time dilation — the mirror-image effect ( $\times γ$ ).
Muon decay experiment — where this contraction is observed for real.
Spacetime interval — the invariant both slices preserve.