How this instrument works
Velocity addition answers a deceptively simple question: if A moves at speed u relative to the ground and B moves at speed v relative to A, how fast does B move relative to the ground? At walking pace the answer is just u + v, and that Galilean rule served physics for centuries. But it silently assumes time and distance are the same for every observer, which special relativity showed is false. The correct rule is w = (u + v) ⁄ (1 + uv ⁄ c²), where c is the speed of light in vacuum. The denominator is the whole story: at low speed it sits at almost exactly 1, so w collapses back to u + v, but as u and v grow it swells and holds w below c no matter how large the inputs get.
The denominator term uv ⁄ c² comes directly out of the Lorentz transformation, the coordinate change that keeps light's speed the same for every inertial observer. Divide two velocities that are each a healthy fraction of c and that product is no longer negligible next to 1, so the sum gets pulled down. Set u to c itself and the formula returns w = c exactly, regardless of v — a photon fired forward from a ship already moving at light speed is still measured at c, never faster, which is the entire point the formula was built to enforce.
The equation is symmetric in u and v, so it makes no difference which velocity is labelled A's and which is B's-relative-to-A; swapping the two fields returns the same ground speed. Engineers rarely meet this correction on the road or in the air, where uv ⁄ c² is smaller than any instrument can register, but it is routine business for accelerator physicists combining beam and particle velocities, for astrophysicists tracking jets ejected from black holes at a sizeable fraction of c, and for anyone checking why GPS satellite clocks need a relativistic correction at all.
- Enter Velocity of A relative to ground — A's speed as measured by a stationary observer, in metres per second.
- Enter Velocity of B relative to A — how fast B moves as measured from inside A's own reference frame, also in m/s.
- Leave both at zero or low values to see ordinary addition; push either past roughly a tenth of c (about 30,000,000 m/s) to watch the correction take hold.
- Read Velocity of B relative to ground — the instrument's one output field, always less than 299,792,458 m/s no matter what you enter above.
Worked example — a ship, a probe, and 200,000 km/s
A ship passes a station at u = 200,000,000 m/s, then launches a probe forward at v = 150,000,000 m/s relative to the ship, both figures measured toward the same direction. Newtonian habit says just add them: 350,000,000 m/s, a good 50 million metres per second past light speed and therefore impossible. Feed the same two numbers into w = (u + v) ⁄ (1 + uv ⁄ c²) instead: the denominator works out to 1 + (200,000,000 × 150,000,000) ⁄ 299,792,458² ≈ 1.3338, and 350,000,000 divided by that gives w = 262,409,137.527 m/s.
That is roughly 262,409 km/s, about 87.5 percent of light speed — well short of the naive 350,000 km/s, and, more importantly, safely short of the 299,792.458 km/s ceiling. The 87,591 km/s gap between the naive sum and the relativistic answer is not measurement error or rounding; it is exactly the amount the universe subtracts so that no combination of sub-light speeds can ever add up to more than c itself.
Questions
Why can't I just add the two velocities together?
Because plain addition assumes every observer agrees on time and distance, which relativity rules out once speeds are a meaningful fraction of c. Adding 200,000,000 m/s and 150,000,000 m/s directly gives 350,000,000 m/s, above the 299,792,458 m/s speed limit — a physical impossibility. The denominator term uv ⁄ c² is the correction that keeps the true combined speed under that ceiling.
What happens if one velocity equals the speed of light?
The result is exactly c, no matter what the other velocity is. Set u to 299,792,458 m/s and v to 150,000,000 m/s and w still comes out to 299,792,458 m/s — a photon launched from something already moving at light speed is measured at light speed, not faster. This is the formula's built-in guarantee, not a coincidence of rounding.
Does it matter which velocity I call A and which I call B?
No. The formula w = (u + v) ⁄ (1 + uv ⁄ c²) is symmetric in u and v, so swapping the two input fields produces the same ground-frame speed every time. Physically this reflects that either object could equally be called the moving observer; the ground-frame result does not depend on which label you chose.
When is the relativistic correction actually negligible?
Whenever uv is tiny compared with c², which covers essentially every everyday speed. Two cars at 30 m/s each produce a correction term of about 3×10⁻¹⁵, utterly unmeasurable, so ordinary addition is fine for traffic, aircraft, and even most rockets. The correction only becomes visible once one velocity climbs past a few percent of light speed, which is why particle accelerators and astrophysics are where this formula earns its keep.
Can the result of this calculation ever exceed the speed of light?
No — that is the formula's entire purpose. As u and v both approach c, the denominator 1 + uv ⁄ c² approaches 2 while the numerator approaches 2c, so the ratio approaches c, never crossing it. Every valid pair of sub-light inputs produces a sub-light or exactly-light output; there is no combination that breaks the 299,792,458 m/s limit.
What if I enter a velocity greater than the speed of light?
The arithmetic will still run, but the output is physically meaningless, since nothing with mass or information can reach or pass 299,792,458 m/s in the first place. Keep both input fields at or below c for a result that describes an actual physical situation; the formula was derived under exactly that assumption.