Watch magnetisation tip, precess, fan out and relax, following the Bloch equations. Pick a pulse sequence and run it, or tip the spins yourself. Drag the sphere to turn it.
Rough values at 3 T.
The sphere is drawn in the rotating frame, which turns with the spins at the Larmor frequency, so spins exactly on resonance stand still. B0 points up the z axis. Each thin arrow is an isochromat, a group of spins that see the same field; the thick one is their sum, M, the magnetisation you can measure. Its part in the x′y′ plane is the signal a coil picks up.
The spread gives the isochromats a Lorentzian range of frequencies, the way a slightly uneven field does. They fan out and the signal decays with T2*, faster than T2 alone. A 180° pulse flips the fan over so it closes again into a spin echo; only true T2 decay is left, which the CPMG train makes plain. Inversion recovery shows T1: the signal vanishes when the 90° pulse lands at T1 ln 2, as Mz crosses zero.
Pulses are hard pulses, drawn slowly so you can follow them; relaxation stops while they play. The lab frame adds the Larmor precession back, a turn a second rather than a hundred million.