To measure very weak magnetic resonance signals without new hardware, we changed how data are collected. Rather than one long sweep, we recorded many short, overlapping pieces, cleaned each piece, and then merged them. On standard instruments and with equal measuring times, this increased usable signal compared with background by about 5 to 10 times across three test samples, helping reveal details that long averaging can hide.