Debugging bad incoming data

hi, I am using Siemen’s Prisma scanner for hercules data acquisition. Our new incoming data look extremely bad, broad linewidth, despite a reasonable FWHM value during manual shimming (~18). We do click ‘apply’ after manual shimming and to our knowledge the scanner doesn’t appear to re-shim before we start acquiring data. Any advice on first steps I should take to track down what’s happening here?

important context: i recently switched institutions and this issue is arising on the new institution’s Siemen’s Prisma. Previously, while using a different institution’s siemen’s scanner, we successfully collected reasonable looking data in ~100 subjects following the same manual shimming procedure

Hmm - is it possible that the shim volume you’re optimizing the linewidth on is accidentally different from the MRS acquisition volume?