There are many ways to do this. As described, this data happens already to be in the data order that an NMRPipe hypercomplex FID is stored. So, a simple approach might be to convert the data directly to binary, and add an appropriate NMRPipe header. According to the info above, the data input has four columns (0 1 2 3) with the actual FID intensities in column 3. And the text file should have 1024*256 = 262144 lines. For example, if the input text file is "fid.txt": getCols fid.txt col.txt 3 txt2bin.tcl < col.txt > col.bin bin2pipe -noIn -noOut \ -xN 1024 -yN 256 \ -xT 512 -yT 128 \ -xMODE Complex -yMODE Complex \ -xSW 10000.000 -ySW 10000.000 \ -xOBS 500.000 -yOBS 500.000 \ -xCAR 4.700 -yCAR 4.700 \ -xLAB F2 -yLAB F1 \ -ndim 2 -aq2D States \ -out hdr.dat -verb -ov cat hdr.dat col.bin > test.fid