... Following up on a previous question about processing interleaved gNhsqc_IPAP.c data where IP/AP is on the inner loop with array=IPAP,phase setting.

When the processing script has:

nmrPipe    -in test.fid \
... usual X axis processing ...
| nmrPipe  -fn TP \
| nmrPipe  -fn COADD -cList 1 -1.1 -time \
...

We have a normal spectum:

But when in the same script we move COADD downstream to X axis processing stage and set -axis Y:

nmrPipe    -in test.fid \
| nmrPipe  -fn COADD -cList 1 -1.1 -axis Y -time \
...

The data looks like so:

I'm a little confused - I thought that -axis Y switch on coadd is equivalent to "transpose, then apply COADD, then transpose again"? What is the difference between the above two ways of COADD'ing?

Thank you!