#48 Band halves swapped with cyclic spectroscopy

[Anne Archibald]

When processing baseband data taken with PuMaII (WSRT, 350 MHz) in cyclic spectroscopy mode, the top and bottom halves of each (-F produced) channel are exchanged (see attached image). I don't know what causes this, but it does not appear to occur with LOFAR data. Perhaps the difference is that the band is inverted in PuMaII data?

[Anne Archibald]

Cees Bassa has seen this effect too in other data; it occurs when dual_sideband is set and not when it isn't. This can be tested by toggling the DSB flag in DADA files and confirming that the saw-tooth effect occurs or not as described. We tested it for real and complex data with and without the DSB flag.

[Anne Archibald]

Fixed in commit abca252bdd9206d584fb711b711ee16e7c3ac0c0