Mixed-mode reflection from 2-port files

The same 0.5 combinations can be written on S11 of four 2-port files to produce SDD11, SCC11, SDC11, SCD11. Treat that as an approximation.

Why it is approximate

Each 2-port S11 was measured with a different termination on the ports that were not connected to the VNA. A true SDD11 is the differential reflection with the other mixed-mode ports terminated in the mixed-mode reference. Four separate 2-port setups do not reproduce that.

Transmission (S21 of each pair → SDD21) is much less sensitive to the unused-port termination when those ports are well-behaved 50 Ω loads. Reflection is first-order sensitive.

What to do

Need	Measurement
SDD21 / SCC21 / CMRR for a cable	Four S2Ps or one S4P — both are used in practice
SDD11 / SDD22, fixture return loss, cascading	True S4P (4-port VNA or vendor-reassembled switch matrix)
Quick “does this launch look broken?”	Approximate SDD11 is a screen, not a compliance number

If you publish approximate SDD11, say so in the dataset name or a parameter (approx=s11-mix) so a reviewer does not treat it as a 4-port result.

Full-matrix conversions

When you do have an S4P, use the library conversion (scikit-rf s2sdd / equivalent) rather than four S11 files. The identities are the same; the data underneath is not.