Richard Wall wrote:
The other explanation is that the device you used for comparing the
frequency response and phase responses was not sufficiently resolving to see
the differences. I note you say almost identical so there was a difference
?
I note Stewart has pointed out that you cannot test without LMDBT. It is
most important that levels are exactly matched but I do not agree on the
merits of DBT.

I dont see how you can fail to agree on the merits of DBT - it removes
both the testers AND the subjects bias.