[Trevor]

On 1/01/2018 1:14 AM, Adrian Tuddenham wrote:
geoff wrote:
And another thought (though not necessarily specific to the particular
conundrum - what is the output load impedance specification of the sig gen ?

That is a very good point. The sig-gen was 600-ohms output and ran
through one of the distribution rack buffers which again delivered from
600 ohms. I seem to remember that I wondered about that at the time and
repeated the test with a completely different low impedance driver - but
although I have forgotten the details, I'm sure it made little
difference.

It's always nice to be "sure" about things without any proof. :-)
Feeding headphones from a relatively high impedance source (Vs the
actual headphone impedance) is well known to cause FR irregularities if
the impedance of the headphones varies as is usually the case. A common
problem with some headphones when fed from a power amp with a simple
series dropping resistor as often used. Anyone with a basic
understanding of ohms law would realise that of course. Easy enough to
check if the dip in response corresponds to rise in headphone impedance
though.

Trevor.