Hi @lewm
"That’s often tolerable but also there is a gradual high frequency roll off."
The statement quoted above will be true if the cartridge is an MM or MI with hundreds of mH inductance.
In contrast, MC cartridges generally have far less inductance (by a few decimal places), and therefore will not behave as claimed.
For example, if the frequency response of a Denon DL-103R (14Ω coils) is measured when the load is 47kΩ, then 47Ω (a thousand times less), the two curves will track each other to within a fraction of a decibel (at the most 0.5dB at 20kHz).
Keep in mind that a low-inductance MC cartridge is a terribly inefficient passive generator that is mechanically driven by the LP groove. As a result, its behavior changes hardly at all when the electrical load is altered.
As a handful of engineers (including Ralph and myself) have been saying for many years, the sonic differences perceived by many audiophiles is caused not by differences in the behavior of the MC cartridge as its electrical load is altered, but rather distortion in the phono stage that is triggered by the electrical reactance / ultrasonic resonance occurring between the inductance of the cartridge coils vs. the capacitance of the tonearm internal wiring, tonearm output cable, and input circuitry of the phono stage.
If a phono stage designed to be insensitive to ultrasonic or low radio-frequency energy, the tonal balance will change very little when the input load is altered, with the side benefit of reduced LP surface noise (much of which occurs within a frequency range that overlaps with the ultrasonic resonances occurring between the inductance of the cartridge and the various capacitances present between the cartridge and the amplification circuitry of the phono stage).