Does that solve ground loop issues though? Other than using optical coupling of some kind, I can't think of an alternative to a transformer for that.
With sufficiently well matched resistors, the CMRR of a differential amplifier is more than good enough.
I will definitely try that (thanks @David Hess as well).
The audio transformers sold for this purpose on Amazon did not measure well... to put it mildly! It shouldn't be hard to make an op amp work better than that...
If you need better rejection, then 0.1% resistors are inexpensive, but the AC common mode rejection becomes more important and will need to be adjusted using a trimmer capacitor. I have only seen instrumentation go to this much trouble. Audio applications do not often need this level of performance.
Below is an example of the common mode rejection of a Tektronix 7A22 differential amplifier which has performance limitations from the same mechanisms as an instrumentation or difference amplifier. The fall in CMR at low frequencies is caused by using input AC coupling capacitors. If AC input coupling is required, then there is another way to do it. The lower curves (10^3 is 60 dB) are similar to what would be seen with an audio instrumentation amplifier with 0.1% resistor matching. The fall in CMR starting at 1kHz is caused by mismatch in the AC CMR and shows why adjustment with a trimming capacitor would be required. 1% resistor matching would put the CMR below the bottom of the graph at about 10^1.7 or 50 where no trimming of the AC CMR would be required.
Good audio transformers have their own limitations which makes using an instrumentation or difference amplifier a very competitive solution if galvanic isolation is not required.