Adding currents; PCM1794A in mono

[Quince]

Regarding the thermal memory distortion, it would make sense that it would be worst with opamps. However, on another forum I've seen the claim that if the opamp is in used as unity-gain or low-gain, it wouldn't be an issue. However, this doesn't seem right to me, as the open loop gain remains high, and the gain reducing negative feedback would have limited ability to correct for a dynamic distortion of this type.

 

[Quince]

I wonder about using Baxandall Super Pair. That's discussed in an 1988 paper by Hawksford: http://www.sx.ac.uk/ese/research/audio_lab/malcolmspubdocs/J10%20Enhanced%20cascode.pdf
Another interesting thing in that paper is the improved cascode by a capacitor from the base of the cascoding transistor to the emitter of the other one.

Then there's this strange schematic of a VAS I found in an old thread at diyaudio: