Yeah @atmasphere, for pre-amp applications, Roger had a hard-on for the 6DJ8 (used in his RM-5 and the Audible Illusions Modulus 2 & 3). He was a man of "firm" opinions. ;-)@bdp24 We used the 6DJ8 in some our early stuff, since the tube is very linear (which is probably why Roger liked it) and you could get good bandwidth as it supported high currents.
But finding one that wasn't microphonic was really challenging! I often found examples you could yell at and hear yourself in the speaker- and that was in an amplifier!
The 6DJ8 was not intended for audio- it was more for instrumentation (look in a Tektronix 'scope sometime- the tube ones are full of them) and television work. OTOH the 6SN7 *was* designed for audio, and like many products there were hiccups along the way but they did get the design right. When we started using it the only source of current production was Russia and they were dreadful. But then the Chinese got in the game and their 6SN7s were pretty good! Since then there has been no turning back and we use the 6SN7 in all of our products.
When we came out with the UV-1, instead of using the 12AU7 (which is based on the 6SN7) as we were doing in our modifications of the Dynaco PAS-3, since we had the room and the current in the power transformer, we opted for the 6SN7 as the centerpiece of that preamp.
What's nice about the 6SN7 is that if a line stage design is at all competent, it will easily keep up with line stages costing a lot more. It does not have to be a complex circuit- in our UV-1 there's only one tube per channel and 4 resistors!
Audio Research makes respected preamps, but I would expect many of the 6SN7-based preamps on this thread to keep up with any of them. This is not an attack on ARC, its simply a statement of how well 6SN7s work in line stages!