The 6922 mod is not complicated; just wire the heater power to pins 4 and 5 instead of 4/5 and 9 for the 12AU7. The other pins (1,2,3,6,7,8) remain the same. Leave pin 9 floating. Once done, the 6AU7 is a drop-in replacement for the 6922, if you want to compare over the long term.
However, the 6922 has a greater need for caution about oscillation; I would use grid and plate stoppers as in the Reduction - of course I AM a very cautious engineer! It will make about as much difference as Integration and Speedball combined, at the cost of an inch or two of solder plus a 6922 or 6DJ8. If you try it, use any old 6922 or 6DJ8 available cheap; if you then like it you can explore better tubes for the fun of it. I did tell you it's a can of worms, didn't I?
The mismatch of analog and digital loudness is a very frustrating problem, which has come up now and then ever since the original Seduction 12 years ago. In theory, and with well-recorded music as documented in the professional audio literature, and compared to digital sources that follow the Redbook standards, Reduction should make about the same loudness level as digital sources. Far too often this does not seem to be the case, and I do not have enough data available to me to determine why. Here is a short list of the three most plausible explanations:
1. The phonograph recording is made at a substantially lower "maximum VU" level than the standard 5cm/sec velocity
2. The digital full-scale (FS) output exceeds the Redbook standard of 2.0 volts RMS
3. The digital recording is more aggressively compressed than the phonograph recording
There is a lot of ranting on the web about excessively compressed digital audio, but I've never seen a good statistical analysis to quantify the problem so I can't say if that's the main issue or not. As I said, very frustrating for an engineer!
