I guess the proper thing to do would have been to put a sinewave out of the source and calibrate the recording levels using the meter. I agree that the gain variance can have an effect on perception, but what I hear in the difference between some of the samples cannot be attributed to the difference in levels.
So here are the sample ids:
A - Neve
B - V3
C - Neve Silk
D - 722
(BTW, when MattH first posted his list naming the preamps, I was amazed at how he nailed it - really nice to see how many people know their shit on this board)
I didn't listen to these yet on my home playback system but I've listened to them alot using Grados plugged into my laptop. I rated them C, A, B, D. On a scale of 1-6, I put them about like this: C=6, A=5, B=4, D=1. In this scale, 1 doesn't suck, I'm just trying to find a way to express the relative margins I applied to the difference between samples. I thought that the Neve silk, Neve, and V3 edged each other out but there was a clear difference to me between all samples. I prfered the silk overall, but I can understand why someone might like the Neve over the Neve silk.
To my ears the 722 clearly lagged in clarity and depth compared to the others. However, it may not be fair to compare a $1700 dedicated preamp to the preamp in a $2400 full featured unit. The 722 preamp might best other preamps in the $900 range and you have to consider the relative size of the two units. I'll never hesitate to use the 722 alone for low profile but after listening to the samples, I feel justified in carrying the extra gear to run an external preamp for open taping. Honestly, if the 722 had matched the others, I'd be very happy to sell the other pieces and always carry a single box.
I'm glad you guys enjoyed the test.