Check out this discussion about the Gold:
http://forum.cakewalk.com/tm.asp?m=1543560
Seems there is some debate over the legitimacy of the ART MPA Gold's
tube claims...
(You have to read about half to three quarters down)
Well, I think they are kinda confused about tubes, honestly. The engineering analysis is fine, but let's step back a minute. They seem to say the goal of a tube is to increase even-order harmonic distortion. Then they criticize the circuit for running at +150V (I have measured higher, however, so the schemo might not be 100% accurate). Then they criticize the circuit for using feedback. Hmmm.
First off, it's incorrect to say that 12AX7 was not designed to run at +150V. That's just plain false, and reading the datasheet will illustrate that. In fact, the datasheet I have (Sylvania) specifies operation at +100V and +250V, with +300V the maximum.
You probably do get an increase in linearity as you increase plate voltage, but I don't know how much.
Next, let's consider the goal of a tube in a design. If the only goal is to increase even-order harmonic distortion, you should go straight to starved-plate at +45V, which ART uses in their other designs. The failing of those designs is not the tube--it generates pretty good sounding harmonic distortion just fine (well, after a tube swap)--but the surrounding solid state stages sound horrible when they are clipped.
OK, so if you use +300V, you likely have a little less THD than at +150V. But either way, you'd have a LOT less than at +45V. If anybody on that thread had measured THD of the ART, they would know that. I have done this often myself.
Now that we have set aside the primary goal of a tube design to be harmonic distortion, it seems straightforward to design a circuit that minimizes such distortion. Indeed, you will often hear that the goal of a "real" tube mic preamp is not to generate distortion, but to minimize it, with the goal of realizing very high headroom. This is usually a statement made to denigrate starved-plate designs.
Now the ART, as noted, will still be restricted by the headroom of its +/-15V supply to its following opamps. So there is definitely a small range between where the tube (on high plate) starts to generate THD, and where the opamps clip. But I would argue the preamp was not designed to operate in that range. In fact, the VU meter set on tube is supposed to be used to help avoid that.
This can actually be a somewhat clean preamp, and is best that way. There are tubes in it, they run at +150V (or higher), those are the extent of ART's tube claims. You are free to buy a Brick if you want a "real" tube pre. Mind its input noise though.