I'm a little confused about how to use ARC2 when mixing. Up until now I've just been impressed with how much better everything sounds, without worrying about how it's affecting the levels and how I'm supposed to take this into account. 
 
The first thing that confuses me is the "Correction on/off" button in the VST. With the trim set to 0, the perceived level sounds much the same with this button both on and off in the project I'm now working on. But I guess I had thought of this button like a bypass switch. Not so - if I turn the VST itself off, the perceived level is noticeably higher. 
 
So that brings us to the trim control. The manual says: 
 

We have worked to make the level difference between the corrected and
uncorrected audio minimal. However, the level difference will vary depending
on the frequency content and level of the program material. To further
minimize this difference and match corrected and uncorrected audio levels
(thus being able to fully appreciate the difference between the “before” and
“after”), use this trim knob. It has a range from -6 to +6 dB, and it should
be set so that the perceived loudness is the same for both the corrected and uncorrected audio.
Once the comparison has finished, please bring back the TRIM knob to its default position (0 dB) to avoid unwanted clipping of the signal.
 

Does this mean that as long as I have the trim control set to unity, I shouldn't worry about the effect that ARC2 has on levels when I'm mixing? In other words, when I eventually turn ARC2 off for the export, the levels I've arrived at with ARC2 on will still be correct? 

I don't have ARC on the MASTER buss. So the levels on arc don't effect the exported mix. I "trim" arc so I don't clip the signal going to monitors. I find "unity" will clip a bit. Remember it is an eq (I won't mention I think it does some phase magic as well, the last time I did that I got a nose bleed :€) I put Slates fgx on the master and try to push a square wave out the other end ;-)
HTH,
Tom

Ah yes of course, sending the master bus to ARC, that had been vaguely at the back of my mind from another thread but it didn't crystallize into a coherent thought until you mentioned it, thanks 

i create a "ARC" bus... and put arc on that, active.......
 
then i route that bus to the main outputs of the audio card.
 
i temporarily send the outputs of 'Master' bus, to the ARC bus, to mix.
 
then, when i output or bouncedown, i turn arc off, put the 'ARC' bus output to NONE, and re-route the master back to the main outputs of the audio card.

batsbrew
then, when i output or bouncedown, i turn arc off, put the 'ARC' bus output to NONE, and re-route the master back to the main outputs of the audio card.

The question has to be... Why?

The whole point of an ARC bus between the Master and your output is that you can leave it ON while you export from the Master...

i turn it off and bypass it, because of the load on the cpu, and also my old school methodolgy-
always simplify a hard wire path, software path, or anything else that can affect the math.
 

Shortest distance between two points is a straight line; Bats' method makes some sense.
I just export (or bounce) from the MASTER.
Line of least resistance, that's my creedo (meaning easy) not always "best" admittedly.
T

well, i don't have a very powerful PC, and any plugin creates a drain, sometimes to the point of fatal errors with the program.
i have to tweak the samples for recording, playback, mixdown, etc.

I totally get it.
Long after I had some horsepower I continued to bounce 2 track "tracking" mixes.
Even with the "Cat", I pretty much freeze all my synths and Amplitube traxs.
Looked at your sig, time to get into the 21st century Bat
 
T

Sharke, here's how I do my stuff. It doesn't totally address your problem, but it works for me.
 
I use a master bus in each project. That bus is set to my main soundcard outs 1/2 during an export.
 
I create 2-3 buses that have an instance of ARC with the speaker correction I need.
 
My master bus outs during mixing will be sent to the ARC correction of my choice so I can hear the right monitors in an instant.
 
This comes up as a template for me. Before I saved it as a template, I ran test tones through the ARC buses with each different correction going on. Instead of messing with the trim on ARC, I adjusted the bus ARC was on in Sonar. I'm not sure how accurate that meter in the plug is and to be honest, I really don't care as I never look at it or even open the plug. What I do watch is my Sonar bus meters and make sure there is no clipping or weird sounds going on. 
 
When I'm read to export, I just set my master bus out to my main soundcard 1/2 and never worry about ARC. I usually play an entire pass of a mix through my master bus without ARC just to see what my final output levels are. I'm usually a little under -3 dB so I'm in good shape.
 
The test tone noise I used was inside my PSP Xenon limiter. It has some handy metering and tools in it so I used them to see how things were looking.
 
So to answer your question....
 
"Does this mean that as long as I have the trim control set to unity, I shouldn't worry about the effect that ARC2 has on levels when I'm mixing? In other words, when I eventually turn ARC2 off for the export, the levels I've arrived at with ARC2 on will still be correct?"
 
In my experience, I have NEVER gotten the same levels with ARC on or off. They are always different even at unity.
 
I pay no attention to the ARC trim and adjust the bus ARC is on. That's what I set using the Xenon test tones. This way I don't ever open ARC. The same is true with ARC 1 for me...which is my preferred weapon of choice over ARC 2. Hope some of this helps.
 
-Danny