AE, could you by chance list what processes you went through to create the test tone, and perhaps what options were enabled
All I did was load my CS2 session (which has all of the default settings I use), go to the Offline panel, uncheck Sync, and hit Generate. What that does is take the same tone it would use in Online mode, and puts it in WAV form on your harddrive. So instead of NAT playing the WAV, recording the output, AND deconvolving, all it does is give you the WAV so you can process it and bring it back, and all NAT does is deconvolve.
So, that's all. CS2 -> Offline -> Generate.
This is the file it generates, and it's the file I sent out to everyone, and when they send it back, all I do is load CS2, go to Offline mode, uncheck Sync, load their recorded file, and hit deconvolve. That's how you have all you have today.
But, that's all with the CS2 default settings. With CS2, I came extremely close, so I stopped moving around too much on it. I kept the settings the same for the most part, and if I found any differences by changing small settings, they were so insignificant that I forget I even did it by now. But that doesn't mean I did everything there was to do, you know? The possibilities are limitless. There are still a lot of settings people can toy with, and I'll dive into that a bit here..
Lemme start by saying that I start with NAT and no session. Giancarlo began helping me while we were still trying to make Nebula match impulses, remember? So when I started trying to do dynamics, I created a cab session from scratch. That's
Cab Session 2.
Opening that in NAT gives you pretty much a blank slate. Click into the Expert panel, and I'll go over a few things..
You'll see that I used the template Template_dynamic_1h (10K), seven 15-second sweeps with a 2-decibel variance between them, and only repeat this process once. This is why the
test tone WAV file is seven sweeps long and are quieter each time.
Now, what would happened if we used ten 20-second sweeps with a 10-decibel variance each time?
I have no idea. But this is a modification to Cab Session 2 that ANYONE can make and try. All you have to do is edit the Length, Repeats, Steps, etc fields in the Expert panel in NAT. If you only edit those things, you'll STILL be using the same test tone, and NAT will analyze and deconvolve them in the SAME way, but there will just be more or less or louder or quieter tones. You will still retain the integrity of the cab programs.
You'll also see that I used Enhanced mode, Mono, and 64-bits. Changing these may give you a different sound, but I'm about 100% confident you'd be taking a step in the wrong direction by doing so because of my workings with Giancarlo. For Kernels, I've set the number at 5, and when I went to 10, I noticed absolutely zero difference except for a file that was twice as large. Maybe it can make a difference to have more or less, but I'm pretty stoked about the sound we're getting now, so I'm cool with it.
I wouldn't even touch anything on the Envelopes tab.
Honestly, that's all there really is to it. If you JUST load CS2, you have exactly the same settings I use for both online and offline sampling. The only changes I would ever make are described above.
+1
list us some info; do you need a reamp device to send the test tone through your setup?
No sir. Reamp boxes are used to change the impedence of the signal from line level to instrument level so the input of your amp can be driven hard. With sine sweeps, you want to BYPASS the preamp section, because it will sound silly and the end result is bad. So, you tap directly into the power section by going into the FX Loop Return. Your power amp (and ANY power amp) is looking for a line level signal (that's what comes from your preamp), so you are perfectly perfect to send a cable from your interface to the FX Loop Return of your amp.
But BE CAUTIOUS. By doing this, you'll PROBABLY be bypassing your master volume, depending on the amplifier of course. What that means for you is that if you just run the sine sweep out into your FX Return, it's probably going to sound like a volcanic eruption in your room. This is NORMAL, and your amp and speakers CAN take it just fine, but you probably don't want to do it that loud. You have two options to turn it down.
If you're sampling offline, you're sending the signal out FROM Cubase or Protools or something. Just turn the sine sweep signal down here on its track properties.
Now, on the Online Sampler, there's a Gain field. What that does is change the output level of the sine sweep(s) in the same way that the volume control on Cubase's track properties will. Generally, you'll get between -50 and -20.
I would think we need say 5 sweeps with 1db (0, -1,-2,-3,-4 db) and then 4 with -5 db (-9, -14,-19,-24 db). Total # of sweeps is 9, b/c nebula can use up to 9 kernels. 
