Satin questions

[Agility]

Hi, got a few questions for fellow Satin users.. these pertain to using Satin to emulate classic tape machines, not create new weird interesting effects.

I've had it for 2 years now and still figuring out how to get the best from it, so many options felt so good initially but now I feel it's almost a curse.

What do you guys feel like is the optimum bias range? Usually I'd have it around 2-3 but lately tried it all the way at 5 and everything did sound very tape like but very dull too.. way OTT? I think 3 is a good number then using the pre emph and gap width to flatten it out...

Also I was reading up about biasing, I heard the way to calibrate it was to overbias you'd raise the bias knob until the curve would rise and then descend again and as it descended, it was overbiasing, is this the same way it works in Satin?

Also, is the headroom knob related to the terms like when people would say they were recording at +3 +6 +9 etc ?

..also(lol), can someone explain what the 0vu meter is all about, I read the manual but too be frank I'm not good with the whole db stuff(to my detrement, ofcourse) but for example, I use Satin on my kick channel and I've been hitting it at around 0db to get that saturation, the knob was always set at -12db but I've been playing with setting it at 0db, this is obviously means driving it harder right.. and then it would appear that hitting it with the meter set at -12db isn't hitting it at 0db.. so which is the 'correct' one, the knob is just adjusting the metering right, not any part of the signal?

Hope that makes sense and someone can enlighten me, some of these concepts go over my head, or maybe they're just not explained in a way I can interpret easily.

Lastly, do you guys pay attention to keeping the curve flat, for example I use Satin on drum/music/masterbuss(Can't afford to use it on each channel unfortunately) and I aim to keep the curve looking flattish(you know, along the lines of http://www.endino.com/graphs/) and what happens is my levels are hitting at like -20 ... from what I understand, that's wayy to low, so should I once I've got a nice curve, then turn the input up until it's glueing the sound(but not into the red) and disregard the shape of the curve?

I used to do it that way when I first got it but then switched to aiming to keep the curve flat.. which does still give a 'tape'ish) sound, I guess but does sound cleaner... which would be more accurate, despite me using it only on the busses?

Any info appreciated, thanks, I feel like I should be an expert at it by now, I spend pretty much every day toying around with it in my projects trying to find the optimum setting but still feel like I'm wandering blind sometimes..

[sascha]

On vacation right now, so just a short comment:
The frequency graph uses a dirac for taking constant measurements. This is fast and causes no visual ripple, as opposed to sine sweeps. But dirac is not ideal when things become nonlinear, therefore the error rises towards DC on the graph at high levels. We might perhaps add sine sweeping later, but atm the curve is most trusted at low-to-medium levels.