On high-end console, fader linearity is not much of an issue, but on mixers in the budget to the project studio level, fader quality can vary and usually the frequency response is NOT linear for the entire path of the fader pot. This means that putting the fader at the say -20 mark results in a different response curve for the the signal than at it's intended range (near the 0db mark)......
Also... the idea is to keep your signal levels wihtin the intended design limits of headroom. Putting the faders too low means you underutilize the summing opamps and lose in the S/N department. Push the faders too high and you push the opamps too hard, causing harsh distortion and graininess.
Incidently, I recently discovered the answer to an inconsistency I've always wondered about - it usually came up when Ed would mention headroom issues with Mackie mixers versus other consoles... He knows what he's talking about as do I, and yet, I've never heard the problems he's encountered in my own gear.
A good part of that contradiction is related to gain structure and not so much any limitation on mixer design. The issue relates to the signal levels the mixer is designed to operate on (+4 dbU in this case) and the digital levels many of us are using.
We all know that for digital recording, we try to get maximum bit count in our recordings, meaning we keep signals as close as we can to 0dbFS to ensure max. resolution... fine... what is not immediately obvious is the correlation between what that means in terms of actual signal level between the mixer and the digital recorder.
A typical 8-buss project-studio board, say Mackie or Ghost is designed to operate at +4dbU.... which means the board's meters are calibrated so that 0db represents the +4dbu ratio 1 volt. And for the Mackie - it allows 28db of headroom above that - a significant amount, actually...
Great... so what, you say.... well - translating that to our digital recorders meters we might mistakenly assume that 0 dbFS on the digital recorder corresponds to 0VU on the mixer meters given that the +4dbU gain structure is maintained to connect between the 2. And unfortunately, we'd be wrong......
The fact is that digital meters have been calibrated in much the same way as analog meters were (for historical reasons and to allow engineers very familiar with analog metering to feel comfortable in a digital setting). This means that when we send a signal that reads 0VU on the mixer over to the recorder, it comes in at the -15dbFS reading on the recorder's meters.
That in itself isn't a problem, except for the fact that we all want to maintain maximum digital resolution, so we really want the signal level to be up a good 10dbFS more at least to make sure all the bits are getting used.......... So we boost the output level of our mic pres (or whatever we're using to track with) to get the level in the right digital ballpark.......
OK... great - we've got max. digital resolution and things are sounding great.... no digital overs - we're laughing! Until............
...we wanna mix... we set the master fader at 0db - its optimal point by design and convention... we set our channel faders as needed... and we notice the signal is generally above 0db on the mixer main meters... actually - it's peaking quite a bit between +2 and 4db.... and we're not even trying! OK... vocal needs to come up now we're peaking between +4 and +6db...
Ok... it's still sounding good... until we look at our digital 2-track (oh, let's say a Masterlink for example) -- digital meters again, and our hot signal is only hitting between -15 and -12dbFS on the 2-track! Gotta watch that digital resolution again........ bring those levels up some more! (and incidently, we're not even close to commercial cd levels at this point!)
Well... need more level out of the board, so we bring the faders up more, bring the master up more and get ourselves up another 4-6 db....
But what's really happening at the board at this point? We are already a good 10db above the nominal design of the the input channel level due to our getting the signals to tape at the requisite hot digital level... so our input channels now have only maybe 12-14db of headroom left... the summing opamps of the master buss is starting to feel the push, because the combined signal level of all those already hot tracks is really eating into its headroom, so that last 4db we want to get out of the board to send a hot mix to the 2-track really grinds the output headroom right down... and suddenly, your mix sounds like crap - distortion, graininess, muddiness.... all the characteristics that we attribute to lowly gear, when in fact, we're simply using it beyond it's intended usable signal range! (mainly because the although our mixer/recorder inputs are matched, the metering is not calibrated the same way, and desired digital levels do not corrolate to the design levels of the mixer!)
So what do we get out of this??? Since the scenario I described will not apply to everyone - it amounts to you really having to know and understand the signal levels that you're dealing with between recorder and mixer and to compensate or adjust accordingly in every phase of the recording process (tracking, mixing) -- otherwise, you're basically shooting in the dark as to what output level you're ending up with.
As a matter of fact, you not only need to know and understand your signal levels, you also need to have a good understanding of the levels your mixer and recorders are designed to handle and adjust your methods to get the best use out of them......
Sorry to interrupt your thread Vurt, but this was kind of related to your question.......
Now back to your regularly scheduled program.....
Bruce