Here's the can of worms that I fear we may be opening. This whole realm of acoustics is rife with changing curves when it comes to graphs of relatives loudness. Most of us have seen a Fischer-Munson graph showing the perceived loudness of different frequencies at different volumes. Actually, it may be Fletcher-Munson. One of them, anyway. The point of it is that a perfect mix, in which all parts are intelligently balanced and the whole mix makes perfect sense, will become a less perfect mix as soon as you change the volume even by a little bit, but it becomes a terrible mix when you change the volume by a lot.
Therefore, if you mix with your monitors turned up to rock-concert levels, someone who listens to your work at normal household levels will not hear the same mix. There is no fixing this problem. It's a feature of being human and having normally working ears.
So, when we have faders and multiple tracks, we fool ourselves into believing that we've found the perfect mix, with all our perfect automated movements, when in fact it all goes out the window as soon as someone adjusts the volume so that it no longer reflects the sound levels heard by the mixing engineer. This isn't speculation; it's the law. Well, I don't know if Fletcher and Munson ever made any laws, but the graph represents something very real which they set out to observe in the interest of improving telephone communications.
Here's one I found on Wikipedia:
Notice that it's not just one line representing sensitivity to sound pressure, but that it implies that every pitch has its own curve. Further, they probably vary from person to person.
Ok now that I've made THAT point, how do you translate these curves into fader movements? And if you can, then how do you make relative fader movements?
To put it simply: any adjustment you make is only meaningful at one pitch OR at one dynamic level. Change one of those, and you may be able to compensate, but mere fader movements cannot compensate for both.
I've wondered all along if MOTU had attempted some sort of compensation. The mere use of a logarithmic scale only compensates for the logarithmic nature of deciBells. When you start making parallel grouped fader moves that are separated by any amount, they are going to have to move by different amounts to maintain "relative loudness." But if you're talking about tracks with different pitches — say, a bass track and a female vocal track — then all bets are off for maintaining any kind of relative loudness through some grouped moves. You'd have to do some digital signal processing with look-ahead pitch monitoring to make this accurate.
For this reason, in my original video on this subject, I called motorized fader movements a convenience. GROUPED motorized faders are also convenient, but they cannot be done accurately without complex computer processing, which is why I said that MOTU simply had to choose a method and go with it. There is no "right" method.
Of course, I got laughed at for talking about theoretical aspects by certain folks who know who they are. For example, 2damnhip wrote:
- I’m really not all that interested in theoretical explanation. I’d have become an electrician if I wanted that. I want all the faders to move by the same db amounts..up or down. +1 for all, not +1 for one of them , +.56 for another etc...crazy.
But I do ‘pologize” to Shoosh because he’s the boss.
An electrician?
Ok, whatever. But it should be obvious that what 2dam and others want is pretty much impossible through mere fader groupings. My advice to everyone would be not to depend on grouped faders to maintain any kind of consistent loudness relationships through their movements other than in a very general way, in a narrow range of values. If what you want requires accuracy, you'll have to do it by ear, unless someone comes up with a complex look-ahead plugin that calculates pitch and loudness and adjusts them to the Fletcher-Munson scale while your faders are moving. That's probably not going to happen.
This is why I've found this thread frustrating. None of us really knows what's going on, and while it appears that
MOTU should fix this for consistency, in reality it's not going to make a lot of difference for reasons described above.
Those are my opinions, and I return to my position from my first video: this is a convenience; not a scientific method of maintaining accurate sound-pressure relationships, because those go out the window as soon as you move those faders, AND as soon as your listener changes the volume level at which it's being played back on their end stereo system.
Some of you will get what I'm saying. Others will ridicule it and suggest I save it for electricians [sic]. But even if MOTU fixes this, we're really not going to get what we want, because it's not possible. It's an interesting topic, and it's been fun chasing down the apparent problem, but I remain skeptical that fixing that will really "fix" anything, though it will look better to anyone watching the numbers. This doesn't mean I like the way it is, or that I choose any side of the debate. It means that I think the whole debate is kind of pointless in the end, because the work of Fletcher and Munson for AT&T pretty much guarantees that it is.
Shooshie