It would be pretty easy to write a “Nyquist script” to measure the positive and negative extremes, provided that the track being measured is not too long (up to about an hour should be straightforward),
Example as a simple script that can be run in the Nyquist Prompt effect:
(this version is for mono tracks only)
It is intended to change the audio signal phase to get the positive peaks greatest than the negatives.
Can we ask what the goal is? I mean the real goal?
You can’t directly compare successive swings of the waveform. The positive and negative swings are never there at there same time. So then you get to decide the time interval. More positive than negative over what time interval.
I have exactly one thing in common with one professional presenter. She and I have profoundly non-symetrical waves when we speak. I can recognize her presentation mixed with some of her other presenters just by looking. No sound needed. In our particular case, all you need is Effect > Invert once at the beginning and almost all of our waves will be formed more up than down.
That’s borderline silly, which leads me to ask why you want that.
There is one side issue in case you’re going to go all scientific on us. Audacity drives scientists nuts. Audacity is a sound editor and sometimes takes actions that while not scientifically accurate, sound petty good. Because that’s our goal.
By experience only (nothing about science), I know best sound is obtained when the positive peaks are larger than negative.
The real world audio is asymmetrical, almost all the time.
It is the main reason because it is a very good idea to preserve the original audio phase polarity and, I guess, it is the reason because all microphones, cables, connectors, etc. have a + and - signs.
But…, there are times when the negative peaks are larger than positives. Specially with brass and percussion instruments.
Whatever the reason is, not always we can “see” it, easily, into the graphical interface.
Also… There was some audio processors, in the past (analogue audio world), with the capability to compare positive and negative peaks and change (invert) the polarity, if the difference takes too much long time. Could it be some plug-in, in the future???