DavidCarlyon wrote: ↑Thu Aug 01, 2019 12:13 pm

One thing i would say - i always have ot check my corellometer when using precedence demo, as sometimes i end up with various frequencies being way out of phase. Don't know if its how i am using it, but will try again.

This is shaping up to be a really powerful combo. Great work

First some notes about mono-compatiblity:

---------------------------

1)

---------------------------

Most mono compatibility meters and things such as the "corellometer" are not actually really completely accurate for the explicit task of measuring mono-compatability. What I mean by this is simply most of these meters are normalized in such a way that gain differences between channels are ignored and only the absolute phase difference is reported. Users will look at such meters and see a strong negative value and panic, but often the result is actually NOT particularly problematic.

If you have a full scale sine wav in the left channel and an identical sine wav int he right channel, such a coloration meter will show +1.0. If you sum them together using 0.5 * (L + R) you get a result that is perfectly identical to either/both input channel.

If you have a full scale sine wav in the left channel and an 180 deg phase inverted sine wav in the right channel, such a coloration meter will show -1.0. If you sum them together using 0.5 * (L + R) you get perfect cancelation resulting in silence in the output. Of course this is a problem and this is something to panic about.

But, what if the left and right signals to NOT have the same gain? Let's say you reduce the right channel by 60dB due to applying some gain panning. Effectively you have have:

L = 1.0 * sin(x)

R = -0.001 * sin(x)

Your correlation meter still shows -1.0, and you panic. But it this situation really something to panic about? Is it really still not "mono compatible"?

If you sum the channels you have:

0.5 * (1.0 * sin(x) - 0.001 sin(x)) = 0.5 * 0.999 * sin(x) = 0.4995 * sin(x)

if you hard-pan the signal you will have:

0.5 * (1.0 * sin(x) - 0.000 sin(x)) = 0.5 * 1.0 * sin(x) = 0.5 * sin(x)

The correlation meter will show +1.0 for a hard pan signal, but -1.0 for an almost hard panned signal with phase inversion. Yet the summed result is almost identical.

In effect the correlation meter is not particularly useful in this case.

Correlation meters typically ignore the gain difference between the signals, yet this is VERY relevant information when considering true "mono-compatability". I am not overly convinced how useful they are for this task.

A MUCH better tool is the:

Goniometer/PhaseScope/VectorScope/Lissajous Vectorscope

such as:

In such a meter "real problems" are easily found when the display starts to approach a horizontal line. If the display generally shows something that is at least slightly taller than it is wide, is not a problem.

---------------------------

2)

---------------------------

In electronic/dance/club/urban/bass music mono compatibility is mostly a concern in the bass range. Precedence retains full mono compatibility in almost any potential parameter settings in the bass range. If you work in these genres, you generally do not have to worry if that if you main concern.

---------------------------

3)

---------------------------

We would really advise to simply put a "mono-maker" plug on your master output and check the final mix in mono with your ears. Something is ALWAYS lost when summing to mono unless you are summing two signals are are identical, which means they are already effectively mono. A hard-panned signal will loose 6dB when summed to mono. A 90deg phase difference between channels will lose only 3dB when summed to mono. Generally well behaved phase difference between channels are not a large problem when summed to mono. Exactly 180 deg phase differences that are ALSO exactly the same gain in both channels are really what to be concerned about, and Precedence does not typically create this situation itself.

---------------------------

4)

---------------------------

we are considering making some kind of metering tool to help show what is and is not something to be concerned about in our estimation to help make these topics more intuitive. We are not yet completely satisfied with the various tools available on the market to check these sort of topics.