Pan law: equal power vs -3dB

Hi All,
In the cubase project setup one can choose out of several pan laws. I don’t understand the difference between the ‘-3dB’ law and the ‘Equal Power’ law.

The equal power law gives equal power regardless of the panning setting (thus, the listener will not experience any level change).

The -3dB pan law increases the level of the mono downmix by 3dB but does not affect the acoustic power (thus, identical to the equal power law).

Thus, how is the -3dB crossfade curve different from the equal power crossfade curve? I have attached an excel sheet (zip) in which I did the calculations. The tabs with Square-law and Sin-Law are both equal power but at the same time do the same as the -3dB law.

Hoping for some insights.
Antoine

PanLaws_v2.zip (54.6 KB)

I don’t understand. What do crossfades have to do with the pan-law settings?
The pan law lets you decide how many dB’s you will decrease a sound if you pan it far left or right.
The crossfades are simultaneously messing with the given level for left and right.
Am I missing something?

If you pan from left to right then the volume in the left channel decreases and the volume in the right channel increases. See for example https://www.youtube.com/watch?v=vMgCSGI45Cw or Loudness Concepts & Panning Laws

If a signal is panned hard left or right, the power of the sum of the channels drops by 3 dB.
But again what does this have to do with crossfades?
I can’t understand your initial question

You’re not doing a crossfade, it’s just a straight volume reduction/boost on either side. If anything the ‘crossfade curve’ would be if you plotted out how long it takes you to turn that pan knob from one side to the other. Otherwise, there’s no ‘fading’ going on once the panning is set. Its ‘this track is at this level on this side, and this other level on this side’.

Ok, I agree that it should not be called a cross fade…back to my question…

which is “how is the -3dB crossfade curve different from the equal power crossfade curve?”
That’s not a crossfade curve, it is a pan law when panning and not when crossfading.
it is not a matter of what it is called.
they are 2 different things.
yes, it involves volume change on both processes.
Still, panning is about changing the balance between the left and right channels while crossfading is a level change for both channels simultaneously regardless of where you have them panned.

Maybe I don’t understand your question, don’t get me wrong, just trying to help.

Yeah you’re missing the point. Forget “crossfade”, that was probably just Antoine thinking it is fading like a cross between left and right. Forgotten. Done. Now we all know we’re talking about pan law, and the question was:

I think if you set your pan to something other than either full left, full right or center you will get different values. In other words at the extremes (including center) there is no difference, but during the pan between the extremes the curve is not the same.

I think that’s in the video you linked to - settings #2 and #3 @ 5 minutes 14 seconds.

Maybe? I’ll check later when I load up Nuendo the next time.

An equal power pan setting will result in an approx 6dB increase in gain when signals are panned from hard left and right to centre.
The -3dB pan law decreases the gain in the left and right signals (by 3dB on each side) as they are panned toward centre to maintain equal gain levels.

In practical use the Pan Law is only important if you are actively changing the Pan Position during the course of the Project. If you just park the Pan Position in one spot, then when you set the Channel’s Level by ear you’ll naturally compensate for whatever the Pan Law is doing.

EDIT

I’d suggest putting the Test Tone Generator (prob not exact name) on an Audio Track Insert and Pan that signal around using the different Pan Laws.

As far as I can see the various pan laws are only really relevant to the maintenance of compatibility between mono and stereo signals.
If I’m doing a mix down to two track on my desk and I want to check the mix in mono, I don’t want a 6dB jump in gain when I punch the mono button, hence the -3dB pan law (or -4.5dB, if the desk happens to be SSL).

Excellent point.

I think that’s backward.

Equal power:

-3dB:

Same.

Now with panning set to halfway between left and right:

Equal:

-3dB

There’s the answer to the question of the thread I believe.

2 Likes

I recall seeing some music engineers that worked on analog consoles saying they had a preference in pan law because of the way it behaved when panning. If the panning was static then a level adjustment is all that was needed after panning was done, but their feeling was that it was faster to get to the desired result with their preferred setting.

And of course if the panning was recorded pan law makes a difference, though again easily compensated for with level adjustments.

[quote=“MattiasNYC, post:14, topic:960506”]
I recall seeing some music engineers that worked on analog consoles saying they had a preference in pan law because of the way it behaved when panning.
[/quote]s

Yes, some guys liked the sides to be a little hotter and panning to centre dropped the levels a bit, and vice versa. There was no ‘standard’, per se.
Sometimes we had to adjust the levels when mono checking because the desk just wasn’t doing what we wanted. Not everything was maintained optimally.

1 Like

I think you might be right. Now that I’m looking at it, I’m struggling with Steinberg’s terminology a bit. They have a 0dB option as well, which is what I think of as Equal Power, which maintains a 3dB per side delta between centre and hard left / right.
But Equal Power here appears to mean maintaining constant levels between hard panning and centre. In my experiences with HW desks, that’s what the -3dB law implies, so now I have no idea what they think the difference is between those two.

And I believe that was what the OP question was.

Take a look at the last two images I posted. In both of them the mono signal has been panned Left 50, so neither center nor full left or full right. Then look at the arrows and you can see that Integrated as well as True Peak have different values. The same can be seen in the video the OP posted.

So to me it looks like the center is the same, the full pans are the same, but between those extremes the rate of change is different, albeit very slightly.

Yes, so that would imply that they’re modeling a difference in pan pot taper, which was a thing in HW consoles, although it wasn’t referred to as equal power vs -3dB. Why they would do it here, god knows.

No, 0dB:

  • No mono compatability: 6dB increase in signal
  • Power (our perception of level): 3 dB increase

@5 min 14 secs you see that is says 'Constant Power Law (-3dB) . ’ This drops the amplitude and power by 3dB. Thus what is then the difference between the -3dB law and Constant Power Law in Cubase?