Page 1 of 2

The Mysterious Panning Law

Posted: Thu Sep 07, 2006 5:00 am
by Robert Henke
Obviously not everyone undertstands what different panning laws do and how this affects sound. In the light of the ongoing sound quality discussions here comes some explanation:

Panning is achived by changing the level of one stereo channel in relation to the other channel. [ Which is quite boring and has nothing to do with real placement of an accoustic source but is the only way to do it without introducing potentially problematic phase or frequency alterations.... ]

How does it work:
If the pan is centered you want the input signal to pass unchanged. Each channel is amplified by 0dB. If you pan completly to one side you want the other channel to be off. But what do you want to happen to the channel you panned to?

You might want it a bit louder because you want to compensate for the energy loss due to the missing other channel.

Different softwares use different rules for the exact curve and the volume boost on the extremes.

What has to be considered:

1. if you feed in a normalised signal one side will clip. If the panning law creates a 6dB boost you will need this much headroom if you want to avoid clipping

2. regardless of the panning law, each static panning can be created with any software by the right settings of the pan and the volume control.

3. the only difference is how it sounds / feels if you simply turn the pan knob from full left to full right.

4. panning only changes volumes, never the sound itself.

5. in Live, the boost at the extremes is + 3dB.


hope this helps

Robert

Posted: Thu Sep 07, 2006 6:15 am
by Johnisfaster
so wait... panning to the left would drop the right and BOOST the left? I was never aware of a boost. I always thought it was simply the opposite side.

do I understand you correctly on that?

Posted: Thu Sep 07, 2006 1:37 pm
by forge
Johnisfaster wrote:so wait... panning to the left would drop the right and BOOST the left? I was never aware of a boost. I always thought it was simply the opposite side.

do I understand you correctly on that?
that's just down to wording - you could describe it as a 3dB dip in the center, or a 3dB boost on the sides - same thing either way really

Posted: Thu Sep 07, 2006 2:37 pm
by anti-banausic
I am one of those who responded in the present thread.

This explanation makes a lot of sense, and I thank you for it.

I, for one, haven't doubted LIVEs engine. Though I have precious little to compare it against.

But thanks for explaining it in such a good way.

AB

Posted: Thu Sep 07, 2006 2:55 pm
by Angstrom
forge wrote:
Johnisfaster wrote:so wait... panning to the left would drop the right and BOOST the left? I was never aware of a boost. I always thought it was simply the opposite side.

do I understand you correctly on that?
that's just down to wording - you could describe it as a 3dB dip in the center, or a 3dB boost on the sides - same thing either way really

hmm, not so sure about that.

0db gain in the center is not 'the same' as -3db gain in the center for the purposes of those 'engine checker' types. Personally I prefer the way Live appears to do it.

What Robert describes would produce a central signal 3db hotter than the 'center dip' method as far as I understand it. So, when Live's output is compared (like people seem to ) against the output of another app which did 'dip in the center' then Live would be louder. (simply because it hasn't attenuated)

0db center signal -3db != 0db center signal -0db ;)

Not that this is a problem at all - just pointing it out.

Posted: Thu Sep 07, 2006 4:03 pm
by dj superflat
questions:

is this the same as HW? that is, what does your standard deck do? (maybe there is no standard, but i never noticed different panning on consoles, whether SSL, neve, or even low rent mackie.) what's the boost at the extremes on HW? it is linear increase from center to extreme or something else in live or in HW?

Posted: Thu Sep 07, 2006 7:42 pm
by tomperson
So, if i understand correctly, when I hard pan to one side, the other one is completely lost?

I mean, I have a drums loop where the hat is mostly on the right channel, and I hard pan the stereo out to the left, then Is it correct saying that my hat is now gone instead of "moved" in the panorama to the right?

Posted: Thu Sep 07, 2006 10:10 pm
by nolus
tomperson wrote:So, if i understand correctly, when I hard pan to one side, the other one is completely lost?

I mean, I have a drums loop where the hat is mostly on the right channel, and I hard pan the stereo out to the left, then Is it correct saying that my hat is now gone instead of "moved" in the panorama to the right?
Correct!

imho the pan control in live is actually a stereo balance control.

If you want to do proper stereo panning then you need to use a utility to narrow the stereo field before panning it.

Posted: Thu Oct 12, 2006 1:39 pm
by mylkoa
Hello All,

I'm not sure if this has to do with the panning law, but I noticed that when I place two tracks of the same stereo clap sound playing and I hard panned them Left and Right respectively, that it sounds different from the stereo sound coming through a single track with no pan.

Can someone offer some light on this? Shouldn't it sound the same? Could it just be a phasing issue/accident, or is this a commonly used mixing technique?

Cheers,
Andrew

P.S. I got the inspiration to try this experiment today from listening to massive amounts of disco in the past few weeks... in particular a song by Franciose Kevorkian got me looking at panning :)

Posted: Thu Oct 12, 2006 2:27 pm
by serotoninsteve
It´s a sort of phasing, this happens when you play exactly the same sounds in different tracks together in Live.
When they are hard panned I assume that this gives then some sort of stereo effect.

Greetings

Posted: Mon Jul 16, 2007 3:52 pm
by [Galo]
Why would you boost the channel and not keepoing it at 0db but just have the other side go down -3db, cause now i know this i know why i need to set my faders to -3db in live (what happend to be the case every mix...)

keep up the good work abletons

-Galo

Posted: Mon Jul 16, 2007 3:58 pm
by mike holiday
does the pan in, say, impulse have this same boost.. or is this only for tracks?

Posted: Mon Jul 16, 2007 4:04 pm
by Tarekith
LOL, so much for clearing things up Robert!

Posted: Mon Jul 16, 2007 5:58 pm
by mikemc
the idea is to keep the relative loudness vis a vis your mix the same, while establishing the 'positioning' of the item. the long and short is that you *might* have to adjust your fader overall slightly if you pan something.

Undoubtedly, real men (like Robert Henke) are aware of this reality and are called upon by their audio-aesthetic sensibilities to do this every day, wielding mighty knobs, vast faders and even huge mice, moving them in accord with this sacred duty, despite the burdensome toil.

:) :roll: :)

Posted: Tue Jul 17, 2007 4:57 am
by Jeroen
Just for my understanding as well. Suppose I have a center signal that is just below clipping level and I pan that to one of the extremes the signal is clipped?

J