Panning, How does it work?

Started by KrazyKatz, February 13, 2007, 21:29:50

Previous topic - Next topic

KrazyKatz

At first I thought that panning effect is simply a higher level of volume on one channel than the other, but I understand it also has something to do with timing...

Eg: We have a left and right speaker. If a sound is set in the middle, how is that we hear it in the middle since the speakers are left and right to us?

Perplexing...
Sonic Brilliance Studios
http://www.sonicbrilliance.com

LPChip

Actually, if you put the same volume on both channels, your brain will combine it and makes it centered.

If you'd put a small change in one of the channels, your brain will still see it as center but it will sound differently. If you would invert the channel then your brain would be tricked giving a nice widened stereo image.
"Heh, maybe I should've joined the compo only because it would've meant I wouldn't have had to worry about a damn EQ or compressor for a change. " - Atlantis
"yes.. I think in this case it was wishful thinking: MPT is makng my life hard so it must be wrong" - Rewbs

Sam_Zen

The brain it is..
(That's why I still have my doubts about the necessity of that front center speaker in 5.1)

The brain does the calculations, and it can do so, because we also have a left and right ear. So if a sound comes from the right, the right ear will hear this sound a fraction of time earlier before it reaches the left ear. Sound travels fast, so the difference of recorded signal between left and right is on micro phase-level. But the ear is fast enough to pass this information.
So, if both signals do have exactly the same content on volume- and on phase-level, the brain locates the source as coming from the centre of the panorama.
That's why we have 2 ears. To define where the sound is coming from. If we only would have one ear, we would need to turn our head around to focus the direction with the biggest amplitude, to locate the source.
If an ear would be just a hole in our head, we could only hear in 1 dimension : a line between left and right. But, because of the special shape of our ears, we can also distinguish if a sound coming from the sky or from the ground, and also if it's from the front or from behind.

This is why I wouldn't call Panorama an 'effect'. Is a volume-slider an effect ?

Quote from: "LPChip"If you would invert the channel then your brain would be tricked giving a nice widened stereo image
This is done in the player option of OMPT of 'Pro-logic Surround' basically, together with some small delay between the channels.
0.618033988

LPChip

Actually, the reason why you have a center speaker in a 5.1 surround situation is because most 5.1 surround setups use the following setup:

front left: speaker designed to output bass and treble
front right: speaker designed to output bass and treble
center: speaker designed to output treble only
rear left: speaker designed to output treble only
rear right: speaker designed to output treble only
sub: speaker designed to output bass only.

As you can see, the center in these cases will give more clarity.

But you also have the situation where you have 5 treble only and a bass speaker. In these cases the treble speakers usually don't output enough watt so the center actually adds value in strength. Remember: the rear speakers usually are much closer to you, and the front+center usually much further away.

And perhaps, the 5.1 decoder will put left to 25% center on the left, 25% left to 25% right to the center speaker, and 25% right to right on the right speaker. I haven't really looked into that, but it would be obvious. :)
"Heh, maybe I should've joined the compo only because it would've meant I wouldn't have had to worry about a damn EQ or compressor for a change. " - Atlantis
"yes.. I think in this case it was wishful thinking: MPT is makng my life hard so it must be wrong" - Rewbs

KrazyKatz

Nice answer Mr. Zen. You mentioned " phase level". What exactly is phase level?
Sonic Brilliance Studios
http://www.sonicbrilliance.com

Sam_Zen

2 LPChip
Yeah, I forgot the division into frequency ranges in this system. And of course this is meant for movies, where the talking head is in the centre of the screen most of the time. Or the singer of a band in case of music.

2 KrazyKatz
You're right asking about 'phase level', because actually it is a wrong expression. It should be : 'phase position'.
It's a time thing, not so much a volume thing. And because the time difference is shorter than one full wavecycle :                  
Here you see the same soundform in both channels, but the R channel (bottom) is a bit earlier.
In this case the difference is 0.38 msec. This means a physical distance of about 13 cm. In air.
0.618033988

LPChip

Quote from: "Sam_Zen"Here you see the same soundform in both channels, but the R channel (bottom) is a bit earlier.
In this case the difference is 0.38 msec. This means a physical distance of about 13 cm. In air.

Really!!!!  :shock:  :o

I didn't know! :) Thanks, this explains alot, and is very usefull :)

Edit: this seems like a very good topic to move to Technical Documents. :)
"Heh, maybe I should've joined the compo only because it would've meant I wouldn't have had to worry about a damn EQ or compressor for a change. " - Atlantis
"yes.. I think in this case it was wishful thinking: MPT is makng my life hard so it must be wrong" - Rewbs

Sam_Zen

One can experiment with this within OMPT :
~ Import a (mono) sample with quite a long duration, or, better, with some looped drum-sequence.
~ Place it on the same row with the same note in two different channels.
~ Set the panning of one channel to the left, the other one to the right.
~ Set the volume of each channel and sample on an equal value.
~ This would result in hearing the sample coming from the center in the stereo.
Of course, if there is a difference in volume-setting between the duplicates, this would affect the position in the panorama, but it still maintains a 'mono' point, only not in the center.
~ To use the space of the panorama one could also do it with the time factor. ModPlug has some codes for this, to use 'offsets'.
XM : 9xx and IT : Oxx or S6x. Apply this to only one of the two.
Or, if the tempo is high enough, one could shift one of the two duplicates one row forward.
0.618033988