Surround Sound Recording Techniques - March 4, 2010
Recording Surround Sound on a budget?
...No?
Well, I am, as part of the Open Source Project. All new Terrene music will be recorded for the 24-bit/96khz/Surround Sound format.
And this is how:
The Mulhausen Rule of Recording Surround Sound Audio Via Triangulation:
Triangulation is important because a single mono signal does not create a convincing enough sound image when panned about inside a 360-degree surround sound field. Instead, the sound is very isolated, lonely, and lacking in impact, as it is being rendered in a very precise place in the listening field by very small speakers.
This rule of triangulation uses natural acoustic techniques to create convincing surround placement in the mix, by making the surround sound field a virtual room in which sounds reverberate realistically, stretching across the width and depth available in surround sound.
The Mulhausen Rule of Mixing Triangulated Audio Images By Quadrant
IN = instrument being recorded.
M# = Microphone, where "#" is the microphone number. Assume the
mic is pointing straight "North."
IN
M1
M2 M3
When placing a three-channel signal that was recorded using the rule of triangulation into a surround sound mix, the surround sound field must be mentally divided into quadrants that govern the seperation of the room mic signals and the close microphone signal across the 2-dimensional aural plane of modern surround sound.
Imagine that a triangle was painted onto the surface of a coin, with its joints labelled M1 at the top, and then M2 and M3 at the bottom. As the close instrument mic (M1) is placed in the mix, imagine rotating that coin so that M2 and M3 both fell inside the "invisible box" of the surround sound field.
Here's an example of what that would look like if a triangulated image were placed in a surround sound mix's extreme Front-Left. The "coin" mentioned earlier has clearly been rolled counter-clockwise:
To define what's really happening when you visualize rolling a coin with a triangle on it, the concept of quadrants comes in. Imagine dividing the surround sound field into quadrants (XX, XY, YX, YY). Keeping that in mind, the rule is as simple as this: Whatever quadrant the close instrument mic is occupying, the two adjacent quadrants receive the stereo room mics. Here's where we must leave the analogy of the coin behind: the room mics must then be panned to the extremity of their respective quadrant, in order for them to use completely seperate speakers and avoid phase cancellation.
M# = Microphone signal from above example, numbered accordingly by #.
| and - demark the invisible box that defines the surround sound field.
--------------
|M1 M2|
| |
| |
| |
|M3 |
--------------
This must remain the case even when the close instrument mic, M1, responsible for providing 90% of the perceived audio, is moved to be more towards the center, as illustrated below. Remember: M2 and M3 would phase if routed to the same audio channel, and M1's travel to the center of the field means that the rear-right speaker is now, ever-so faintly, contributing to the image for M1. However, it is not, for M2. Thus, the perceived effect is that the sound is close by the user, reverberating slightly off the left and forward walls.
--------------
| M2|
| M1 |
| |
| |
|M3 |
--------------
And if the engineer, busily mixing, continued to move M1 in that direction so that it completely crossed from the front-left to the rear-right, that would, in the name of having the most accurate, clean, and deep-sounding surround sound image, produce:
--------------
| M3|
| |
| |
| M1 |
|M2 |
--------------
Always remember that M2 and M3 must be panned as though they were bound by invisible glue to the extreme corners of the surround sound field, and harbor a natural aversion to their being in the same quadrant as M1. As a result, you'll be rewarded with a natural-sounding, phase-free surround image that only took 3 channels to produce.
CONSIDERATIONS FOR CENTERED AUDIO SIGNALS:
The two big corner cases are those times where M1 is perfectly centered *in some way.* The two ways that an audio image can be centered, and the mixing placements that compensate, are:
1) Centered on the X (Left/Right) or Y (Front/Back) axis, but not both: The room mics will be in the two quadrants not occupied by the two which contain the close instrument mic signal. In the below example, M1 is being shared by the two lower quadrants, resulting in the opposing quadrants receiving the room mic signals:
--------------
|M2 M3|
| |
| |
| |
| M1 |
--------------
2) Centered on both the X and Y axis (dead center in the surround sound field): Next time use five mics and make a domino pattern with M1 in the center. But this time: Align M3 and M4 at opposite ends of the Y axis, along the X axis. Then place M1 at the center of the X axis. Like so:
--------------
| M3 |
| |
| M1 |
| |
| M2 |
--------------
The image here will be confusing for the listener, as it will have depth, but no width, however it will at least be sonically balanced and in the center of the mix.
CONSIDERATIONS FOR FRONT-HEAVY MIXERS:
Naturally, this is all written from the point of view of a technical study in imaging audio correctly in a surround sound field using as few microphones as possible, where all speakers recreating the surround sound field are assumed to be equal in ability and volume. In the real world, surround sound speaker systems of the sort in home theaters today are extremely "front-heavy," and the likely mixing strategy for many will be to use the front channel's additional center speaker for the isolation and greater clarity of key song elements like vocals and snare, and to merely use the back speakers for "depth information" like room mic signals.
If the engineer is producing front-heavy mixes that use rear speakers "for depth" (i.e. room sounds, reverb, audience noise, but no instrument detailing), the likely scenario is the quadrants XX and XY will contain all the close instrument mics (the M1's of their respective instruments) and quadrants YX and YY will contain the M2s and M3s, respectively (and exclusively), using strategic balancing of the M2 and M3 signals to acheive the desired placement in the surround sound field.