Recording - April 2012 - 40

DAW, or what-have-you. But, once you have this information in this format, what can you do with it? How to decode ambisonics You can’t play the B-format recording directly, but you can decode it using fairly simple mathematical transforms to drive a multi-speaker array. Using eight or more speakers you can get a remarkably accurate stereo image, complete with depth cues, height cues... you can tell how high an object is in the soundfield. The soundfield accuracy is as good in the front as it is in the back; you can turn your head around 360 degrees and everything in the soundfield retains its position accurately. The problem is, nobody wants a huge array of eight or more speakers in their house. People want to listen on stereo or 5.1 surround systems, and it is possible to translate the ambisonic formats directly down to stereo or 5.1 surround. The problem is that doing this you lose some information, including height cues. The advantage is that you get to choose the information that you lose when you operate the decoder. This means you can start with a 4-channel B-format recording, load it into your workstation, and— using the right software—listen to it in stereo while adjusting the width of the stereo image, the height of the stereo image, and the direction. You can’t adjust the balance between ambient and direct sound, but you can make a remarkable number of changes, steering the soundstage in the decoding process. This makes it a very powerful tool for live recording, because in the field when you do not have proper monitoring environments or enough time for proper setup, you can delay some of the decisions that you would normally make with an ordinary stereo recording and deal with them at mixdown time. 5.1 and UHJ It also means you can get a free 5.1 surround mix in the bargain, and the system translates very well to 5.1. Because you mathematically have all the information about the wavefront at one point in space, you can translate that into any playback configuration possible. One note: One of the options that the ambisonic systems give you is UHJ-format. UHJ is a “matrix surround” format, rather like Dolby Stereo or the old SQ or QS 1 quadrophonic record formats. It allows you to record limited surround information onto a 2channel distribution medium. Back in the 1970s when 2-channel distribution was the only practical way to get audio to the consumer, this was a big deal, and if you read any older Ambisonic literature, you will hear a lot about UHJ-format. This isn’t the 1970s any more. We have digital distribution formats. Some people like to use UHJformat as a kind of extra-wide stereo, just listening on two channels without decoding, and that’s still a valid use of it today. What’s on the market For many years the only ambisonic hardware available was the SoundField system, originally made by Calrec and now made by the SoundField company in

S By

o tD

y rse

I think people should be interested in ambisonics, which is a very interesting way of recording accurate stereo and surround from a single point in space. It’s been around for 30 years, but it’s of special interest now that there is relatively inexpensive ambisonic recording gear out there and that recording the multiple tracks required has become quite cheap. Michael Schulze has mentioned it a couple of times in his recent articles about Mid/Side stereo recording, saying, “If you think this stuff is cool, check out ambisonic.net for something even cooler.” I agree. Let me first point out that this is a British invention, and the British are famous for doing things in ingenious but totally different ways. These are the people who drive on the other side of the road, in cars using nuts and bolts made with Whitworth thread that isn’t metric or SAE standard... it’s just British. One look inside a Quad loudspeaker will show you what I mean by “ingenious but different”! B-Format and A-Format There are a lot of stereo recording methods that can reproduce an accurate stereo image from one location; these methods are very popular for classical and other acoustic music recording. You may have heard of standard stereo methods like ORTF or coincident cardioids or spaced omnis. In the 1970s, British scientist Michael Gerzon reasoned that you can effectively extrapolate the shape of a complete soundfield by looking at the air pressure and velocities at a single point in space. By using one omnidirectional microphone (for pressure), and three figure-8 microphones (for velocities), he constructed an array that produced a four-channel output that contained all the information about the sound passing through the microphone array. The four channels give you absolute pressure, then up-down, right-left, and front-back localization information. This is called a “B-Format” signal, and contains all you need to know about precisely what direction in 3-D space a sound is coming from. The problem is that figure-8 capsules are difficult to build, but it turns out that a much easier way to get the same information is with an array of four cardioid capsules mounted in a tetrahedral array. This produces what he called “A-Format”, which could be converted into B-format with a fairly simple mathematical transformation. There’s a geometric proof that the two formats give equivalent information, but the math is well beyond the scope of this article, so you’ll have to take my word for it. (Or you can read Gerzon, Michael: “The Design of Precisely Coincident Microphone Arrays for Stereo and Surround Sound”, AES preprint L-20, presented at the 50th AES convention, March 1975.) So... a typical ambisonic recording system has a single microphone with four capsules and a black box that converts the output into B-format. The box has four outputs for W (pressure), X (front/back), Y (left/right), and Z (up/down) signals, which you can record on your 4-track tape machine, digital recorder,
40
RECORDING APRIL 2012

http://www.ambisonic.net

Recording - April 2012

Table of Contents for the Digital Edition of Recording - April 2012