Ambisonics was invented by Michael A. Gerzon, Peter Fellgett and John Hayes (and building on the work of other researchers) in the early 1970s. It is a series of recording and replay techniques using multichannel mixing technology that can be used live or in the studio. By encoding and decoding sound information on a number of channels, a 2-dimensional ("planar", or horizontal-only) or 3-dimensional ("periphonic", or full-sphere) sound field can be presented.

Benefits

Proponents of this technique assert that Ambisonics represents a superior approach to surround sound for a number of reasons: It has been suggested that Multichannel audio be merged into this article or section. ...

• It is isotropic in that sounds arriving from all directions are treated equally (as opposed to surround systems that assume that the main sources of sound are frontal and that rear channels are for ambience or special effects).
• because it does not rely solely on relative channel 'levels' to position a sound, but also includes other criteria (including psychoacoustic ones), localisation is superior to conventional panpotted techniques, which tend not to localise well between the speakers (sources are pulled into the speakers). Side and rear localisation in particular are enhanced.
• it allows better localization of sound outside of the sweet spot
• only 4 channels of information are required for transmission and storage of a full-sphere soundfield. Full-sphere replay is by more than 4 loudspeakers, the signal for each speaker position being derived using appropriate circuitry or software.
• the loudspeakers do not have to be positioned in a rigid setting - most regular polygons and (with somewhat more complex technology) a number of irregular figures are acceptable - allowing the speaker configuration to be matched more effectively to real listening environments (such as home living rooms, etc). The decoding circuitry or software is tailored to the speakers in use.
• the signal format used for Ambisonic distribution contains all the information required to reproduce the soundfield via a suitable decoder at the replay end. The decoder is configured to match the target array of replay loudspeakers, but the incoming Ambisonic signal can remain the same whatever the target array. Thus, unlike conventional surround media, the signal channels are not mapped one-to-one with the loudspeakers in the target array. In fact, the signal content is independent of the resolution of the target array. In a conventional surround system, which has 1:1 channel to speaker mapping, this is not the case, and the addition of more speakers, for higher replay resolution, requires additional channels, one speaker per channel. This allows flexibility for a composer or recording engineer to produce a "perfect" mix without worrying about how the mix will later be decoded.

Isotropic means independent of direction. Isotropic radiation has the same intensity regardless of the direction of measurement, and an isotropic field exerts the same action regardless of how the test particle is oriented. ... A sweet spot is a place, often numerical as opposed to physical, where a combination of factors suggest a particularly suitable solution. ...

First-order Ambisonics & B-Format

In the basic version, known as first-order Ambisonics, sound information is encoded into four channels: W, X, Y and Z. This is called Ambisonic B-format. The W channel is the general mono level of the signal, corresponding to the output of an omnidirectional microphone. The X, Y and Z channels are the directional components in three dimensions. They correspond to the outputs of three figure-8 microphones, facing forward, to the left, and up respectively. (Note that the fact that B-format channels are analogous to microphone configurations does not mean that Ambisonic recordings can only be made with coincident microphone arrays.) :For other senses of this word, see dimension (disambiguation). ...

The B-format signals are based on a spherical harmonic decomposition of the soundfield and correspond to the instantaneous sound pressure (W) and the three components of its gradient (X, Y, and Z, which are related to the particle velocity) at a point in space. The loudspeaker signals are derived by using a linear combination of these four channels, where each signal is dependent on the actual position of the speaker in relation to the center of an imaginary sphere going through all available speakers. In more advanced decoding schemes, spatial equalization is applied to the signals to account for the differences in the high- and low-frequency sound localization mechanisms in human hearing. A further refinement accounts for the distance from the listener to the loudspeakers. Spherical Harmonic is a fantasy novel by Catherine Asaro which tells the story of Pharaoh Dyhianna (Dehya) Selei, ruler of the Skolian Imperialate, after the Radiance War fought by the Imperialate and their enemy Eubian Concord. ... This article is about compression waves. ... Particle velocity is the velocity v of a particle (real or imagined) in a medium as it transmits a wave. ... Closeup of a loudspeaker driver. ... In mathematics, linear combinations are a concept central to linear algebra and related fields of mathematics. ... Sound localization is a listeners ability to identify the location of origin of a detected sound or the methods in acoustical engineering to simulate the placement of an auditory cue in a virtual 3D space (see binaural recording). ...

For the simplest (two dimensional) case (no height information), and spacing the loudspeakers equally around a circle, we can derive the loudspeaker signals from the using W, X and Y channels:

where N is the number of speakers, n is the speaker under consideration, k is a constant which can be between 1 and 2 and θ_n is the angle at which the speaker is located on the circle. The larger values of k make for better performance at the centre of the array, at the expense of the size of the sweet spot. By choosing k carefully, we can arrange for the signal presented to "match" the encoded B-Format signal.

More generally, the full three dimensional case (with loudspeakers equally spaced around a sphere) is given by:

where φ_n is the angle above or below the horizontal plane. Readers should note that the coordinate system used in Ambisonics follows the right hand rule convention with positive X pointing forwards, positive Y pointing to the left and positive Z pointing upwards. Horizontal angles run anticlockwise from due front and vertical angles are positive above the horizontal, negative below. Horizontal plane is used in radio to plot a antennas relative field strength (which directly affects a stations coverage area) on a polar graph. ... In mathematics as applied to geometry, physics or engineering, a coordinate system is a system for assigning a tuple of numbers to each point in an n-dimensional space. ... The right hand rule is also an algorithm used to solve Mazes In mathematics and physics, the right-hand rule is a convention for determining relative directions of certain vectors. ... A clockwise motion is one that proceeds like the clocks hands: from the top to the right, then down and then to the left, and back to the top. ...

Relationship to coincident stereo techniques

Different linear combinations of W, X, Y and Z can create signals equivalent to those picked up by any conventional microphone (omnidirectional, cardioid, hypercardioid, etc) pointing in any direction. Thus the signals used in any co-incident stereo microphone technique can be generated directly from the B-format signals (for example: MS with forward-facing cardioid using M = (W + X) / 2 and S = Y, or "Blumlein" crossed figure-8s using L = (X + Y) / 2 and R = (X - Y) / 2). A microphone, sometimes referred to as a mike or mic (both IPA pronunciation: ), is an acoustic to electric transducer that converts sound into an electrical signal. ...

Thus we can consider first-order B Format as a series of sum and difference channels: W = front + back + left + right + up + down (mono: omni mic); X = front - back (figure-eight mic facing forward); Y = left - right (figure-eight facing left); and Z = up - down (figure-eight facing up).

Recording techniques

The soundfield microphone

Many Ambisonic recordings have been made using a special microphone - the soundfield microphone (SFM). This microphone has also become popular with recording engineers, since it can be reconfigured electronically or via software to provide different stereo and 3-D polar responses either during or after recording. The Soundfield microphone is a recording studio microphone comprised of four closely spaced figure-of-eight microphone capsules positioned in a tetrahedron- the capsules MUST have some spacing, for providing phase difference between the capsules, which are four subcardioid capsules (not 4 figure-of-eight). ...

'Native' microphones

The SFM uses a tetrahedral array of capsules, the outputs of which are matrixed together to generate the component B-Format signals. However it is entirely practical to generate B-Format from a collection of coincident microphones (or mic capsules), each with the characteristics of one of the B-Format channels listed earlier. This is referred to as a "Native" Ambisonic microphone or microphone array. The primary difficulty inherent in this approach is that HF localisation relies on the diaphragms approaching true coincidence, which is difficult to achieve with complete microphones. However electronic coincidence compensation can be used, and this can be effective especially where capsules and not whole microphones are employed.

Thus if you wish to generate planar B-Format (WXY), you could use an omnidirectional mic coincident with a forward-facing and a left-facing figure-8. Exactly this technique was used by Dr Jonathan Halliday at Nimbus Records to record their extensive and continuing series of Ambisonic releases.

Ambisonic mixing

A popular and unfortunate misconception is that Ambisonic recordings can only be made with the SFM, and as a result there is a widespread, and erroneous, belief that Ambisonics can only be used to capture a live acoustic event (something that accounts for a tiny proportion of modern commercial recordings, the vast majority of which are built up in the studio and mixed from multitrack). This is not the case. In fact, Michael Gerzon's designs for Ambisonic panpots pre-date much of his work on soundfield microphone technology. Ambisonic panpots - which allow mono (for example) signals to be localised in B-Format space - were developed as early as the 1970s, and were incorporated into a special mixing console built by Alice Stancoil Ltd for the IBA surround-sound test broadcasts.

Ambisonic panpots, with differing degrees of sophistication, provide the fundamental additional studio tool required to create an Ambisonic mix, by making it possible to localise individual, conventionally-recorded multitrack or multimic sources around a 360-degree stage analogous to the way conventional stereo panpots localise sounds across a front stage. However, unlike stereo panpots, which traditionally vary only the level between two channels and thus rely on only one technique that humans use to localise sounds, Ambisonic panning provides additional cues which eliminate conventional localisation accuracy problems, especially in surround, where our ability to localise level-only panned sources outside the front quadrant is severely limited (especially to the sides).

Other tools included 'spreaders' which were designed to "de-localise" a signal (typically by varying the virtual source angle with frequency within a determined range) - for example, in the case of reverb returns - however these were not developed further.

Legacy hardware

By the early 1980s, studio hardware existed for the creation of multitrack-sourced, Ambisonically-mixed content, including the ability to incorporate SFM-derived sources (for example for room ambience) into a multichannel mix. This was thanks primarily to the efforts of Dr Geoffrey Barton (now of Trifield Productions) and the pro audio manufacturers Audio & Design Recording, based near Reading, UK. Barton designed a suite of outboard rack-mounted studio units that became known as the Ambisonic Mastering System. These units were patched into a conventional mixing console and allowed conventional multitrack recordings to be mixed Ambisonically. The system consisted of four units:

• Pan-Rotate Unit - This enabled eight mono signals to be panned in B-format, including 360-degree 'angle' control and a 'radius vector' control allowing the source to be brought in towards the centre, plus a control to rotate an external or internal B-format signal.
• B-Format Converter - This connected to four groups and an aux send and allowed existing console panpots to pan across a B-Format quadrant.
• UHJ Transcoder - This both encoded B-Format into 2-channel UHJ (see UHJ format) and in addition allowed a stereo front stage and a stereo rear stage (both with adjustable widths) to be transcoded direct to 2-channel UHJ
• Ambisonic Decoder - this accepted both horizontal (WXY) B-format and 2-channel UHJ and decoded it to four speaker feeds with configurable array geometry.

A significant number of releases were made with this equipment, all in 2-channel UHJ, including several albums on the KPM production music library label, and commercial releases such as Steve Hackett's Till We Have Faces, The Alan Parsons Project's Stereotomy, Paul McCartney's Liverpool Oratorio, Frank Perry's Zodiac, a series of albums on the Collins Classics label, and others, most of which are available on CD. See The Ambisonic Discography in the External links for more information. Engineer John Timperley employed a transcoder on virtually all his mixes over the course of over a dozen years until his sad death in 2006. Unfortunately the albums, film soundtracks and other projects he created in UHJ over this period are largely undocumented at present, and thus remain unlisted in the Discography.

The lack of availability of 4-track mastering equipment led to a tendency (now regretted by some of the people involved) to mix directly to 2-channel UHJ rather than recording B-format and then converting it to UHJ for release. The fact that you could mix direct to 2-channel UHJ with nothing more than the transcoder made this even more tempting. As a result there is a lack of legacy Ambisonically-mixed B-format recordings that could be released today in more advanced formats (such as G-Format). However, the remastering - and in some cases release - of original 2-channel UHJ recordings in G-Format has proved to be surprisingly effective, yielding results at least as good as the original studio playbacks, thanks primarily to the significantly higher quality of current decoding systems (such as file-based software decoders) compared to those available when the recordings were made.

Current mixing tools

The advent of digital audio workstations has led to the development of both encoding and decoding tools for Ambisonic production. Many of these have been developed under the auspices of the University of York (see External links). The vast majority to date have been created using the VST plugin standard developed by Steinberg and used widely in a number of commercial and other software-based audio production systems, notably Steinberg's Nuendo. With the lack of necessity to interface to a conventional console, the encoding tools have primarily taken the form of B-Format panpots and associated controls. Decoder plugins are available for monitoring.

There are presently some issues with implementing B-format groups and other channel structures in current DAW software which is often either stereo-based or based inflexibly on conventional surround configurations. The ability must exist to use plugins with one input and multiple outputs, for example, and it must be possible to create B-format buses of some sort and hook up decoder plugins to them, record their contents, and perform other operations. Documentation is being assembled to assist engineers wishing to work with these tools.

There are also stand-alone software tools for manipulating multichannel files and for offline decoding of B-Format and UHJ files to standard arrays, plus software players capable of playing and decoding standard B-Format files and other Ambisonic content.

The plugin field is a particular growth area for Ambisonic production tools at the present time.

UHJ format

Ambisonic B-Format is the standard format for use in the studio. While it is possible to distribute B-Format recordings for decoding and listening by end-users, this is only starting to be more widespread with the advent of software-based players. Traditionally, Ambisonic recordings have been distributed in the form of 2-channel discs, CDs etc using the 2-channel version of the UHJ encoding hierarchy. UHJ, which was developed by the Ambisonic team, incorporating work done by the BBC (on Matrix H) and Duane Cooper (on Nippon Columbia's UD-4) and others, takes the 4-channel B-Format signal and encodes it into a hierarchical set of up to four channels (L, R, T & Q). All four channels deliver full, with-height 3D periphony, with a level of accuracy identical to 4-channel B-Format. Removing Q, the fourth UHJ channel, removes the height information: the resulting 3-channel UHJ delivers the same accuracy as 3-channel (WXY) B-Format to provide horizontal (planar) replay. The third channel, T, may be bandwidth-limited without serious degradation of localisation accuracy. This "2 1/2 channel" format was tested by the Independent Broadcasting Authority (IBA) in the United Kingdom as a method of broadcasting surround recordings, with the 1/2-channel applied to the FM stereo signal via phase-quadrature modulation. Removing the third channel results in 2-channel UHJ, which offers planar surround with rather less localisation accuracy than planar B-format. This is the format commonly used in the past for distribution of Ambisonic recordings. It is mono/stereo compatible in that, without decoding, the listener perceives a stereo image, but one that is significantly wider than conventional stereo. The L & R channels can also be summed for a very high degree of mono-compatibility. Replayed via a UHJ decoder, the full surround capability is revealed.

G-Format

The lack of availability of Ambisonic decoders (only a handful of hardware decoder models are currently available, although software-based players are now emerging) led to the proposal that Ambisonics could be distributed by decoding the original signal (preferably B-Format but also legacy 2-channel UHJ recordings) in the studio instead of at the listening end. A professional software or hardware-based decoder is used to decode the Ambisonic signal to a conventional surround speaker array (eg 5.1) and the resulting speaker feeds are authored to a conventional multichannel disc medium such as DVD. This is known as "G-Format".

The obvious advantage of this approach is that any surround listener can be able to experience Ambisonics: no special decoder is required beyond that found in a common home theatre system. The main disadvantage is that the flexibility of rendering a single, standard Ambisonic signal to any target speaker array is lost: the signal is targeted towards a specific "standard" array and anyone listening with a different array may experience a degradation of localisation accuracy, depending on how much the actual array differs from the target.

In practice, Ambisonics in general has proved to be very robust, however. Examples of G-Format recently released by Nimbus Records used 2-channel UHJ decoded to a square array of four speakers (this is conventional for decoding planar Ambisonic recordings: a rectangle of sides with ratios of between 2:1 and 1:2 can be used, a square being midway between the two). The resulting 4-channel (LF, RF, LS, RS) signal was authored to DVD-Audio/Video discs and although many listeners will be listening on arrays other than a square, the results have proved very encouraging. Nimbus Records is a British record company specializing in classical music recordings. ...

Some releases of G-format sourced from B-Format have also occurred, for example the album Swing Live by Bucky Pizzarelli (available on Chesky Records, DVD-A or SACD), where a B-Format SFM recording was "manually decoded" to 4.0 speaker feeds in the mixdown process.

Recovering B-Format from G-Format

It is theoretically possible to recover B-Format from a G-Format signal, in which case Ambisonic listeners with their own decoders could recover the B-Format and decode it for their own array, thus achieving more accurate localisation. However for the greatest accuracy in smaller environments such as a living room, the decode process includes shelf filtering that may cause the decode to be irreversible.

It is possible that as a result of current development work (primarily by Dr Peter Craven) on hierarchical systems for audio rendering, these problems can be overcome (and G-Format superseded) by distributing a common signal that plays back as 5.1 on 5.1 systems (and so on) but can also be decoded Ambisonically if listeners have the right equipment.

Downloadable B-Format files

An official file format for downloadable B-Format files, called ".amb" format, has been defined. Several score such files are available for free download from the Ambisonic Bootlegs website listed in the External links. (And they are not bootlegs.) The website also gives details of ad hoc software players. 20 (twenty) is the natural number following 19 and preceding 21. ...

The ".amb" file format is defined for B-Format files upto third-order, full-sphere (16 channels), although most of the files currently available are first-order, full-sphere (4 channels).

Current developments

The Ogg Vorbis project has shown interest in implementing Ambisonics as a means for including surround sound in their project. In addition there is a growing series of freely-available developments such as VST plugins, enabling common DAW systems (such as Nuendo) to be used to encode and decode B-Format and generate decoded speaker feeds: see External links. Ambisonics is also in use in a number of commercial products, primarily in the professional audio industry, and some consumer electronics manufacturers include Ambisonic decode capability in their surround systems. This article is about the audio compression codec. ...

Higher-order Ambisonics

A particularly active area of current research is the development of "higher orders" of Ambisonics. These use more channels than the original first-order B-Format and offer benefits that include greater localisation accuracy and better performance in large-scale replay environments such as performance spaces.

FMH-Format (Furse-Malham Higher Order Format) is a set of coefficients for creating second-order B-format channels, using nine channels rather than the four used by first-order B-Format. Significantly more spatial information is captured in this format. At present, "real" recording techniques using this are in their infancy, however it is straightforward to compose synthetic recordings. The FMH set of coefficients applies weightings to the channels such that all the spherical harmonic coefficients have a maximum value of unity. Whilst this approach is not rigorously "correct" in mathematical terms, it has significant engineering advantages in that it restricts the maximum levels a panned mono source will generate in some of the higher order channels. Some experimental software extends this approach to higher orders as well.

The following links provide details of some of the work currently going on in higher-order Ambisonics:

• http://gyronymo.free.fr/audio3D/download_Thesis_PwPt.html
• http://www.muse.demon.co.uk/3daudio.html
• http://www.york.ac.uk/inst/mustech/3d_audio/higher_order_ambisonics.pdf
• http://users.skynet.be/solaris/linuxaudio/downloads/hoafilt.pdf

The following table lists the various higher-order combinations which are possible:

In the table, note that as you move from horizontal to full-sphere, or from lower to higher orders, backwards compatibility is always guaranteed because channels are only ever added. This means, for example, that a first-order, horizontal decoder can still decode a third-order, full-sphere soundfield by simply ignoring 13 of the 16 channels.

Intellectual property

Most of the patents covering Ambisonic developments have now expired (including those covering the Soundfield Microphone), and as a result the basic technology is available for anyone to implement. Exceptions to this include Dr Geoffrey Barton's Trifield technology, which is essentially a three-speaker stereo rendering system based on Ambisonics (US patent 5671287); and so-called "Vienna" decoders, based on Gerzon and Barton's Vienna 1992 AES paper, which are able to decode to irregular speaker arrays (US patent 5757927). 1992 (MCMXCII) was a leap year starting on Wednesday. ...

The 'pool' of patents comprising Ambisonics technology was originally assembled by the UK Government's National Research & Development Corporation (NRDC), which existed until the late 1970s to develop and promote British inventions and license them to commercial manufacturers - ideally to a single licensee. The system was ultimately licensed to Nimbus Records (now Wyastone Estates Ltd) who hold the rights to the 'interlocking circles' Ambisonic logo (UK trademark 1500177). Nimbus Records is a British record company specializing in classical music recordings. ...

References

• Michael A. Gerzon. ``Periphony: With-Height Sound Reproduction. Journal of the Audio Engineering Society, 21(1):2--10, 1973.
• Peter Felgett. ``Ambisonics. Part One: General System Description. Studio Sound, 1:20--22,40, August 1975.
• Michael A. Gerzon. ``Ambisonics. Part Two: Studio Techniques. Studio Sound, pages 24--30, October 1975. Correction in Oct. 1975 issue on page 60.

See also

• Soundfield microphone
• Surround sound

The Soundfield microphone is a recording studio microphone comprised of four closely spaced figure-of-eight microphone capsules positioned in a tetrahedron- the capsules MUST have some spacing, for providing phase difference between the capsules, which are four subcardioid capsules (not 4 figure-of-eight). ... It has been suggested that Multichannel audio be merged into this article or section. ...

External links

• Ambisonic.net website
• Ambisonic Surround Sound FAQ
• AmbisonicBootlegs, a repository of Ambisonic recordings and compositions
• Ambisonic Discography, a list of record releases, broadcasts and other Ambisonic content
• Ambisonics resources at the University of Parma
• 3D Audio Links and Information
• Ambisonic resources at the University of York
• Binaural Ambisonics rendering at the Institute of Electronic Music and Acoustics (IEM)