Once upon a time, in the incorrectly-named ‘Good Old Days’, audio systems were relatively simple things. There was a single channel of audio picked up by a gramophone needle or transmitted to a radio, and that single channel was reproduced by a single loudspeaker. Then one day, in the 1930s, a man named Alan Blumlein was annoyed by the fact that, when he was watching a film at the cinema and a character moved to one side of the screen, the voice still sounded like it was coming from the centre (because that’s where the loudspeaker was placed). So, he invented a method of simultaneously reproducing more than one channel of audio to give the illusion of spatial changes. (See Patent GB394325A:Improvements in and relating to sound-transmission, sound-recording and sound-reproducing systems’) Eventually, that system was called stereophonic audio.
The word stereo’ first appeared in English usage in the late 1700s, borrowed directly from the French word ‘stéréotype’: a combination of the Greek word στερεó (roughly pronounced ‘stereo’) meaning ‘solid’ and the Latin ‘typus’ meaning ‘form’. ‘Stereotype’ originally meant a letter (the ‘type’) printed (e.g. on paper) using a solid plate (the ‘stereo’). In the mid-1800s, the word was used in ‘stereoscope’ for devices that gave a viewer a three-dimensional visual representation using two photographs. So, by the time Blumlein patented spatial improvements in audio recording and reproduction in the 1930s, the word had already been in used to mean something akin to ‘a three-dimensional representation of something’ for over 80 years.
Over the past 90 years, people have come to mis-understand that ‘stereo’ audio implies only two channels, but this is incorrect, since Blumlein’s patent was also for three loudspeakers, placed on the left, centre, and right of the cinema screen. In fact, a ‘stereo’ system simply means that it uses more than one audio channel to give the impression of sound sources with different locations in space. So it can easily be said that all of the other names that have been used for multichannel audio formats (like ‘quadraphonic’, ‘surround sound’, ‘multichannel audio’, and ‘spatial audio’, just to name a few obvious examples) since 1933 are merely new names for ‘stereo’ (a technique that we call ‘rebranding’ today).
At any rate, most people were introduced to `stereophonic’ audio either through a two-channel LP, cassette tape, CD, or a stereo FM radio receiver. Over the years, systems with more and more audio channels have been developed; some with more commercial success than others. The table below contains a short list of examples.
Some of the information in that list may be surprising, such as the existence of a 7-channel audio system in the 1950s, for example. Another interesting thing to note is the number of multichannel formats that were not `merely’ an accompaniment to a film or video format.
One problem that is highlighted in that list is the confusion that arises with the names of the formats. One good example is ‘Dolby Digital’, which was introduced as a name not only for a surround sound format with 5.1 audio channels, but also the audio encoding method that was required to deliver those channels on optical film. So, by saying ‘Dolby Digital’ in the mid-1990s, it was possible that you meant one (or both) of two different things. Similarly, although SACD and DVD-Audio were formats that were capable of supporting up to 6 channels of audio, there was no requirement and therefore no guarantee that the content be multichannel or that the LFE channel actually contain low-frequency content. This grouping of features under one name still causes confusion when discussing the specifics of a given system, as we’ll discuss below in the section on Dolby Atmos.
|Edison phonograph cylinders||1896||1|
|Berliner gramophone record||1897||1|
|Optical (on film)||ca. 1920||1|
|Fantasound||1940||3 / 54||3 channels through 54 loudspeakers|
|Q-8 magnetic tape cartridge||1970||4 .|
|CD-4 Quad LP||1971||4|
|SQ (Stereo Quadraphonic) LP||1971||4|
|Dolby Stereo||1975||4||also known as 'Dolby Surround'|
|Digital Compact Cassette||1992||2|
|DTS Coherent Acoustics||1993||5.1|
|SACD||1999||5.1||Actually 6 full-band channels|
|Tom Holman / TMH Labs||1999||10.2|
|DVD-Audio||2000||5.1||Actually 6 full-band channels|
|Dolby Surround 7.1||2010||7.1|
|NHK: Ultra-high definition TV||2005||22.2|
|Auro 3D||2005||9.1 to 26.1|
|Dolby Atmos||2012||up to 24.1.10|
Looking at the column listing the number of audio channels in the different formats, you may have three questions:
- Why does it say only 4 channels for Dolby Stereo? I saw Star Wars in the cinema, and I remember a LOT more loudspeakers on the walls.
- What does the `.1′ mean? How can you have one-tenth of an audio channel?
- Some of the channel listings have one or two numbers and some have three, which I’ve never seen before. What do the different numbers represent?
Input Channels and Speaker Roles
A ‘perfect’ two-channel stereo system is built around two matched loudspeakers, one on the left and the other on the right, each playing its own dedicated audio channel. However, when better sound systems were developed for movie theatres, the engineers (starting with Blumlein) knew that it was necessary to have more than two loudspeakers because not everyone is sitting in the middle of the theatre. Consequently, a centre loudspeaker was necessary to give off-centre listeners the impression of speech originating in the middle of the screen. In addition, loudspeakers on the side and rear walls helped to give the impression of envelopment for effects such as rain or crowd noises.
It is recommended but certainly not required, that a given Speaker Role should only be directed to one loudspeaker. In a commercial cinema, for example, a single surround channel is most often produced by many loudspeakers arranged on the side and rear walls. This can also be done in larger home installations where appropriate.
Similarly, in cases where the Beosound Theatre is accompanied by two larger front loudspeakers, it may be preferable to use the three front-firing outputs to all produce the Centre Front channel (instead of using the centre output only).
The engineers also realised that it was not necessary that all the loudspeakers be big, and therefore requiring powerful amplifiers. This is because larger loudspeakers are only required for high-level content at low frequencies, and we humans are terrible at locating low-frequency sources when we are indoors. This meant that effects such as explosions and thunder that were loud, but limited to the low-frequency bands could be handled by a single large unit instead; one that handled all the content below the lowest frequencies capable of being produced by the other loudspeakers’ woofers. So, the systems were designed to rely on a powerful sub-woofer that driven by a special, dedicated Low Frequency Effects (or LFE) audio channel whose signals were limited up to about 120 Hz. However, as is discussed in the section on Bass Management, it should not be assumed that the LFE input channel is only sent to a subwoofer; nor that the only signal produced by the subwoofer is the LFE channel. This is one of the reasons it’s important to keep in mind that the LFE input channel and the subwoofer output channel are separate concepts.
Since the LFE channel only contains low frequency content, it has only a small fraction of the bandwidth of the main channels. (‘Bandwidth’ is the total frequency width of the signal. In the case of the LFE channel it is up to about 120 Hz. (In fact, different formats have different bandwidths for the LFE channel, but 120 Hz is a good guess.) In the case of a main channel, it is up to about 20,000 Hz; however these values are not only fuzzy but dependent on the specifications of distribution format, for example.) Although that fraction is approximately 120/20000, we generously round it up to 1/10 and therefore say that, relative to a main audio channel like the Left Front, the LFE signal is only 0.1 of a channel. Consequently, you’ll see audio formats with something like ‘5.1 channels’ meaning `5 main channels and an LFE channel’. (This is similar to the way rental apartments are listed in Montr\’eal, where it’s common to see a description such as a 3 1/2; meaning that it has a living room, kitchen, bedroom, and a bathroom (which is obviously half of a room).)
LFE ≠ Subwoofer
Many persons jump to the conclusion that an audio input with an LFE channel (for example, a 5.1 or a 7.1.4 signal) means that there is a ‘subwoofer’ channel; or that a loudspeaker configuration with 5 main loudspeakers and a subwoofer is a 5.1 configuration. It’s easy to make this error because those are good descriptions of the way many systems have worked in the past.
However, systems that use bass management break this direct connection between the LFE input and the subwoofer output. For example, if you have two large loudspeakers such as Beolab 50s or Beolab 90s for your Lf / Rf pair, it may not be necessary to add a subwoofer to play signals with an LFE channel. In fact, in these extreme cases, adding a subwoofer could result in downgrading the system. Similarly, it’s possible to play 2.0-channel signal through a system with two smaller loudspeakers and a single subwoofer.
Therefore, it’s important to remember that the ‘x.1’ classification and the discussion of an ‘LFE’ channel are descriptions of the input signal. The output may or may not have one or more subwoofers; and these two things are essentially made independent of each other using a bass management system.
If you look at the table above, you’ll see that some formats have very large numbers of channels, however, these numbers can be easily mis-interpreted. For example, in both the ‘10.2’ and ‘22.2’ systems, some of the audio channels are intended to be played through loudspeakers above the listeners, but there’s no way to know this simply by looking at the number of channels. This is why we currently use a new-and-improved method of listing audio channels with three numbers instead of two.
- The first number tells you how many ‘main’ audio channels there are. In an optimal configuration, these should be reproduced using loudspeakers at the listeners’ ear heights.
- The second number tells you how many LFE channels there are.
- The third number tells you how many audio channels are intended to be reproduced by loudspeakers located above the listeners.
For example, you’ll notice looking at the table below, that a 7.1.4 channel system contains seven main channels around the listeners, one LFE channel, and four height channels.
|Left Front Height||x|
|Right Front Height||x|
|Left Surround Height||x|
|Right Surround Height||x|
It is worth noting here that the logic behind the Bang & Olufsen naming system is either to avoid having duplicate letters for different role assignments, or to reserve options for future formats. For example, ‘Back’ is used instead of ‘Rear’ to prevent confusion with ‘Right, and ‘Height’ is used instead of ‘Top’ because another, higher layer of loudspeakers may be used in future formats. (The ‘Ceiling’ loudspeaker channel used in multichannel recordings from Telarc is an example of this.)