Joint (audio engineering)

For multi-channel audio can be for the purpose of a more compact representation, exploited by channel coupling dependencies between them. Corresponding processes are used in the digital processing means as the digital audio data compression.

By using only the information on the difference to an existing or a new channel is written into a middle channel, the repeated description of common content can be avoided.

Most common is also increasing use of multi-channel systems in dual-channel configuration, ie in the area of stereophonic. According frequently is of English Joint Stereo (in short: JS) and its varieties such as the Pegeldifferenzstereophonie, mid / side stereo ( English) and the parametric stereo out of the question.

These varieties have their counterpart in the traditional audio engineering, eg as intensity stereophony and MS stereophonic. You will in part be associated with different degrees of sound loss. Many users know only the lossless MS method.

The difference signals can be lossless stored ( lossless coupling ) or quantized and coded according lossy. The quantization case falls into the center of the stereo image, where it is well masked by the remaining signal.

Mid / Side stereo

Mid / Side stereo stores the average of the two stereo channels into a central channel (English "mid channel" ) and the respective difference value in a side channel (English "side channel" ). This eliminates the multiple detecting common content and the difference signal has lower complexity. For digitally represented signals is sufficient therefore a smaller amount of data to describe, whereby a data compression is achieved.

This encoding a lossless principle; that is, from the available information ( the stereo channels here ), the original data is reconstructed exactly. Since separated in the difference signals are present in this structure, the area information, but they are also more accessible to data reduction. The data reduction of the space (Page signal) designed so subjectively very low loss, since the observable tend to be more compared and usually more important information in the middle of now separate signal can not be touched them.

The method is widely used in audio data compression with newer lossless codecs such as FLAC as well as in lossy as LAME and can already be found in the original MP3 codec of the Fraunhofer -Gesellschaft again. It is also used in broadcasting at FM stereo.

As for stereo signals usually a high correlation of the two stereo channels is present, the difference value is often very small or zero, and the information can be stored efficiently as in matrix representation.

The reconstruction of the signals can now be done from the available information.

When calculated, the center channel as follows

And the side channel with this formula

Then the left channel calculated from

And the right of

The complete Signal flow on the example MP3 with channel coupling would be this:

Parametric multichannel audio

In the case of highly lossy parametric stereo (English "intensity stereo" ) the signal is analyzed and parametric descriptions stored, from which the decoder can then by appropriate arrangement of certain signal components on a virtual soundstage restore an approximate impression for example.

It is a mono channel and direction vectors for reconstruction of stereo data will only be stored.

Pegeldifferenzstereophonie

The Pegeldifferenzstereophonie (English "intensity stereo" ) is a simple case of the parametric stereo. When you only the level difference between the channels is held; the maturity -related time delays of certain signal components, be disregarded and phase information is lost. However, the phase shift may be occurring is no longer perceptible above a certain frequency. Nevertheless, the original stereo information is lost; the process is lossy and is generally considered only for low data rates, where you instead of mono- coding but would prefer a little stereo information.

In the digital audio data compression it is usually used in encoders only at very low data rates. Many encoder decide based on the selected data rate, whether they want to use intensity stereophony, but there are some, mostly older codecs that rely exclusively on this technique.

Assessment

Especially in the early days of this compression technique by many proponents of " pure " Stereos has been argued channel coupling destroy the stereo feeling and so affects the listening experience.

Often, however, MS stereo were cast out with Pegeldifferenzstereophonie (or MS stereo with data reduction of the difference signal ) in a pot with such arguments, probably due in part due to the unfortunate combination of the two techniques under one name.

Meanwhile seems the number of strict rejecters of the algorithm but declined; this is also due to skillful implementations in modern audio codecs. Especially in lossless algorithms and at lower data rates, the technique is practically established.

In case of a data reduction of the difference channel and the center signal arising as against a separate quantization of the channels the same tonal advantages and disadvantages. Absolute advantage is the consistent direction of dominance that in the data reduction of two separate stereo channels, especially at low data rates often turns out to be a problem because it makes there by migration of the stereo image of your left and right noticeable. However, since joint stereo is encoded in the mid and side signal, here, this effect occurs as a pulsation of the stereo width from center to the side on which much cheaper or just perceived as significantly less disturbing is tonpsychologisch than that the stereo - center, the in the mainstream media area contains most of the important parts Sing Alan, is clearly not be found or it changes its position. Disadvantage is that in the case of a silent page signal, such as in echo chambers, the depth of field suffers to the sides. If they are different, the left and right channels are strong, also arises with MS stereo is worse than the separation by the conventional channels. Since the decision whether stereo is coded as left-right or mid / side signal can be taken in many audio formats for individual blocks (sections over a few hundred milliseconds), select Encoder for such formats the presentation, which can be encoded better unless the user has not set a specific representation.