Shortly after completing the building of Hear No Evil recording studio early in 1990, I was called upon to mix the music for a German cigarette commercial in Dolby Stereo. Andy Day from Dolby came over with their SEU4 and SDU4 encode/decode units and helped me position my spare NS10s in suitable parts of the control room. The final slice of the chocolate surround gateau was the ritual of setting the levels with the infamous Tandy audio level meter. Once Andy had left, the real fun started, and within an impossibly short time I was hooked forever into the wonderful world of surround. Ten years, and probably some 500 mixes later, I'm still in there and still learning as the goal posts keep moving forward from LCRS through 5.1, 6.1 & 7.1 to 10.2.

                     Although mixing music in surround formats is in many ways easier than mixing in conventional stereo, there are many issues to be aware of if you   are to avoid confrontations involving eggs and faces. This article is about some of the things I've discovered both on a practical and theoretical level       in the course of working in surround.. Firstly, the never ending story of the different surround formats.

                     Dolby Stereo/Dolby Prologic/Dolby Digital/DTS/SDDS

Dolby Stereo is a matrixing format that relies on phase differences to encode and then decode the channel information. When mixing for Dolby Stereo you will generally supply your mix as a 4 track LCRS music stem but you must monitor through the encode/decode hardware (codec) so that you can hear the unpleasant effects of the processing. Unless you mix very wide, the matrix will not see sufficient phase difference and will tend to collapse the audio into the centre. Loud bass heavy signals placed in the centre will also have the same collapsing effect. The key factor is this: just because you place a signal in a specific channel does not mean it will be reproduced in the same position after the encode/decode process. Although you may be clever enough to get your perfect surround music mix when listening through the matrix there may well be additional elements of sound effects and dialogue added at a later stage that will scupper your best efforts. The consumer version of the matrixing process is Dolby Prologic; all home theatre amplifiers have it and you should be aware that people rarely turn it off. If you are working in stereo, you should at some point listen to your mix through a prologic decoder so that you can hear how approximately 100 million people will also hear it. And because Prologic works on any stereo signal, this means ALL formats. And if you have mixed & monitored through the matrix, you should label your master Lt & Rt to signify that fact.

                     Dolby Digital is a format that uses a perceptual data compression (about 12 to 1) algorithm to encapsulate the information into an AC3 digital datastream. The inputs and outputs are generally 16bit 48kHz and as such it is a very lossy codec and not ideal for situations where the music will be heard under a high degree of scrutiny. Quality aside, the big advantage is that the surrounds are in stereo and placements will always be maintained. If you put the kick drum into the left surround channel, it will always be played back in the left surround speaker (at least until bass management kicks in, more about that later). And of course, you have a separate sub bass channel.

                     Digital Theatre Systems (DTS) & Sony Dynamic Digital Sound (SDDS)  are the other cinema audio formats. DTS is a 5.1 format that is similar to Dolby Digital, however, the data is stored on a synchronised CD and so the compression is far less severe, about 4 to 1. DTS have adapted this technology to Red Book CDs, and so it is possible to play back DTS 5.1  20 bit encoded CDs on any CD deck with the help of a decoder that takes its feed from the digital output. You can buy a cheap software encoder from a number of manufacturers so at last you can mix in 5.1, encode it, burn it onto a CD, and take it home to play it through your home system! DTS have taken it upon themselves to release about 200 CDs in the format; these range from re issued classics from Steely Dan & the Eagles to 5.1 versions of mainstream releases. Some are great, but I'm sure that I've learnt far more from the bad ones! Whatever their merits, I consider them to be the Bible of surround mixing, and an essential tool to anybody wishing to get involved in the practice. SDDS is the Sony 7.1 format that has 2 extra channels placed inside the left and right front channels. This is a cinema format and not really relevant to music mixing as yet. Finally, there are the Dolby & DTS 6.1 formats, where there is an extra channel at the back for the centre surround, but these don't need to concern us for the moment.

                     Although most engineers will leave the mastering and final encoding to dedicated facilities, a knowledge of some of the issues involved is necessary. Decoders in home entertainment systems have the ability to read coded 'metatags' in the datastream which decide how the audio information will be interpreted by systems that, say, have no centre speaker, or no sub bass, or maybe just a Left Centre Right. The decoder will then downmix the 6 audio channels according to the metatags if there are less physical channels. The algorithm can also apply dialogue normalisation (a type of floating reference level), and dynamic range compression for replaying film soundtracks at low levels (so as not to annoy the neighbours with apocalyptic sub effects late at night). All these various options are decided during the encode process and so obviously will have a great effect on how your final mix will sound on a domestic system and you need to be aware of them when mixing. The home entertainment amplifier will also have an element of bass end managementso that the bottom end will be directed to the sub if you are using    bandwidth restricted monitors.

                     DVDs have a stereo LPCM soundtrack and an optional surround mix. You should be wary about letting an algorithm decide on how your carefully planned 5.1 mix should sound in stereo — it is far better to do a specific stereo mix in the way that you always have in the past and you should do this if at all possible. If you can't go back to the original multitrack elements to mix in stereo, you can create a stereo mix from the six tracks on your 5.1 master by folding in the respective surrounds to the left and right and adding in the centre and sub until you get the rightbalance. However, you've probably spent many years perfecting the art of mixing for stereo and it would be unrealistic to expect a downmixed 5.1 to compete on any real artistic level. Many of the techniques that you normally use in terms of compression, equalisation and stereo placement are no longer relevant because you have a greatly expanded soundfield to play with. No longer are your elements jostling to be heard, they now have space and dynamic range.

                     Monitoring

                     The bass management systems of home entertainment systems will filter off a low frequency element of the five channels, combine it with the sub or .1 channel, and then direct it to the sub bass speaker. If a home system is without a sub, that information is generally ignored. It is therefore important that vital mix information is not placed solely in the sub or you risk the danger of losing it. The normal practice is for just the bass, or maybe some kick drum to be placed there, but never without having it in one or more of the main 5 channels. Historically, the sub bass comes from cinema practice where all the front speakers are full range and the sub was just used for special low frequency enhancement (LFE) for explosions and special high energy effects.

                     I've tried many different combinations of monitors over the years, from having three full range main monitors at the front with smaller monitors of the same type at the rear, to moving the rears to the side, and then substituting all the monitors to five of the same type. I've found that using large main monitors is too overwhelming; it feels like I'm drowning in my very own black forest surround gateau. My current preference is for using five nearfield monitors such as the Genelec 1031s, placed in the positions recommended by the ITU, that is, at 30 degrees and 110 degrees from the front centre position. I occasionally move the rears further to the back for certain projects that require a more ambient type of mix, such as orchestral film scores that need envelopment without too much imaging. I'm lucky that I have a large control room and am able to have the monitors placed freely on stands. This also works for other engineers who may have their own preferences for where they place the monitors. I've also set up a secondary 5.1 system in my live room with a domestic bass management system so that I can get more insight into my mixes in the same way that you switch between different monitors when working in stereo. Whilst a stereo mix has to sound good on any format, whether radio, CD, boombox, car or television, a 5.1 mix will only ever be heard on a home theatre system, or maybe a car with a central driving position.

                     There is also a question of how you actually control the level of six speakers simultaneously so that they level track in a constant way. Although it is possible to set up a group of six faders on the returns from your six track master and then send the outputs of these to your monitors, its both a clumsy solution and uses up too much console real estate. If you have three separate stereo monitoring busses that will each route to a separate pair of speakers you can do it this way, but calibration is awkward, and you're never quite sure if the three stereo outputs are tracking properly. The best way to tackle the problem is by the use of a, surround monitoring device such as the Magtrax. The most basic units will just have 6 line level inputs, a volume knob, and 6 amplifier outputs. Features on more sophisticated units have are one or more external inputs so that you can monitor the outputs of a DVD player and switching so that you can choose to monitor either your console output busses or the outputs of your master 6 track recorder. Some form of output calibration for your speaker amplifiers, and maybe a downmixing switch so that you can hear your 5.1 collapsed down to stereo are useful if only to prove to you that you're going to have to do a separate stereo mix anyway!

                     What do I put where?

                     What is the function of the centre speaker in 5.1? Should you not use it at all, creating in effect a 4.1 mix? On the positive side, the centre channel can be used as an anchor in a way that you can never achieve with a phantom centre. If you move away from the sweet spot in a stereo mix, sounds placed in the phantom centre will also appear to move in relation to the L & R. By using a hard centre you can go a long way to stop this happening, making the image more stable when you shift away from an ideal listening position. It will also sound punchier and you will avoid the 2kHz dip that you get with a phantom centre due to non-coincident wave ronts reaching the ears. You can also go a long way to avoid the push and pull of conflicting frequencies in the same speakers, e.g. a kick drum could be placed as a hard image in the centre channel with a bass line as a phantom centre, equal in the left and right. On the down side, cheaper home systems rarely have three similar speakers in the front and a mismatched centre speaker could easily throw off your whole mix. It is far more likely that only the L & R will be balanced properly, thereby making it much safer to stick to phantoms. This is a situation where analysis of the mixes on the DTS CDs can give valuable insight to what works and what doesn't, but this something that you really have to make up your own mind about.

                     
                     What are you trying to achieve with your surround mix? You should probably be trying to create a large listening area, to provide a high degree of listener envelopment, and to provide a conductor's or an audience perspective so that the mix sounds great without you necessarily being aware that you're listening in surround until you hit that stereo button!

                     
                     Some music sounds best in a natural acoustic space. Classical & jazz are two good examples. To hear an instrument coming at you solely from the rear speakers in a classical recording is disconcerting, and in fact, sometimes downright annoying. What the mixer should be trying to recreate is a natural acoustic that envelops the listener without distracting.The ideal perspective is that of the conductor himself, who has all the instruments wrapped around him in a semicircle; he hears most of the instruments directly, enhanced by early reflections and the general reverberation of the concert hall.

                     The other case is that of a pop record. Here, the instruments, if there are any, have generally been recorded on a piecemeal basis and there is little or no spatial information involved. The engineer has to create the illusion of space by his judicious use of reverbs, delays and processing. This has no basis in the real world and so the engineer can have the freedom to be a lot more aggressive in the placement and use of dynamic panning in his 5.1 mix.

                     It is an extremely interesting exercise to slowly pan from a front speaker to the equivalent rear. If you apply equal level to front and rear, the audio will seem to be coming from 45 degrees in front and not 90 degrees, as you'd expect. If you keep panning towards the back, the sound will then break up so that you can almost hear it as two discrete sources with slightly different frequency content, and then finally it will zip to the rear. This phenomenon is upheld by psycho acoustic research which shows that the ear's frequency response to sounds coming from the rear is radically different from that of sounds coming from the front. This is in great part due to the physical geometry of the outer ear and lobes affecting the frequency response of the ear canal, and it can be of great help to get an assistant to gaffer back your lobes when mixing. I've also found that you can help this problem by slightly equalising the source as you pan from front to back.

                    
                     How many times do you place source hard left or right when it isn't part of a stereo pair? Sources that are placed solely in one channel can sound obtrusive; you become aware of the positioning of the speaker rather than the positioning of the sound. It's more musical to pull instruments slightly;into the room by placing small amounts of the signal in the other channels so that you almost feel you can walk behind them. This also helps to widen the critical listening position. The same principal extends to the use of reverb. It's good to use different reverbs in the different planes. For natural ambience, I tend to use the four outputs from a Lexicon as my master reverb. I pan the outputs to the four corner monitors, but bring in the front pair a small amount so that will be a very limited amount of return to the centre. Any close miked instrument that I then send to this reverb will immediately get a context within the room. I'll then set up reverbs for the different planes according to the content of the mix. I've found that it's better to change the position of the reverb returns so they are in a different place from the source. It's also good to have different reverbs in the front & rears, and don't forget, there's nothing to stop you having a horn section for example, panned between left and left surround, being fed to a stereo reverb that is panned hard right and right surround. A vocal in the centre could also have a delay in the surrounds, to create the feeling of the reflections of a large stadium.

                     Practical issues of 5.1 Vs stereo mixing on same session

                     Many engineers use their stereo mix as the basis for their surround mix either by starting to work on the surround whilst their mix is up on the board or by recalling their stereo mix subsequently. However, there are times when this is not possible. Sometimes it is impractical to mix the surround in the same session as the stereo, often because of budgetary considerations, or because the record company is not yet ready to commit to the format. In this case, a good solution is to lay off elements of the stereo mix to another multitrack format, preserving effects and dynamics that are integral to the sound of the stereo mix. Formats such as Radar and Protools are ideal for this purpose. The elements can then be archived for later retrieval when the time is right; in this way the integrity of the original mix can be preserved even if the 5.1 mix will be undertaken at another studio by a different engineer. This task can easily be undertaken at the completion of the stereo mix in about an hour.

                     Another situation is where it is unfeasible to mix in the same room due to lack of suitable monitoring or limitations of the console itself. While it is   possible to mix in 5.1 on any professional desk, some lend themselves to the process far more easily than others. Digital consoles have a natural advantage because much of their functionality is software based, and it only needs the correct algorithms to mix in the various surround formats; even moderate priced digital desks like the Yamaha 02R and the Mackie D8B have a good implementation of surround. However, the older analogue consoles are more difficult to configure without using up much valuable console real estate. On desks with more than one stereo bus it is possible to use one bus for front L&R and the another for surround L&R with the centre and sub being addressed by auxiliary busses. This is a clumsy situation because it makes panning between front & back difficult, and smooth pans through the centre speaker well nigh impossible. Dynamic panning is also tricky in all but a very basic way. I chose to buy an Euphonix because as an analogue console under full digital control, it gives you full automation of every surround function and the ability to mix in over a dozen differing surround formats: useful for Imax and special event audio systems.

                     I mentioned earlier that in many ways a surround mix is easier than a stereo mix. You now have six speakers with the equivalent of a greater useful dynamic range so that you can get much better bass, more separation between instruments and a creation of space in the mix. When working in stereo, an engineer has to spend much of his time eqing and compressing to fit a large number of signal sources into a stereo perspective so that they can not only be heard, but that they are balanced without masking each other. And of course this perspective is in one plane only. Many of these problems evaporate when mixing in 5.1 because you now have a 3D perspective in which to place your sounds. Although you now have 4 planes between adjacent pairs of speakers (front, rear, left side, right side) you can also bring sounds forward into the room so you can literally think of your space as a stage on which you can place the various instruments. You now don't have to eq and compress just to pull something through on a mix, and strangely, even balance becomes slightly less critical.

                     Surround sound mixing is still in its infancy and sadly there are many aspects that I've had to pass over through lack of space. At the moment there are limited outlets for surround sound mixes; to whit, music for film, concert remixes for DVD, and albums released on the DTS CD format. However, consumers who have bought their home entertainment centres with a wide screen television and all the associated paraphernalia will quickly become accustomed to listening to films and film music in surround and will expect to hear the same quality and aural spaciousness from music albums. Lately, after several weeks of mixing solely in 5.1 I had to go back to stereo, and my sense of loss was palpable. Moving back to 5.1 was like coming home.