Continuing from Steve Howell's series on editing, Simon Croft explains how he uses these techniques when producing.

The process of mixing is vital. It can make the difference between a perfunctory demo and a song that has real impact. In this article we will be analysing the ingredients of a successful mix and explaining some of the techniques you can use to achieve these results. Though the type of equipment you use is relatively unimportant, you will find it very difficult if not impossible to do complex edits on a cassette machine so the use of an open reel recorder is assumed.

We are undoubtedly in the age of the remix. There are engineers who spend a great part of their working life just remixing material that was recorded elsewhere. Some record company executives seem to spend more time thinking about the remix than they do the song itself. If a song unexpectedly flops, they commission a remix. If it's a hit, well, there's the 12" to consider and maybe a 'harder' version for the American market.

Remixes undoubtedly helped Frankie keep 'Two Tribes' in the charts, thanks to Trevor Horn's unceasing ability to construct yet another version.

Example

To provide a reference point, we will assume throughout this article that you have a song on multitrack that looks something like the example in Figure 1 and work through it, rather like Steve did when presenting his approach to producing extended re-mixes. As you can see, it's a very simple structure. For that reason, it should give plenty of opportunity for imaginative treatment.

Before we get to specifics, let's take a look at some important points in a mix.

Our example has no more than a count in, or a couple of bars of drum machine, before the first verse. That's hardly the sort of thing that will impress a record company executive or get Joe Public reaching for his wallet. Those first few seconds are important.

Most hit singles have an instantly identifiable 'tag' at the start. Remember the strident guitar chord that starts 'A Hard Days Night', the a cappela vocal intro to 'Solid as a Rock', or the initial orchestral sounds on 'Slave to the Rhythm'? All these records grab your attention from the first second.

All the Same

One of the failings of demo mixes is the lack of variety. The whole band comes in on the first bar and they stay at that level until the end. Only the changing lyric identifies one verse from another. It is a fact that if you've got 16-tracks you'll probably fill them. Mix the same way and the arrangement sounds repetitive.

A lot of chart material has a very sparse first verse. Think of the way 'In the Air Tonight' by Phil Collins builds up. If those beefy drums had been there from bar one, the whole effect would have been lost.

Of course, some songs benefit from an immediate display of high energy. You'd probably want to get to the point pretty quickly on an up-tempo dance record. In such a case, consider making the arrangement more sparse later on. In our example, one verse has a solo over it. This might sound exciting on stage but the recorded arrangement could sound more effective with less going on. Listen to 'Pass the Dutchy' or 'Don't Mess with my Tutu'. Both of these have 'breakdown' sections, created by muting large parts of the mix.

Over and Out

It is generally considered 'uncommercial' to end a song with a theme that has not appeared earlier on. It is also considered bad form from a more structural point of view. There are exceptions to this, such as the unexpected 'I do like to be beside the Seaside' reprise on 'Seven Seas of Rye' by Queen. In that case though, it is really nothing more than an addition to the basic structure of the song.

Sometimes the unexpected is used for effect. The sudden cut out on 'Virginia Plain' by Roxy Music is a good example.

If you want a song to end by fading a repeating chorus, try to avoid placing two identical choruses back to back. Even introducing a tambourine can provide some lift the second time around.

Start By Listening

A successful mix starts with an objective view of the song on multitrack. Don't just listen to a rough mix, get the multitrack and pick out interesting parts. Get familiar with the location of these parts on tape and the general arrangement.

By the same token, the end of a long recording session is not a good time to start a mix. You may be very familiar with the tracks on tape but you are in no state to be objective. After several hours, your concentration is probably wearing thin and your ears could do with a rest. It's much better to make a rough mix for later consideration and start afresh another day.

As far as sound balancing is concerned, the subject has been covered in some detail over the months so I won't go into too much detail here. Being familiar with the EQ, pan and fader functions, you have probably obtained some pretty reasonable results already. However, it's easy to lose sight of the basic considerations when you are dealing with new techniques. So let's short-form some of the main points.

The balance between the main vocal part and the other tracks is crucial. Even full time producers can go astray on this at times, (doubtless through worrying too long about the perfect snare sound!) Try listening to your mixes on a car cassette player. It's surprising what the combination of small speakers and engine noise tells you about the basic sound balance.

A touch of compression is always useful (if not essential) for keeping the vocal constant in the mix. A small reverb setting can be effective in the verses, where the singer is basically telling a story, with a bigger reverb to build the chorus up. This is the generalisation but can prove to be effective.

When it comes to drum kits, the snare is often the most important ingredient but must be considered in the context of the song. A bigger snare isn't always a better snare. Just imagine what the power of a gated reverb snare would do to a Sade track.

When EQing bass drum, pay attention to the way it sounds with the bass instrument on the track. Can you distinguish one from the other? Bass instruments are best panned dead centre so there is no spatial separation to help. At low frequencies, any stereo effect is negligible so you might as well have the low end handled by the left and right speakers equally.

Overpowering hi-hat is a dead give away on drum machines and tends to create monotony unless the programming was exceptional and equally avoid boosting the top end on hi-hat and cymbals to the point where the tape saturates.

That leaves the mid-range. In general, bear in mind that EQs do have cut as well as boost. You can sometimes clean up a sound far better by removing the irritating frequencies than boosting the desirable ones.

If you have tracks that are recorded as stereo pairs, the way you pan them depends to a certain extent on just how they were recorded. In theory, two tracks recorded with a crossed pair of mics will have the most realistic image panned hard left and right, as they would be if you had gone direct to ½-track.

Watch out though for close miked pianos and electronic keyboards with stereo outputs. They can start to sound ridiculous if you make them appear 20 feet wide. As any one who has listened to Jeff Beck's 'Blow By Blow' album in headphones can tell you, hard panning a Rhodes piano when the stereo vibrato is on really gets on your nerves.

Delay units are versatile things. You can of course fatten up sounds by panning a short delayed version of the original to the opposite speaker. Unfortunately, you can also make the track sound really imprecise if you overdo things. It's always worth making the effort to get the delays in time with the music.

For the more mathematically minded, (Any Ben Duncan fans reading this — Ed.) if you know how many beats per minute there are in your song, you can work out a suitable delay precisely. Take 60 (secs) and divide by the 'BPM'. For example 120 BPM will give you an answer of 0.5 seconds or 500mS. If you think about it, that's exactly the time each beat takes in milliseconds. You can use any multiple of this delay time and know that your echoes are not going to conflict with the speed of the track.

The Gear

To follow the rest of this article through on a step-by-step practical basis, you will need to have at least the following:

Multitrack and mixer (combined or separate).

Open Reel stereo recorder: preferably half track. If you have 1/4-track, use virgin tape.

Outboard gear: up to you but a reverb or delay would help to illustrate certain points.

Monitoring: a set of cans will do.

Editing kit: splicing block, single sided razor blade, chinagraph pencil, splicing tape, leader tape and a couple of empty reels. Tape.

The Plan

For the sake of clarity, we are going to use the typical song in Figure 1. If you look at Figure 2, you will see that a substantial rearrangement has been made. Some parts are used several times, others have been discarded.

The technique that allows this to be done is often called 'mixing in sections' and was covered in detail by Steve Howell last month.

It allows complex changes in level and effects settings that could only be executed otherwise using a wide range of outboard units and a fully automated desk. Some mixes can be done in one pass while others would need the combined qualities of a concert pianist and a human octopus to carry off in one go!

Mixing piecemeal, maximum use can be made of the effects you have. When a track does not appear in a certain section, an effect unit can be performing a different task.

The greatest benefit though is that of artistic freedom. Sometimes it's hard to see the wood for the trees. It can be easier to sit down after the recording is done and decide the best possible arrangement of the component parts.

This is not a licence to make sloppy recordings or forget about the arrangement until afterwards. No one would claim that the technique can rescue utter dross and you will save yourself a lot of hard work if you plan things out in the first place.

Going back to Figure 2, let's talk about why you might want to make such radical changes.

For one thing, our example song has no intro on the multitrack. Intros can often be created by isolating an interesting lick or motif somewhere else on the tape. In this instance, we are imagining that the intro will be taken from part of the solo. To make the effect more catchy, the phrase is going to be repeated by joining two identical mix sections together. Repeating several small sections will give an effect similar to back tracking a record or triggering a sample.

As an alternative, you could try a sound effect like a door bell or a spoken word. Try and avoid the obvious for songs about motorbikes, unless you are hell bent on yet another remake of 'Leader of the Pack'. Similarly, bits out of news broadcasts are getting somewhat clichéd. Steve's trick of taking a sustained chord and reversing the tape can be quite effective. If you remove what will now be the end of the chord and splice the remainder onto the first beat of the song, it creates an interesting crescendo.

Moving to the first verse, you will see that it is abridged. The chorus is normally the catchiest part of the song and the verses are sometimes mundane by comparison. Chopping a couple of lines out of the first verse gets you to the chorus that much faster. Whether this is an appropriate tactic depends very much on the melody line and the importance of the lyric. In our example, it's the end of the verse that has been deleted. The start, or even the middle, may sometimes be more effective.

The first chorus is where you would expect it to be. You will notice too that the first chorus has been used every time a chorus is required. Sometimes you will find that one take is definitely superior to the others. Perhaps the vocal phrasing is better or the harmonics tighter. As choruses are fundamentally identical, there is no reason why you should not use the best version every time round.

A fuller arrangement is used for the second verse, creating a bigger sound and introducing elements that have not appeared before. This leads into the second chorus, which is identical to the first — although the mix could be different if you wanted.

Because the original arrangement only had two verses, the first verse is repeated. This time though, it is used in full with the complete backing.

Instead of a solo, a 'breakdown' version of the chorus is used. This could start with bass drum, vocals and handclaps to give a sing-along feel. As the chorus progressed, other instruments could be introduced, building up to the final chorus. The song ends on a fade out. In reality, you could well need more material to fade on, depending on the chorus length.

That is basically the plan but do bear in mind last month's comments about signal levels, keep the rhythm section constant from one section to the next otherwise you'll end up with some obvious level shifts which is not normally what you want.

Let's review what you have just done. You have set up the mix as if you were going to do it in one pass. Therefore, you have set the level to tape for the loudest section of the mix. By using Mute and subgroups for level changes you have ensured that any section of the mix can be duplicated by pushing a few buttons.

Provided that you do not alter the fader settings after you have started mixing onto the stereo recorder, you should not have any tell-tale jumps in level.

Timing

Now that so many recordings are done to click track, or sequenced from SMPTE code, people are very sensitive to inconsistencies in the beat. Inevitably, you will make edits that are not really precise. When this does happen, carefully peel the edit apart and start again. Don't just say 'Oh, it will do.' Once you have altered all the settings on the mixer, it will be much harder to go back and replace a section. If you tackle the problem straightaway, you can always re-record a section that is beyond repair. Most times though, you will be able to redo an edit without any audible problem. Lay any discarded tapes over the back of a chair. That way, you won't find you have trodden on them if they are needed again.

For editing you will find standard play tape far more durable than extra or extended play varieties and of course, the higher the tape speed, the more precise the edits will be.

Edit Technique

To get good, tight edits, I find it best to use the bass drum as a guide. Put your stereo machine into the edit mode, where the tension is taken off the reels but the tape is still in contact with the heads.

If you rock the tape backwards and forwards, by turning the NAB adaptors, you will be able to home in a particular point. Listen out for the distinctive 'woomph' that the bass drum makes. Because the tape is running so slowly, this is much lower and softer than you would imagine. Occasionally, you may get confused and pick out a tom or even the snare. Don't worry, it is one of those things that gets better with practice.

Don't forget, even a sharp attack sound like bass drum has some leading edge to it. The sound does not appear from nowhere but builds to full volume over a short time. Be careful not to cut that leading edge off. Rock the tape until you are sure that you are marking a fraction in front of the beat.

Sometimes, you have to edit a tape that either has no percussion or has a heavy synth bass, making the correct beat hard to locate. You can always try editing 'on the fly'. Play the tape through as normal and make a quick dab with a sharp Chinagraph when you hear the edit point you need.

Do also make sure that you are marking in the middle of the Play head. Someone who looks unlike the writer has been known to mark against the record head in absent minded moments. This obviously ruins an otherwise okay edit.

Inch by Inch

Finally, lets run through the mix in Figure 2 and see if there are potential problems.

The example intro is constructed from a phrase on the multitrack, repeated for emphasis. Step one is to locate that part and record it twice onto the stereo machine. Two possible difficulties arise, both related to the use of echo or reverb.

The phrase you are recording will be the start of the number. You will need to give both your tape machines a bit of 'run in' time, to make sure they are at the correct speed. There is a danger that your starting note will be shadowed by an echo or trace of reverb from a previous note on the multitrack. If this is a problem, either mute your return channel from tape or the effect input until the desired moment.

The second problem concerns the end of the phrase where it joins verse 1. If you have a significant amount of reverb or delay, you may find that this is chopped short by editing and sounds unnatural. Sometimes this is an interesting effect but it can often be eliminated if necessary.

To prevent the sudden cut-off, the reverb will have to continue at the start of the next section, (in this case the start of the first verse). For this to happen, the opening phrase will need to be on the multitrack tape, in the correct physical location. As this is the start of the song, nearly all the tracks should be free before verse 1 begins. If you are careful, you should be able to bounce the opening phrase from the stereo recorder onto the multitrack. Even if you do not get the timing dead on, it should still place that last bit of 'hangover' in about the right place.

You may get away with bouncing the phrase, reverb and all. It would be better to bounce the phrase, onto the stereo machine and back, 'dry'. It then serves only as a trigger for the reverb when you record verse 1.

Because the opening phrase may not be very rhythmic, try marking the join points 'on the fly' as explained earlier. While you are making your first few splices, you might as well locate the starting note and join on some leader tape.

The first verse should be fairly straightforward. The mix is not particularly busy and there is a steady beat to edit against. If you're shortening the verse, watch out for the vocals. You may find that the next vocal line comes in before your edit point. Mute, rather than be left with a random syllable.

As the chorus is likely to have plenty of reverb, there should be no problem with 'hangover'from the verse. If there were, the end line of the verse on multitrack may well be similar enough to provide the effect. Otherwise, drop in the vocal line if you can find a spare track.

Going from the chorus to verse 2 should be no problem. The only reason there may be an edit point there is to allow routing and effects changes.

The breakdown chorus is the next point that may be slightly complicated. That would be another point to watch the reverb at the end of the previous verse. To get a build at the end, you might need to ride the faders for some tracks.

As stated before, that should be done at the subgroups, not the channels that return from the multitrack.

And so to the fade. Because you may want to have a number of attempts before finding the best fade time, continue to avoid the main faders. Fade from a couple of subgroups instead — but watch how your effects are routed. It will sound pretty strange if the music fades, leaving the reverb because it is still fed to the main L/R outputs.

Splice on some leader for a hiss-free end to your creation and spool it on to a spare reel. Don't forget to label it, otherwise you could accidentally record over your masterpiece.

Next

Now that you understand the technique, you can apply it in a number of ways. To practice tight edits, try recording a track from an LP. See if you can remove a section, such as a verse, without the join showing.

After that, dig out some of your old multitrack tapes and see what improvements you can make. Better still, borrow a few tapes from someone else and astonish them with your new version!

Editing takes a little time to master and your new mixing techniques will not always work at first. Keep at it and you will have a valuable new skill to help you shape your music.