Recording Techniques (Part 14)
Part 14: Effects, in the latest installment of his studio cookbook. David Mellor looks at ways to-make your recordings more tasty.
As we all know, variety is the spice of life, and the variety of effects in your rack can have the same effect on your music as the spice rack in your kitchen — helping the flavours of the instrumental and vocal sounds to blend together, or adding a bit of fire to your productions! We have so many types of effects at our disposal these days that it's almost confusing, and it can be a time consuming process to choose the preset that adds that special 'something' to a musical line. So, when there's so much on offer from the various manufacturers, it's probably a good idea to step back and look at what the basic effects are, why they were invented, and what they can do. With this knowledge, you can develop your own combination of effects from basic ingredients, just as a chef will carefully select from his range of herbs and spices rather than casually throwing in a handful of curry powder. But where do we start? Let's go back in time to a point when recording techniques were so primitive that there were no effects at all, and ask the question: Why do some of those early recordings still sound so good?
The one effect that has been around almost since the beginning of recording is compression, which I dealt with in a previous part of this series. In fact, compression was not originally an effect as such; it was developed because of the need to compress the dynamic range of real life sounds to suit the restrictions imposed by early recording media. It was only later that compression was used as a true effect to enhance sounds, rather than solve a problem. The same applies to equalisation, which was invented to compensate for shortcomings in the frequency response of equipment, and later found favour as a way of adjusting the frequency balance of a sound or a mix.
Before the 1960s, engineers just didn't have a need for effects as we do now. The men in white coats didn't go around saying things like, "Bing's voice could do with a touch of Quadraverb", in the studio canteen. They had enough problems getting the sound clearly on to tape in the first place, without messing around with it further. In a sense, the equipment itself produced all the effects you might ever need: EQ from resonances in the microphone; compression and distortion from the tape recording process (and from all the valves in the equipment); reverberation and echo from print through on the tape; time domain effects from the wow and flutter produced by motors and bearings in tape recorders, cutting lathes and record players.
Ironically, as the performance of equipment got better and better, the more we needed effects to thicken up the sound, right up to modern times where synths and samplers send an ultra clean signal directly along a piece of wire through a mixing console whose performance approaches the limits of the possible, on to an almost noise-free digital tape recorder. With equipment this clean, we need a whole rack full of effects units to get back to square one. Is this progress? Well of course it is, because we can now dictate how we want things to sound, rather than having to put up with conditions imposed by less-than-perfect equipment. The reason some early recordings still sound very good is that musicians and engineers adapted their styles and their techniques to the circumstances in which they found themselves, in much the same way as contemporary musicians have exploited the unrealistic sound of a sampler transposing a real instrument up or down in pitch as a musical device in its own right.
Reverberation is a natural phenomenon that improves the perceived quality of even the weakest singers. We all know how pleasurable it can be to sing in the bath, because the high level of reflected sound adds a fullness and roundness to the sound that in reality may be totally absent in the original. It's not convenient or desirable to set up a recording studio in a bathroom — you'd keep slipping on the soap — so it's better to have some artificial way of generating reverberation, or reverb as it has come to be called.
Since a room with hard tiled surfaces produces a lot of reflections, leading to reverberation, studios hit on the idea of building an echo chamber, probably in the basement. An echo chamber (see Figure 1) is a room containing a loudspeaker, one or two microphones, and lots of irregular hard surfaces. Sound is sent down from the mixing console through a power amp into the speaker, picked up by the mics (which are angled away from the speaker), and put back into the mix. It's an obvious thing to do when you think about it. In actual fact, the sound produced by a real echo chamber isn't all that brilliant, mainly because small rooms have strong resonances which colour the sound too much.
Nevertheless, it was a lot better than nothing, and echo chamber reverb has been used on many recordings. There is one interesting point about echo chambers that is still relevant to contemporary effects — the feed to the chamber need only be mono, to a single speaker. If there are two mics, they will each pick up a slightly different combination of reflections, thereby giving an entirely adequate stereo effect. The benefits of having a stereo send to an echo chamber, or to a modern equivalent unit, are sonically so slight as to be negligible.
Because echo chambers were rather bulky and not easily transportable (!), various mechanical alternatives were tried. The two that actually worked were the spring and plate reverbs. Spring reverbs have gained a high degree of notoriety, basically because most examples of the type sound truly awful. Even at their highest state of development, for example in a number of AKG products, no-one could put their hand on their heart and say that they produced the perfect reverb. A spring reverb works by having a loudspeaker-like transducer impart a mechanical vibration to a long spring. This vibration rushes to the end of the spring, bounces backward and forward along its length, and generally gets very confused. The sound that is picked up and transformed back into electricity and then into an audio signal has a lot of the qualities you would like reverb to have — plus some qualities that you wouldn't wish upon your worst enemy. Spring reverbs, to put it bluntly, twang. Well, you'd expect it I suppose. Percussive sounds and spring reverbs just do not go well together, because you get that horrible twanging sound on every beat. The spring reverb that you are most likely to come across secondhand these days is the Great British Spring. I had one myself when digital reverbs were fantastically expensive (and still not very good). The Great British Spring's main advantage was that its sound was very dense, and it really gave a good thickening effect on sustained synth voices. Even now, a digital reverb can be judged by how much thickening power it provides.
If spring reverbs have never been up to much, the other mechanical reverb is a different kettle of fish. The plate reverb actually generates a good clean, smooth, thick sound. An EMT plate reverb is still worth having in the studio, alongside current devices. The plate works in much the same way as a spring, by using one transducer to make the plate vibrate, and another two to pick up the stereo reverb. If I had room for one, and a few pounds to spare, I would definitely be looking for an echo plate for my own use, because I think they have a sound quality that will always be valuable.
Moving up to date, digital reverbs offer such convenience that they are found everywhere. Someone will probably even make a digital reverb for use in the bathroom one day. There is a lot of difference between different units, and it's something that you can't really find out about by looking at spec sheets: you have to listen. On any unit there will be a variety of programs to simulate different environments such as small rooms, medium size rooms, halls and cathedrals. The essential difference between each program lies in the pattern of reflections, and you are not usually given much control over this. Each program usually has a number of parameters that allow you to modify the effect. 'Reverb time' is simply the time taken for the level to drop from its maximum by 60 decibels. 'High frequency damping' cuts down high frequencies more than low frequencies as the sound decays. (This simulates the sound of a room with wall surfaces that absorb high frequencies more than low.) 'Pre-delay' is a delay inserted before the signal is fed to the effects circuitry. Having a slight gap between the clean signal and the reverb can help to maintain clarity while making a sound more dense.
Going back to basics once again, the original source of delay was the good old analogue tape recorder. Studios would keep a trusty old Revox A77 loaded up with a reel of long play tape, specially to provide echo effects. Delay was added by sending a signal via an auxiliary to the tape recorder, where the signal would be recorded on tape via the record head, and recovered by the separate play head and brought into the mix via a spare channel — the delay effect is a result of the physical separation of the two heads. Repeat echo could be obtained by sending some of the delayed signal via the aux back to the Revox, thus creating a feedback loop. This technique was extensively used by Jamaican dub producers who would ride the echo fader through the song, and occasionally push it up into positive feedback where the echo would get higher and higher in level, and eventually crack up in distortion. The big drawback to working with a tape recorder was that the tape would run out every half and hour or so, depending on what varispeed setting you had used. I had a craving for stereo echo, so I would find myself jumping up and down like a yo-yo during a lengthy mix session, just to take care of the two recorders.
The invention of digital delays provided an answer to this problem, and of course now that it's so easy, the effect is going out of fashion! But even if you hardly hear an echo these days (apart, for some reason, from on Kylie Minogue singles), it doesn't mean that you can't use it at a subliminal level to improve your mix. Now, delay is incorporated into multi-effects units, with a number of useful parameters that give you control over the treatment: 'delay time' is obviously the time difference between the original signal and the repeat. It would be useful if all delay units offered the option of giving the delay time in beats per minute, because that would help the engineer find a musically useful setting, rather than having to work out the number of milliseconds on a calculator (more on this later in the series). 'Feedback' is used within the unit to give repeat echos, rather than going through the desk. To simulate the softening effect of multiple repeats through an analogue tape machine, most delay units — or delay programs on multi-effects units — have a high frequency damping feature, which filters the signal a little on each pass around the feedback loop. What digital delays don't yet have, and probably never will, is a setting to make them sound exactly like old analogue tape delay, which just goes to show that you can rarely completely replace a piece of gear with something more modern.
Time domain effects include chorusing, phasing and flanging; all subtly different varieties of the same thing. Figure 2 illustrates how chorusing can be achieved with an analogue tape recorder with a varispeed facility (this is probably easier than trying to imagine what goes on inside a digital black box). As shown, the tape recorder will produce a single echo and nothing more, which will thicken the sound up a little. But when you apply your musically trained fingertips to the varispeed control, you will find that as you increase the speed, the pitch of the echo will rise, and when you decrease the speed, the pitch will fall. Note that the effect only occurs when the speed is changing, not when you leave it at the faster or slower setting, so you have to continuously move the varispeed knob up and down. This might be a little tedious, but you will find that it will miraculously transform a thin and weedy vocal into something with depth and body. If you could do this with two varispeed tape recorders running out of sync, the result would be twice as good, and you could pan the resulting signals left and right for a stereo effect. Isn't it great that we can get all of this in a rack mounting box these days?
The terms phasing and flanging have changed their meanings somewhat over the years, but there are two distinct effects here so I'll describe both and let lexicographers argue about the terminology. Figure 3 shows a multitrack tape recorder with both sync and replay heads, together with a quarter inch that provides a delay. The signal that forms the main component of the processed audio comes from the replay head, but if a separate output is available on the sync head (as it would be in a pro recorder), then an output can be generated which is in advance of the replay output. This is fed to the delay recorder, whose varispeed is set so that the delay is almost exactly the same as the delay between the sync and replay heads on the multitrack. These two signals are then mixed together at the console. If the delays are exactly matched, then in theory the signals will reinforce each other and simply become 6 decibels louder. But if the delay is slightly out, then an effect known as 'comb filtering' occurs, which reinforces some frequencies while cancelling others out almost completely. This doesn't sound very good as a steady state effect, but when you shift the varispeed up and down, as before, then the sound becomes very rich, lively and interesting. Another way of changing the delay, rather more crude than using a varispeed control, is to press the flange of the supply reel of the delay recorder — hence the term 'flanging'. These days, we would more likely think of the effect as being 'phasing', because of the phase cancellations that occur, and save the term 'flanging' for when positive feedback is added to make the effect stronger. Phasing and flanging processes have of course now been digitised and incorporated into multi-effects units, as has chorusing.
Another time domain effect which has mechanical origins is the Leslie loudspeaker simulation. The Leslie speaker basically has two horns from which the sound emerges, and these horns rotate. At any time, one horn will be approaching the listener and one will be receding. In the same manner as the two-tone siren of a police car seems to fall in pitch as the car goes past (due to the Doppler effect), the sound from the horn which is coming towards you will be raised in pitch and the pitch of the sound from the receding horn will fall. This rise and fall in pitch is not constant, but varies with the angle between you, or any listener, and the mouth of the horn. In theory this should give an effect similar to chorusing, but reality is quite different, and I don't think there is anything on the digital market that can rival the true sound of a Leslie in the flesh. But there will be, one day.
The original Aural Exciter was made by Aphex, and it was considered to be so wonderful that studios couldn't even buy it — they had to lease it by the track-minute! Exciters in general work by adding a controlled amount of distortion to the signal. You can distort the signal in a thousand ways between the microphone and the loudspeaker, but not in quite the way that an exciter does. The secret is to heavily compress the signal before passing it through the distortion producing circuitry, and then add this almost constant amount of 'fizz' to the clean signal. It may sound crazy, but the fact is that it works. Particularly on material that is lacking in high frequencies, the exciter can add more brightness than any EQ HF control.
My last effect for this installment is the pitch changer, or Harmonizer as it was called by its inventors, Eventide. The pitch changer exists only in the digital domain as there is no mechanical or electronic means of doing this job effectively. (I'm told that the pitch changing tape recorder with a rotating head did not work too well, and I'm not surprised.) Pitch changing can never be performed perfectly, at least not at the current state of the art, and some devices will cope better with certain sounds than others. The main use for pitch changing is thickening up vocals by adding small amounts of signal a fraction of a semitone up or down from the normal signal. This produces an effect similar to chorusing, but without the cyclic up and down motion. For a novelty effect, delayed feedback of a pitch-changed signal is great. If you have never heard the effect of a repeat echo that changes in pitch on every repeat, then you owe it to yourself to go down to your local dealer and try it out. It's not for every day usage but who says that serious engineers can't have a bit of fun now and then?
One of the problems with processing these days is that multieffects units make it so easy to try one effects combination after another. You'll sit there thinking "No... not quite... almost... I don't like that..." and so on ad infinitum. I use an ART SGE, and I'm well pleased with what it can do, but there are times when I just want to go back to basics. I plug my Fender Stratocaster guitar into a spare Alesis Microverb (which makes a great high impedance guitar preamp with the reverb turned right down), I take this to my Drawmer DL221 compressor to add some bite to the signal, which then passes to a Fender Champ practice amp (the output of which is cut down from 6 Watts to 1.5 Watts by a series/parallel resistor network to turn some of the sound into heat rather than make a lot of noise). The highly distorted guitar sound that emerges then goes through a chorus program on my Yamaha SPX90 multi-effects unit, and then through a Drawmer DS201 gate to get rid of the noise that has by now built up. A bit of EQ on the console and the sound is...
I almost said perfect, but the point I want to make is that it can be a lot more fun to build up an effect than just to select a favourite preset on a multi-effects unit. Effects provide a way to make your sound different from everyone else's, so go to it: be creative and let effects add spice to your musical life.
Feature by David Mellor
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!