Return To Zero (Part 2)
Following on from last month's glossary of terms, we begin our run through the complicated business that is recording, with a back-to-basics approach to setting record levels.
Following on from last month's extended glossary of terms we continue our back-to-basics series with a subject fundamental to the recording process — setting the recording levels.
Whether you've just unwrapped your brand new four-track cassette machine and can't wait to record your polyphonic synthesiser, or your goal in life is to capture the sound of the Flying Scotsman roaring through York station, the process of making good recordings, no matter what the sound source, rests on your ability to correctly set the recording levels.
Unfortunately, the more you become interested in the art of recording and start to move away from the domestic Hi-Fi approach of machines with built-in automatic record level controls (like Walkmans and mono cassette recorders), the more the medium demands of your personal skills. Making this initial step across into the 'real world' of recording, for want of a better description, is, for many people, a little tricky as several new concepts and a fair bit of technical jargon have to be understood.
There are two basic approaches to recording sound onto tape:
1. Microphone and instrument sound sources plugged directly into the tape recorder.
2. Sound sources plugged into a mixer and then out into the tape recorder.
The difference between the two is that method one involves the setting of only one recording level control, whereas method two may result in up to four separate controls having to be adjusted prior to making a recording. Obviously, the first method is straightforward and more simple but lacks flexibility, whilst the second method has three variables which all need to be set correctly yet, in its favour, offers far more creative potential in terms of signal processing.
So, faced with these two different approaches, you need to make a decision about which method you're going to use. Ask yourself the following question, 'what am I going to record?'. If you purely want to make a stereo recording of classical music or jazz using a couple of good quality mics to capture that live atmosphere, then go for the first method as a mixer isn't required and would unnecessarily over-complicate matters. However, if your need is to record several instruments or sound effects each onto its own tape track and then mix them together at a later date, a mixing desk is obviously very important. In fact, on some of the larger multitrack tape recorders it's obligatory, as they lack input level controls of their own.
Setting a recording level is not a particularly difficult thing to do as long as you follow a few basic ground rules. There is, as they say, no substitute for practical experience. For example, after you've recorded a drum machine and then played it back only to find that the cymbals sound like the pet cat tap dancing on bacofoil, then it's time to improve matters by applying a little thought and some experimentation.
The quest before us (Figure 1) is to record a sound that utilises the full capacity of the tape medium but at the same time avoids the two big dangers that lie in wait: noise and distortion. Just like Indiana Jones in 'Raiders of the Lost Ark', the recordist is faced with two evils. When Jones was trapped in the Egyptian tomb there were hissing snakes covering the floor. In our terms, this is the noise floor of the tape and we don't want our precious sound falling into it. Jones could not escape through the ceiling of the tomb either, as the exit had been sealed. Again, for us, this is the horror of distorting the sound. We can't let the signal go up too high or else it runs out of headroom and overloads the tape to the, point of saturation. Objective: to keep our head down at the same time as avoiding falling into the pit of hissing snakes.
The answer, as Indiana Jones so cleverly realised, is to steer a course between these two evils and progress sideways until safety is reached. Translated into our terms, this means setting the recording level to an optimum point where the peak levels in the sound don't push the tape into the ceiling and cause distortion, whilst at the same time not allowing the signal to fall to the noise floor and be covered in tape hiss.
Finally, just as Jones had a guiding light to help him through his treacherous journey, we too can employ a means of help in the form of the VU meters found on the tape machine and mixer.
Most tape recorders are fitted with VU style (volume unit) meters, usually of the mechanical kind ie. moving needle (Figure 2). All such meters are scaled in decibels (dB) and reading from the left will be marked something like -20dB, 0dB at the point where the scale turns red and +3dB at the right-hand end. Underneath the dB scale, some meters are also marked with a percentage scale -100% being equal to the 0dB mark ie. it represents 100% utilisation of the tape's recording channel.
However, these days this is not entirely true as modern audio tapes are capable of handling much higher signal levels (in the region of +8dB) and consequently offer you a greater safety factor or headroom as it's known. You can therefore pile plenty of signal level onto the tape and record well away from the noise floor.
If you own a tape machine with meters which have a PPM-type (peak programme meter - often a multicoloured LED bargraph) response, and therefore a fast ballistic movement, you'll be able to set the recording level more accurately as this type of meter reacts much faster to peaks giving a more realistic indication of the signal's true record level. Some PPMs also incorporate a peak hold facility which continuously displays the maximum level reached by the signal.
Method one: having plugged the microphone into the tape recorder's mic input socket, you can then begin to adjust the record level control. At this point you need to make a mental note of the type of sounds you're recording as these will affect the setting of this level. Sound sources with fairly fast attacks - drums, bass guitar played 'slap' style etc - can easily be over-recorded as the standard VU meters found on most tape machines don't react fast enough to sound peaks and therefore give a false reading of the actual record level.
What you find is that if the level is set too high, each beat of a snare drum, say, will cause the meter to deflect into the red area and result in a distorted sound. To cure this you turn the level down a bit, however, if you only turn it down a little so that the meter is just going into the red, you may still be over-recording for the very reason stated before ie. a VU meter's inability to react quickly enough. In practice, you may need to alter the level several times and make a few test recordings before the desired result is obtained.
Remember, you should always try to record the sound with the highest possible recording level but without forcing the tape into distortion. This doesn't necessarily mean that you can't go into the red. Any sound with a slow attack will register more accurately on the meter allowing you to really push the record level up high without fear of overloading the tape. Consequently, you can let the needle sit in the red area without any real problems.
When making a live recording where you don't have the luxury of a trial run-through to check your levels, then you should under-record the music. If you over-record the sound it's impossible to reduce the resulting mess from the distortion incurred and restore it to some kind of listenable form. However, if you do under-record the sound, at least it won't be distorted. Instead, it will probably be a little hissy though some of this can be removed later by using such devices as a dynamic noise filter or a graphic equaliser and fairly pleasing results can be achieved.
Method two: when making a recording that uses several sound sources recorded onto different tracks of the tape recorder eg. six microphones around a drum kit, a synth, bass guitar and lead guitar; you need to use a mixer. Having made the initial recording you are probably going to overdub extra instruments and a vocal track which again requires a mixer.
So what about the recording levels (Figure 3)? Almost every mixer has an input gain control usually found at the top of the channel, a fader at the bottom which controls the channel output level, and an overall level control known as the group output fader, usually located at the right-hand end of the mixer; three level controls in all. We've already spoken about setting the tape recorder's own input level but now that we've introduced a mixing desk a further three controls are available and each has to be set to the correct position. Tricky business you might say; well here's how to approach it.
Consider this. Each of the variable level controls mentioned so far are, in essence, small amplifiers in a big circuit. One of the properties of amplifiers is that when you turn them full-up, noise and distortion can occur. Therefore it would make sense not to turn the mixer faders full-up, but to set them around half-way. If you think about it, this is a fairly obvious starting place as it allows you to both increase and decrease the signal level at will, a facility not available to you if the fader is already set fully up but there still isn't enough level.
But what about the number scale written down the sides of the mixer faders? Well, these vary on different manufacturers' products but some are scaled in a similar manner to the VU meter ie. minus numbers at the bottom end moving up to zero and then on into plus numbers. The one thing you mustn't do is go around setting all your faders to the zero position in the hope that it will send the sound to the tape recorder and read zero on its VU meters. The zero mark or your mixing desk faders may well be at the maximum position which means the mixer's own noise would be high and there'd be insufficient room to increase the levels further.
One thing you should try to do is use a test tone oscillator to calibrate your recording level and to line up the mixer with the tape machine. The April 1984 issue of HSR covered this topic in more detail, but here's a brief summary.
With your test tone oscillator set to the required frequency (usually 1kHz), adjust the group output fader until the mixer VU meter reads 0VU. Then with the mixer output connected to the tape recorder adjust the input level of the recorder to again read 0VU and match that of the mixer. Having done this, both the mixer and tape recorder are now lined up, so you know that any signal which registers 0VU on the mixer will be recorded at that same level on tape. Once set, you can forget about the tape machine levels and concentrate on the three mixer level controls. If you haven't got a test tone oscillator, then a pure tone (sine wave) from a synthesiser will do. You can then generate the correct frequency by playing the note C two octaves above middle C, as this corresponds roughly to 1kHz.
Now, having lined up the mixer and tape machine you can proceed to set the levels on the mixer itself. An initial starting point would be something like this.
1. Set input gain control to around half the scale amount.
2. Set all tone controls to the 12 o'clock flat position.
3. Set channel fader to half-way.
4. If a channel needs to be assigned to a group output, press the relevant button to route the channel to an output group.
5. Adjust group output fader to half-way position.
6. Listen and look at VU meters.
If the sound you're recording is being picked up by a microphone, then you'll probably require a little more input gain, so turn it up to about three-quarters of the scale and then check the signal peaks on the meter again. When several signal inputs, such as six different mics around a drum kit, are being mixed together onto one mixer group, it's important to work with the input channel faders to balance each individual sound in the overall mix before you start moving the group fader.
Once the required balance has been achieved, you can proceed to adjust the group fader to give the best recording level on the VU meter. Obviously, had you started with the group fader set on maximum you wouldn't have been able to increase the recording level. Any further increase could only be achieved by moving each of the individual channel faders up a little, which is a pretty dumb move when you have just achieved a perfect balance to the drum sound isn't it?
One final tip. When applying tonal equalisation (EQ) to the sounds, you should be aware of the fact that excessive variations in the equalisation control settings can drastically alter the signal record level. Also, when you use large amounts of treble boost, high frequency noise becomes more prominent if not masked by the sound content.
Next month, we shall look at the use of noise reduction systems and see how they can improve your recordings.
Feature by Paul Gilby