These short clips from well-known classics were programmed entirely within the Garritan Personal Orchestra (GPO4) including the use of the built-in reverb, Ambience. Other more expensive convolution reverbs could have been used but it’s so convenient and fast to work within one mixer when time is limited. Garritan are, apparently, considering the inclusion of a convolution reverb within GPO 5. What a good idea.

For many, writing the music is the easy bit. The difficult part is getting everything hooked up so that your music exactly fits the director’s video frames. Split second timing and rock solid synchronisation is essential.

Even if you’re not interested in becoming the next Hans Zimmer, you may well need to synchronise computer generated MIDI tracks with an external audio recorder. For devices like audio recorders, video decks and sequencers to be considered synchronised they must run together at exactly the same time and tempo. And in order to do that one of them – the ‘master’ – must dictate the tempo to the others – the ‘slaves’.

Role play

For example, you might set up a video recorder to transmit time code to a sequencer like Cubase. In that case Cubase would be the ‘slave’ and the video recorder the ‘master’. Of course, you could just as easily set them up the opposite way around with Cubase as the ‘master’. Furthermore, you could have Cubase act as both ‘master’ and ‘slave’ by having it transmit MIDI Clock to a drum machine whilst simultaneously receiving time code from the video recorder. But what exactly is MIDI Clock and time code, and what’s the difference between them? First a little history.

Impulsive behaviour

Back in the 80s, before MIDI came along, electro musicians used to synchronise their analogue sequencers and drum machines by sending regularly spaced electrical impulses (Pulse Clock) from one to the other. Before long a method of encoding these impulses known as FSK code (frequency shift keying) was developed and boffins in studios across the land were ‘slaving’ their beat boxes to tape recorders.
Unfortunately FSK had no position indicator and you always had to start at the beginning of a song (no intros or outros). Once MIDI arrived on the scene FSK was soon replaced by MIDI Clock, a tempo based synchronisation signal.

Used In conjunction with SPP (song position pointer) a ‘master’ device could now send 24 MIDI Clock messages for every quarter note (12 for every eighth note and so on) and our electro musicians could now start and stop their drum machines anywhere within a song and vary the tempo. All modern devices still use this system today for tempo based synchronisation.

Naming and framing

With the demise of FSK the recording industry adopted a method of synchronising used in the film world and began ‘striping’ analogue tape with SMPTE time code. Laid down as a standard for encoding film frames by the Society of Motion Picture and Television Engineers, SMPTE (pronounced ’simptee’) sends a 24 hour clock type of synchronisation related to hours, minutes, seconds and frames (divisions of a second). With SMPTE seconds are not divided into tenths and hundredths but into frames.

You’ve probably seen the abbreviation EBU. It’s exactly the same time code as SMPTE and was renamed by the European Broadcasting Union when they adopted it for use with European frame rates.

As with most synchronisation formats, nothing’s ever perfect. Because the different rates of alternating electric current varies from continent to continent (USA: 60Hz, Europe: 50Hz), so too do the frame rates and 24, 25, and 30 fps (frames per second) are now in common use. For this reason, when you connect up a series of devices they must all be set to identical frame rates.

Logic’s Synchronization Window (General settings) contains sync parameters for running Logic as a ‘slave’.

Types of time code

There are three main types of time code. VITC (Vertical Interval Time Code – pronounced ‘vitzi’) is a video format, stored in the actual visual images. MTC (MIDI Time Code) is a translation of SMPTE which is transmitted via a MIDI cable. LTC (Longitudinal Time Code) is another name for SMPTE/EBU and is recorded onto the audio tracks of both audio and video recorders.

Although LTC can be output from an audio recorder during fast-forwarding and rewinding it can only be output from a video recorder at the normal playback speed. Obviously this isn’t a satisfactory state of affairs when adding frame-synchronised music to the picture. To do that you need a synchroniser like Emagic’s Unitor8 that reads VITC. In fact it reads and writes both LTC and VITC so it’s good for both audio and video work.

Sound to picture professionals generally use a hardware synchroniser like the Emagic Unitor8, Neundo Time Base (the first pro synchroniser to support VST System Link) or the MOTU Digital Time Piece. These machines will read and write all the different forms of time code and synchronise most digital audio equipment with knife-edge precision.
The MOTU Digital Timepiece synchronizes digital audio equipment with knife-edge precision

Talking clock

If you happen to work in a completely virtual environment you’ll never need to worry about external synchronisation but it is, of course, happening in the background. Audio, MIDI and video is synchronised by your computer’s very own digital audio clock. It’s usually a very stable situation. However, this stability is weakened when you use external time code synchronisation.

Although your sequencer and recorded audio tracks play back in perfect sync (because you’re using the computer’s clock) there’s no guarantee that the system synchronised to time code will do the same (your video or audio tape and MIDI equipment). There’s room for drift. What’s needed is a single master clock, to control the whole system.

Word Clock is purely a timing reference for digital audio devices and is really just a replacement for the sample rate clock in your audio card. It doesn’t carry any time code information and can’t be used to synchronise devices on its own. It can however provide a common timing reference to multiple devices enabling them to be synchronised with one another. Devices that use word clock include digital mixers, stand-alone hard disk recorders, computer-based digital audio workstations and computer audio cards.
Apart from using Word Clock to stabilise a synchronisation set-up you can also use it when making digital transfers. For example, to make a transfer between certain digital audio devices – for example, a MOTU 2408 audio interface and a digital mixer – they must be connected via word clock cables. Doing so will ensure that both devices are synchronised to the same rate of digital audio, enabling consistent playback. In fact the combination of SMPTE and word clock is the ideal sync scenario.

Machine control

Yet another type of synchronisation called MIDI Machine Control (MMC) enables you to control all your studio devices from a single MIDI source, usually a sequencer, by sending three simple messages – Play, Stop, and Locate. For example, you could have Cubase or Logic send an MMC ‘Play’ message to a synchroniser like MOTU’s MIDI Timepiece AV. The AV will respond to the message by generating and distributing time code to all your linked devices, perhaps an ADAT (via an ADAT sync cable), a hard disk recorder (via MIDI cables), and the computer itself (via serial or parallel cables). Of course MIDI Machine Control can also be used to operate just a single device like a hard disk or tape recorder.

Those clever folks at Steinberg have recently developed a network system for digital audio called VST System Link. It can be used to synchronise several computers without Ethernet cards, hubs or the usual paraphernalia associated with networking. You can use digital audio and MIDI cables instead as long as each computer in the system is equipped with a suitable ASIO compatible audio interface.

Using VST System Link you can do all manner of powerful things like having one computer run your VST instruments while another another gets on with the task of recording your audio tracks. Or, if you needed extra audio tracks in a project, you could free up CPU demands on one computer and add the extra tracks on another machine.

Another possible scenario might be to set up a single computer to act as a virtual effects rack, and run all those CPU intensive plug-ins we so dearly love to use. It’s also a cross-platform system so you can take advantage of effect plug-ins and VST instruments that previously, you could only use on either a Mac or PC.

Wired for sound

Propellerhead Software’s Rewire and Rewire2 protocols are yet another form of synchronisation and provide sample accurate audio synchronisation between programs like Reason and Cubase SX/SL. Any two Rewire compatible programs can share the same sound card, transport controls, automatic and audio mixing controls. You can also route MIDI tracks from Cubase to Reason and control it’s various built in synthesisers.

Tech terms

Drop-frame dfps: A type of SMPTE time code used in the US for colour video production. Every minute ( except the 10th minute), two frames are dropped, to match the ‘clock on the wall’.
Striping: When SMPTE time code is recorded to tape as an audio signal, recording engineers refer to the process as ‘striping’ the tape.

Sync track: A track, reserved on a multitrack audio recorder for SMPTE time code.
TDIF: Tascam Digital Interface – a protocol for sending digital signals to and from Tascam multitrack recorders like the DA-88.

S/PDIF: Sony Phillips Digital Interface. Uses either coaxial cable with RCA connectors or optical cable with TOSLINK connectors.

Copyright 2006 Keith Gemmell

Mello-traumatic

With the advent of multitrack recording, rock musicians decided to add ‘strings’ to their records themselves, using electronic keyboard instruments such as the Mellotron. This extraordinary contraption was notoriously unreliable, which is not surprising really because each note on the keyboard had it’s own six foot length of tape which played back the sounds of pre-recorded violins and cellos for up to eight seconds.

Arguably the forerunner to the modern sampler the Mellotron was a favourite with the Beatles and the Moody Blues, who used it on just about every song they recorded in the 70s. Jean Michel Jarre still uses his Mellotron. Rick Wakeman, allegedly, burnt his.

Strings in a box

By the mid 70s dedicated string machines arrived on the scene the most famous being the ARP Solina String Ensemble. Using cheap electric organ technology it was renowned for it’s warm lush sound, achieved with a sophisticated built-in chorus circuitry. God knows what it would have sounded like without it. Used by the Eagles and Elton John and featured on numerous disco hits, it didn’t sound anything like a real string orchestra but was suitable for background pads or slow moving melody lines.

Enter the sampler

Things changed in the 80s and 90s with the birth of digital sampling technology. For the first time ever musicians could write and record complete orchestral arrangements using samplers like the Akai 1000. Using a sampler, an Atari computer and a sequencer, musicians could now add the sound of real strings to their songs.

In the early 90s EM-U Systems took things a stage further and used their high end sampler, the Emulator 3, to record a realistic set of string samples and stored them as onboard ROM in the Proteus/2 Orchestral sound module. An updated version is still available under the name of the Virtuoso 2000 and provides a quick and easy way to knock up a realistic string orchestra.

As a concept the sampling process is very simple. For example, a single violin or a complete string section makes a sound that strikes the diaphragm of a microphone and a corresponding voltage is generated. Analogue to digital converters measure the voltage which is then stored in memory as a series of numbers. To play back the sound, the numbers are converted to voltages again with a digital to analogue convertor. Finally they’re amplified and sent to a speaker which converts the voltages to sound waves.

Meet the orchestra

Before you embark on writing string arrangements it’s important to get to know the instruments themselves; how they’re played, their sound characteristics, their playing range and so on.
The string orchestra consists of four sections covering the complete musical spectrum – soprano, alto, tenor, bass (SATB).

Violins usually carry the melody and cover the high, soprano range. They’re usually divided into two sections – violins 1 and violins 2. A safe, practical working range for the violin section is from G2 (the lowest open string) to about C6 on a MIDI keyboard. When played in the high register the second violins are often used to double the line at the octave below, to provide strength and cohesion.

Violas cover the mid, alto range and are darker in character than the violins. A safe, working range for the violas is C2 to about C5 on a MIDI keyboard. They’re used mainly to provide the inner harmonies of a chord and can be divided into sub groups for this purpose. Viola players are rarely featured and other musicians frequently tell ‘viola’ jokes behind their backs (all in good fun, of course).

Cellos cover the lower ranges. They’re strong and powerful in their low, baritone range and expressive when played in their higher, tenor range. They’re sometimes used to carry the melody. A comfortabe playing range is from C1 to G4.

Double basses, of course, cover the bass range and frequently double the cellos at an octave below, for extra strength. Their playing range is C0 to C3 but they rarely go very high.

Painting by numbers

For many musicians the cheapest and easiest way to re-create strings is to use a General MIDI sound module like the Roland Sound Canvas. Used carefully, they can provide a pretty convincing string orchestra mock-up.

The basic GM sound set contains individual instruments starting with violin, at program number 42 through double bass at 44. The quality varies considerably depending on the manufacturer but generally speaking, they don’t sound too realistic.

Far better are the ensemble strings at program number 49. Using just this program will produce a decent all round string orchestra from violins to double basses. Strings 2, at program number 50 generally have a slower attack time and are better for slower passages.

Programs 51 and 52, usually named Synth Strings 1 and 2 are more synthetic in character and better suited to keyboard style pads and backgrounds.

Program number 45 contains tremolo strings, an eerie effect produced by short rapid back-and-forth bow movements near the bridge. Number 46 houses the pizzicato strings, the sound of players plucking the strings with their fingers.

The Halion way

Many film composers write their string scores using a sequencer like Logic and a high end sample library before recording a real orchestra. Unfortunately most of these libraries cost a absolute fortune. However, with a moderately powerful computer and the Halion String Edition Vol.1 you can do the same. It was developed by Wizoo as a tool for arrangers and composers orchestrating their scores on a budget and is supplied with 8 CDs full of very useable, high quality string samples.

The beauty of a sampler player such as this is that much of the hard work has been done for you with a variety of bowing styles and articulations on offer, to help create authentic sounding parts. Legato, pizzicato, spiccato, tremolo and trills are all there. For beginners, special ‘4 in one’ programs are included, where all four sections (violins, violas, cellos and basses) are combined into one. Of course, that’s a compromise but once mixed into the background (as they usually are) only the trained ear will spot the difference.

Keep it simple

Scoring strings can be a daunting task for the beginner. Listen to recordings and read books on the subject. But don’t allow the lack of technique to prevent you getting started. String sections are in vogue at the moment and feature on quite a few chart hits but listen closely and you’ll discover that more often than not a single melody line is all that’s necessary. As a rule of thumb keep it simple behind the vocal with flowing lines and leave the flowery bits for filling gaps and maybe solos.

Shaping and scraping

If you’re into creating and programming analogue synths and you’re not too concerned about authenticity, it’s fairly easy to create string sounds. Choosing the correct waveform to begin with is crucial. If you read Ian Waugh’s earlier Ten Minute Master on Synthesis you’ll already know that sawtooth wave forms are best for synthesising strings.

Shaping the amplifier envelope is just as important. As you probably know string players (sometimes unkindly referred to as scrapers) produce a sound by striking a string with their bow and then draw it across the string in an up and down motion. The Attack, Decay, Sustain and Release controls of the synthesiser’s Volume Envelope should ideally reflect the way you want your virtual player to behave. For example – the harder a string is initially struck, the shorter the envelope’s attack time will need to be.
A typical string patch for normal bowing would have moderate Attack, Decay and Release times (knobs around 12 o’clock) and Sustain set at it’s highest level. As the synthesiser’s key is pressed there’s a slight crescendo at the beginning of the note due to the moderately slow attack time.

If you lengthen the Decay (3 o’clock) and reduce the Sustain level (9 o’clock) the note will drop in level after the initial Attack time. This too is a fairly common and natural sound.

If you were to set the Sustain level to it’s minimum level only the Attack and Decay will be heard – holding down the key has no effect. Moving the Attack setting to it’s shortest possible time and setting the Decay at 12 o’clock produces a plucked pizzicato sound.

If you have a copy of Cubase SX, you can easily re-create the classic sound of the Solina String Ensemble by using the A1 synth. It’s found as a pre-set in the Pad section. Two oscillators are used to create sawtooth waveforms which are treated heavily with a chorus effect, just like the original.

Tech Terms

Legato – connected notes, without gaps between them. Strictly speaking, all notes are played with one stroke of the bow, without changing the direction of the bow. In a string section all the players change their bowing direction at different times.

Pizzicato – The player plucks the strings with the fingers. The higher the note, the thinner the sound.
Spiccato – often confused with pizzicato but actually completely different. The bow is bounced on the strings to produce short, quickly decaying notes. Often used in fast passages.

Tremolo – rapid back-and-forth strokes of the bow are used to create a sense of urgency and anticipation.

Copyright 2006 Keith Gemmell

Swish sounds

It’s a scientific fact of life, but unfortunately, all electrical circuits generate a certain amount of noise. To prove it, with no music playing and all the faders up, turn the master volume knob on your mixer to the full on position. You’ll probably hear a swishing noise, similar to the sound of wind blowing through trees.

So what happens when you feed too low an audio signal into the average electrical circuit? The answer, of course, is that a high proportion of hiss will be added along with the audio. Because the audio signal is low, in all probability, when you arrive at the mixing stage you’ll want to turn it up. However, in this case, turning up the audio means turning up the hiss, because they’re inseparable. What you have is a low signal to noise ratio.

On the other hand if you overload a circuit with too high an audio signal you’ll cause it to distort. Obviously we need to find a happy medium here and the trick is to feed as much of the audio signal into the circuit as possible without causing distortion. That way you’ll achieve a high signal to noise ratio. Having said that, it’s wise to leave a safety margin (headroom), in case there are any unexpected high signals. And that’s why most items of equipment have a level meter of some kind with a normal operating level marked 0dB. The actual headroom available is chosen by the manufacturer and varies according to the make and model of the equipment.

You may have come across the term ‘dynamic range’ on your equipment spec sheets and wondered what it means. It’s just a measurement of the difference between the two problems mentioned above. For example, the dynamic range of an audio device is the difference between the lowest signal level it can handle without too much hiss and the highest signal level it can take without the risk of distortion. This measurement is usually expressed in decibels.

To dB or not to dB

But what exactly are decibels? Originating from the early days of the telephone they’re a measurement of ratio used to measure signal levels. Voltage is expressed as dBs and worked out as log (to the base 10) of V1/V2 where V1 (measured voltage) and V2 (reference voltage) are the two signal levels being compared. However, if you’re not too hot at maths, don’t worry because remembering a few figures is all you need do, to make intelligent guesses when operating faders and reading meters. For a start, if you double a signal its level increases by 6dB. If you double it again it increases by another 6dB, making a total of 12dB. That’s four times larger than before. The principle is the same in reverse; if you half a signal, you’re lowering its level by -6dB.

Audio production generally includes several pieces of equipment and it’s important to set up a gain structure and match the signal levels at every link in the chain. Most equipment has an input level indicator, either the old fashioned VU meter (for measuring volume units – below)

or a LED peak level indicator. How you set the levels depends on whether a device is digital or analogue. Analogue equipment levels such as signal processors and mixing desks are set to approximately 0dB. However, digital equipment levels, such as soundcard inputs (below), should be set a little lower, -6dB to -10dB or so, depending on the type of signal.

This is because in the digital world, if your signal goes above 0dB it will distort immediately (clipping). If the peak is very short, you’ll probably get away with it but if it’s continuously above 0dB, you won’t.

INFO: Audio signals sent to analogue devices can exceed 0dB by small amounts but those sent to digital devices must remain below 0dB, to avoid clipping.

Impedance matters

There are few subjects as complicated or baffling as impedance (Z) and what follows is a gross simplification of the subject. However, understanding the concept rather than the technical details is enough to ensure that all your gear is properly matched.

Take a look at the spec sheets for your studio gear – recorders, mixers, effects units and so on – and you’ll probably find their output and input impedance specified as ohms. However, you’ll not find an input impedance specification for your microphones. That’s because the input of acoustic signals is a mechanical process as opposed to an electrical one.

When you connect two devices together, the one outputting the signal is the ‘source’ and the one receiving or inputting the signal is the ‘load’. Leaving aside the technical stuff, roughly speaking, the impedance of the source device is a measurement of the electrical current (the audio signal) it can supply. The more current it generates, the lower its output impedance. The impedance of the load device is a measurement of the current it needs to operate properly. An example of this might be an effects unit (source) feeding a signal to a mixer (load).

The right gear

One way to better understand impedance is to use the analogy of a car climbing a hill. Think of the engine as the source and the car as the load. If the car is in a low gear, the engine provides higher revs and easily propels the car up the hill. However if the car is in a very high gear, the engine will struggle with the load.
Impedance matching is similar in principle. If the output impedance of the source is lower than the input impedance of the load (car climbing hill in low gear), there’s unlikely to be a problem. However if the load input impedance is lower than the source output impedance (car climbing hill in high gear), the circuit will come under strain. Just as the engine is going to struggle and lose power, the resulting audio signal will suffer from distortion and level loss.

Of course, just like the car’s engine, an electrical circuit doesn’t want to be performing flat out all the time and a kind of gearing ratio (bridging impedance) is introduced. The general rule is that the load impedance should be at least ten times higher than the source impedance. Most modern devices are connected this way – low Z source to high Z load – because bridging results in the maximum possible transfer of voltage from one device to another. Follow this simple rule when connecting your studio gear together and not only will you prevent signal overload, you’ll probably have enough ohms in hand to split a source signal between two or more load devices.

Power amps and speakers are an exception to the ten to one rule. Their impedances should be as near identical as possible, to maximise the transfer of power. Speakers usually have impedances of 4 or 8 ohms. Generally speaking (no pun intended), a lower speaker impedance will ensure more power from the amp. However some amps will overheat with a 2 ohm load, so to be safe, it’s best to keep the speaker load at 4 ohms or above.

Match boxes

What about equipment with unmatched impedance? How can they be connected? Well, we’ve established that impedance is similar in principle to the gears on a car so obviously you’ll need an electrical ‘gearbox’ of some kind. For guitars and synths you’ll need either a DI box or an impedance matching adapter. For example, you might need to connect a bass guitar (high Z) to an XLR mic input (low Z). Although it’s possible to use an impedance matching adapter such as a HOSA MIT-129 (beow),

an active direct box such as the BSS AR133 (below) is a better solution.

For example, you might have a mixer with 1/4 inch phone-jack inputs (high Z) and a professional condenser mic (low Z). The solution here is to use a matching transformer such as the HOSA MIT-176 between the mic cable and the input jack (below).

Incidentally, HOSA also manufacture a very handy transformer (MT-156) with an XLR (low Z) to 1/8 inch mini-jack connection (high Z) that enables you to connect a professional microphone to the input of personal portable recording devices, camcorders and computers.

And what’s considered high and low impedance for mics? 150-300 ohms is low, 600-2000 is medium and 10,000 or more is high. If you can, always use low Z mics. One reason being low Z mics can be used with a very long leads, over long distances, without hum or high frequency loss.

Tech terms

Z (Impedance) – An electrical circuit’s resistance to alternating current. High Z circuits contain high voltage and low current. Low Z circuits contain lower voltage and high current.

Transformer – Electrical component used to alter the voltage, current or impedance of an electronic circuit magnetically, instead of a direct connection.

SNR (Signal to Noise Ratio) – The ratio, expressed as decibels, between the level of the required signal and the level of unwanted noise.

Dynamic Range – The range, expressed as decibels, between the maximum output of a device (before distortion) and its noise floor.

dB (Decibel) – Unit for measuring audio levels.

Copyright 2006 Keith Gemmell

Ask a variety of music professionals to name their favourite sequencer and you’re likely to receive completely different answers. Musicians might answer Cubase, Live, Reason, Logic or Sonar, because those products provide an all-in-one composition, recording and production solution. A music teacher might point you to Band-in-Box, because it generates automatic play-along accompaniment for their students. Recording engineers and producers are likely tell you that Pro Tools is the ‘guvnor’ because of its high-end audio editing capabilities. The word ‘sequencer’, depending on the musical need, can mean very different things to different people.

Wall of sound

To Raymond Scott, an American film composer, who’s generally acknowledged to have invented the first sequencer as far back as 1953, a sequencer was the giant electromechanical device residing in his laboratory. He referred to it as his ‘Wall of Sound’. According to his friend, Bob Moog, the inventor of the Moog synthesiser, it was comprised of countless racks stacked high with relays, motors, steppers, and electronic circuits.

However, despite a great deal of interest and a visit from the bosses of Motown Records, the Wall of Sound never saw the light of day. You see, Scott wasn’t really interested in making money from his invention; he was already a wealthy man. Despite the fact that he was a prolific composer himself, his real goal in life was to perfect an automatic composition machine.

The first commercial sequencers were analogue devices consisting of a series of modules which could be set to particular voltages. When triggered sequentially, these voltages were interpreted by an external synthesiser as a sequence of notes. Such repetitive action was clearly more suited to digital technology and before long these unwieldy analogue machines were replaced by computers. They’re now rare as rocking horse pooh but for the retro minded, Algorithmic Arts have developed software emulations of these big old beasts – SoftStep and BankStep – which output MIDI data instead of control voltages (CV). If you’re a Windows user, you can download demos from their website.

BankStep, from Algorithmic Arts, functionally emulates a large $50,000 bank of hardware sequencers and supporting modules – without the dangling patch cords

Steps ahead

For electro composers in the 70s computerised step sequencers were an absolute boon and ideally suited to the robotic style of music associated with groups like Kraftwork. Composers simply typed a sequence, step-by-step, into machines like the MC-4. The 16 notes of the chromatic scale would be represented by the numbers 12 to 27 and a bar of 4/4 was allocated a number of steps. For example, if a bar had been defined as 192 steps, a half-note (minim) would be entered with a value of 96.

The software equivalent of step-entry survives today and can be found in both Logic and Cubase. To try it out in Cubase SX, open a MIDI editor and press the step-entry icon (staircase icon). Now specify note spacing and length, using the Quantize and Length menus and enter the notes with your MIDI keyboard. In Logic, you can use the Keyboard Window. Cubase’s Drum Editor also provides step entry as does Logic’s Hyper Editor.

Cubase provides a staircase icon for activating Step Input mode

Logic provides a keyboard window for step inputting notes

The Oxford Dictionary defines a sequence as ‘the following of one thing after another; a set of things belonging next to each other in a particular order’. To a songwriter, that ‘set of things’ could be interpreted as an intro, a verse, a chorus and a bridge. Until the 80s, pop music was composed using acoustic instruments and song construction (because that’s what composing is, beyond the initial bursts of inspiration) was a mental process. But the Atari computer and pattern based software such as Steinberg’s Pro 24 and C-Lab’s Notator changed all that.

Notator’s arrange page allowed pattern based sequencing, ideal for constructing and ordering the verse chorus and breaks in a pop song

Crazy patterns

Using pattern based sequencers musicians could now compose complete songs and easily restructure the material sequentially. For example, the centrepiece of the arrange page in C-Lab’s Notator was displayed as a single pattern. Within each pattern were 16 tracks. Musicians would simply record verses and choruses into patterns and assemble them in what ever order they liked in the ‘pattern list’. The finished songs, although conventional in structure, often had a robotic feel to them. In the mid 80s this method of writing gave birth to a whole new style of music known as Synth Pop; think Depeche Mode and Pet Shop Boys.

Steinberg abandoned pattern based sequencing when Pro 24 was reborn as Cubase. Notator, which eventually became Logic, soon followed suit. If you’re a regular reader of MTM and follow the tutorials for these applications you’ll be familiar with their linear approach to sequencing. However it’s interesting to note that Cubase has partly returned to its roots and now offers a more pattern-orientated way of working with the inclusion of the new ‘Arrange’ track. Once again musicians can move verses, choruses and other relevant sections of their songs about . Great stuff. Why did they wait so long?

The new ‘Arrange’ track in Cubase SX3 marks a return to the pattern-based approach to sequencing last seen in its ancestor, Pro24

Maybe Steinberg are responding to the success of FL Studio (yes, Fruity Loops has grown up). FL Studio is a pattern based and step-entry sequencer combined. Patterns can be welded together using a playlist. And just like the ‘Arrange tracks’ in Cubase, FL’s playlist also supports audio tracks (audio is streamed from the computer’s hard drive).

All aboard

With the success of Notator and Cubase, Atari computers were popping up in commercial studios everywhere. However, not everybody was computer savvy and before long sequencers were being installed in hardware synthesisers like the M1 which featured an on-board 8-track sequencer capable of holding 10 songs and 100 patterns. So good was the M1 at that time that entire backing tracks were sequenced using just the onboard sounds.

Korg have continued the tradition to this day and their Triton range of synths all feature 16-track onboard sequencers, with cue lists, for arranging patterns and facilities for both realtime and step-entry recording. The Triton Karma, in particular, is rather special. It features Kay Algorithmic Realtime Music Architecture (KARMA), a MIDI data generating technology that takes input notes and controller movements and generates complex musical phrases and effects. For anybody who owns a Korg Karma, special editing software written by Peter Kay, the inventor of KARMA., is available for PC and Mac. And that brings us neatly to another type of sequencing software.

The generation game

Apart from just recording data and editing it, sequencers can be put to work in other, more creative ways. Mathematicians often make good musicians and vice versa. And ever since the Atari days, musical boffins have been producing fractal music composition software. FractMus 2000, a freeware download by the way, is one such program. It generates notes and melodies using twelve algorithms from, wait for it… number theory, chaotic dynamics, fractals and cellular automata.

A similar program called MusicWonk also generates algorithmic MIDI music and, if you’re the experimental type, provides you with an opportunity to create riffs based on non traditional patterns. It’ll generate tunes from such unlikely sources as DNA sequences, stock market trends, star maps and any other algorithmic data source you care to throw at it. MusicWonk’s big brother, ArtWonk not only produces MIDI music but graphic animation too, also based on algorithmic processes.

Of course sequencers such as these will not actually generate complete compositions (not yet anyway) but saving the results as MIDI files and importing them into notation software or a conventional sequencer, can provide an almost inexhaustible source of melodic material to work with.

Audio young dudes

In response to the growing popularity of loop based music production, a few years ago several ‘audio only’ sequencers appeared on the market. Ableton Live for example provided facilities for mixing and matching samples of any tempo and pitch. Several other popular audio sequencers such as Fruity Loops and Acid did much the same thing. These easy-to-use sequencers were, and still are, extremely popular and attracted a whole new generation to computer music making. However these programs appear to be growing up (perhaps along with their users) and now, depending upon which version you buy, include MIDI functionality, albeit rather limited.

Sequencing software like Fruity Loops may have made sequencing easy but it was Apple who brought it to the masses, with GarageBand. Free with all new Apple Macs, GarageBand is a versatile entry level sequencer suitable for audio recording and basic MIDI sequencing. However it’s as an audio loop production tool that it really excels and Apple provide thousands of loops, packaged as Jam Packs, to use with the software.

Cool Edit Pro, yet another loop based sequencer, was recently acquired by Adobe Systems, given a makeover and re-branded as Adobe Audition. With it’s ability to exchange files with Premier Pro and After Effects, Adobe are clearly aiming their new sequencer at video professionals who need to knock up a quick-and-easy soundtrack. But is that really such a good thing? It’s doubtful that many musicians and composers would think so. Except one perhaps, if he were still alive. Maybe, fifty years on, Raymond Scott’s dream of a sequencer that composes automatically is nearing its reality after all.

Tech Terms

Realtime recording – A process which can be conducted while the sequencer is actually running. For example, tempo changes recorded using a MIDI controller would be done in ‘realtime’ without stopping the program.

Step Input – A method of entering MIDI note data into a song, either with the mouse or using a MIDI keyboard, while the sequencer is stopped.

Resolution – A sequencer’s smallest unit of time (tick) that’s possible within a bar. For example, in Logic this is 1/3840. Some people refer to resolution as ‘pulses per quarter note’ (PPQ). In those terms Logic’s resolution is 960 ppq (3840 divided by 4).

Copyright 2006 Keith Gemmell

Thomas Edison completed the task, in 1877, when he recorded his voice onto the worlds first recording medium – a strip of paper, coated with paraffin. Later that year he built a phonograph and used a cylinder to record the sound, this time covered in tin foil. In 1887 he updated the whole thing and began using a solid wax cylinder.

The commercial possibilities of such a momentous discovery were enormous and it wasn’t long before Emile Berliner, a US German immigrant, invented the gramophone, in 1888. He used a wax coated disc, which, after the recording, he immersed in acid, to expose the grooves made by the stylus.

Magnetic wire and tape

Although the wax analogue recording process remained much same until around the time of the Second World War, record companies began experimenting with two new formats – magnetic wire and magnetic audio tape. Both formats were developed simultaneously and battled for supremacy. In fact quite a few commercial reel to reel wire recorders and playback devices were on sale in the US in the 1940s. But tape eventually won through.

In both cases the recording principle is the same. A length of magnetic tape (or wire) is sped past a recording head. An electrical signal, captured by a microphone, is fed to the recording head and recorded onto the tape as a magnetic pattern. The tape is then passed over a playback head which converts the changes in the magnetic field back into electrical signals (sound).

Wire recorders like the Webster – Chicago 288 were common in the US, in the 1940s (photo: Tom Albrecht at antiqueradios.com).

By the 1950s professional recordings were made using 1/4” wide tape running at speeds of 15 and 30 ips (inches per second). And because analogue tape can carry multiple tracks in parallel, in perfect synchronisation, it wasn’t long before 2 track stereo superseded mono recording. The natural progression from here was 4 track recording and by the late 1960s, professional recording studios began using 8 and 16 track recorders. The tape size used for these machines increased accordingly, to a width of 2”. In the 1970s, 24 track recording eventually became the standard and recorders such as the Studer A80 and the Otari MTR90 remain a common sight in top flight studios, even today.

The Studer A80, a classic analogue 24 track tape recorder, still used in many recording studios around the world

Why did analogue tape become so popular? Because different instruments could be assigned a track each and mixed down to 2 track stereo. Also, it could be easily spliced and edited. And its disadvantages? Tape hiss, caused by the magnetic particles, was the main problem, particularly at speeds of 15 ips and below. However, in the 70s, various companies tackled the problem using volume compression and expansion. The most successful system was developed by Dolby laboratories. Other flaws included distortion and wow and flutter (pitch variation) but, like hiss, these problems were practically inaudible on high-end machines.

Open reel recorders

Although it’s all digital now, the home recording boom began with the advent of analogue tape machines like the Teac 3340, which in the 70s, offered musicians affordable 4 track open reel recording on 1/4” tape. 8 tracks on 1/2” tape soon followed, with the Tascam 38, but it was Fostex who first squeezed 8 tracks onto 1/4” tape with their revolutionary A8. The race to produce affordable 16 track machines using 1/2” tape began, and soon after, the Fostex B16 and the Tascam MSR16 became a common sight in budget studios around the world. It’s ironic that just as analogue tape recording had reached a state of near perfection it was almost completely abandoned, with the sudden rise of digital technology.

Compact cassette recorders

Also popular in the early days of home recording, the portastudio provided a cheap way of making demos on a 1/8” cassette tape. Tascam developed the concept first and perhaps one of their best machines was the 488 which, remarkably, allowed 8 tracks to be recorded onto a standard cassette tape.

DAT recorders

Whereas analogue recording captures sound as magnetic particles, digital recording converts audio signals into binary code. Pioneered by Sony, DAT (Digital Audio Tape) first appeared in 1987. The technology employed is similar to that used in video recorders. A rotating head is used to record the audio data on to 1/8” magnetic tape, housed in a shell approximately half the size of an ordinary compact audio cassette.

Most DAT machines use four sampling modes: 32 kHz at 12 bits, and 32 kHz, 44.1 kHz or 48 kHz at 16 bits. Some though can record at 96 kHz and 24 bits. The quality of the sampling depends on the duration of the recording. For example, 32 kHz at 12 bits provides six hours of recording while 96 kHz at 24 bits will provide just 90 minutes on the same length of tape.

DAT tape was commonly used for mastering in pro and semi-pro studios alike during the 80s and 90s. However, the advent of CD-R has led to a gradual decline in sales, although the portable variety remains a firm favourite for location recording.

MDM

Modular Digital Multitracks were directly responsible for the demise of personal open reel multitracks. They are easy to use, require little maintenance and can record 8 tracks of audio onto a cheap video cassette tape. Like DAT recorders, they use a rotating record head. From the early 90s onwards, Alesis and Tascam dominated the market with the ADAT-XT and DA-88 machines respectively. The ADAT machines recorded on S-VHS tape and the DA-88 and its descendants use Hi-8mm tape. They can still be found in many studios today and Tascam’s current flagship model is the DA-98HR. These machines are great for combining old world tape recording with digital editing and sessions are often recorded first on an MDM, transferred to a computer hard disk , for editing, and sometimes returned to the MDM for mixdown.

At the head of the TASCAM DTRS range, the DA-98HR is perhaps the ultimate modular multitrack recorder

Random access recording

Video cassette, of course, is not the only digital recording medium. Much more common these days are the various random access multitracks which record to hard disk and magneto-optical disk. Unlike the tape based digital formats, random access recording allows you to edit tracks. You can also record virtual tracks – several takes perhaps, of a single instrument – and choose the best one afterwards.

DAW

Digital Audio Workstations are probably the most commonly used random access systems used for home recording today. All you need is a reasonably powerful computer, recording and editing software and a sound card. Audio is sent to the sound card’s analogue inputs where it’s converted into a digital signal and recorded on the hard drive. Alternatively, a USB or FireWire interface such as M-Audio’s Omnistudio or Audiophile can be used instead. The advantages of this system are virtually unlimited tracks and digital signal processing (depending on the processing power of your computer), mixing to stereo on the hard drive, graphical faders and waveform editing.

HD

Not everybody likes computers, and for some, a stand alone Hard-Disk recorder is a better option. In the 90s Akai produced the first affordable machines like the DR4 and the DR8. However, due to the fixed hard disk within these machines, backup and storage was slow. Today, hard disk recorders are much more sophisticated. The ADAT HD24, for example, has two hot-swappable media bays which provide convenient access to the recording drives, reducing data backup time to just minutes.

MO

Magneto Optical recorders provide similar features to the hard-disk type. A laser is used to magnetise the disk and read the recorded data. MO disks are slower than hard disks but are reliable and cheap to buy. Otari manufacture a mono machine (popular with the talking book industry) as well as 2 track stereo and 8 track units (both popular in the world of film/video post production and broadcasting).

MD

MiniDisc is another ‘laser-read’ magneto-optical format and two types are available. One for 2 track recording (used in regular consumer MiniDisc machines) and MD data discs, used in multitrack recorders. The main attraction is, of course, the size and portability. However, to fit the data onto such a small disc, the MD format uses a form of data compression (ATRAC) and the sound quality is not as good as a DAT machine.

Tech terms

DAT – Digital Audio Tape, supplied in 1/8” cassette format. Commonly used for 2 track mastering.

MDR – Recordable Minidisc, often used by local radio, to record and play back programs.

DA8 – Digital audio cassette for Tascam’s DTRS modular recorders such as the DA-98HR.

ADAT – Digital audio cassette for Alesis modular recorders.

AAC – Advanced Audio Coding, an audio codec that may well replace the mp3 format. A favourite with the audio industry because it features built-in copyright protection.

MP3Pro – The latest advance in mp3 technology offers stereo compression at half the size of the standard mp3 format.

More information

Recording History – Lot’s information plus a film clip showing a demonstration of the Ediphone, circa 1907.

Copyright 2006 Keith Gemmell

Keyboard controllers

Keyboard controllers are modelled on a standard piano keyboard. When you press down a key a Note On/Note Off message is transmitted to a receiving device, a sampler maybe, telling it exactly which note to sound. At almost the same instant a velocity message is transmitted, relaying just how hard the key was struck.

Compared to a real piano, most keyboard controllers have small keys and provide a playing range of just a few octaves. For this reason professional players favour larger, fully weighted keyboards like the Kurzweil K2600 (below), to capture the full range and nuances of their performance.

To give musicians full control of their sequencing software, most professional keyboard controllers have a set of MIDI control sliders which can be assigned to any of the 127 CCs (Continuous Controllers) contained in the MIDI specification. For example, a slider could be assigned to CC#7, to control volume changes. Another might be assigned to CC#74 (Frequency Cut-off), to add a touch of brightness to the sound. In addition to a set of eight sliders, which can also be used as organ drawbars, the Kurzweil K2600 features a touch sensitive ribbon controller. This rather ingenious device enables the player to transmit Continuous Controller messages by simply moving a finger along the ribbon.

Of course you don’t have to be Rick Wakeman to use a keyboard controller and there are plenty of cheaper alternatives to the Kurzweil range on the market. Many musicians use small desktop controller keyboards for real-time control of their music software and hardware. M-Audio’s USB Oxygen8 (below) is perhaps the best known example.

However, for advanced MIDI control and a larger keyboard the Evolution MK range is a better bet with plenty of assignable knobs and sliders. Edirol also manufacture a range of compact MIDI controllers and their brand new PCR-1 model even features a built-in audio interface.

Guitar controllers

There are two recognised methods of transmitting MIDI data with a guitar. One is to use a specially manufactured MIDI instrument like the Ztar, manufactured by Starrlabs. On a conventional guitar you select notes with your left hand by pressing the strings on the fingerboard, against the frets. On the Ztar, you select notes using pressure sensitive keys instead. And with your right hand, you strike a string-like trigger on the body of the instrument. However, despite the lightning fast response and excellent control, the instruments are expensive and the player may need to learn a few new playing techniques.

Another much cheaper method uses pitch tracking. A special hexophonic pickup is mounted on the guitar which independently detects the pitch of all six strings and converts it into MIDI data. Blue Chip Music Technology manufacture these pickups for both guitar and bass along with the Axon AX-100, a rack-mountable guitar to MIDI convertor that’s capable of recognising the exact pitch of a note immediately it’s played. Being able to hear the conventional guitar along with the MIDI controlled sounds is a big advantage with this system.

Percussion controllers

Most early drum synths consisted of analogue drum pads with their own built in sounds. But the pads were prone to wear, the sensors were none too sensitive and many of the sounds were awful. In fact they were pretty crude devices compared to the percussion pads available now. Modern examples like the Roland SPD-20 have high quality drum pads capable of transmitting MIDI data from drum rolls and flams.

Roland also lead the way at the higher end of the market, with their V-Drums – full size kits featuring mesh pads, realistic cymbals and a drum sound module. The pads have a wide dynamic range and even include rim shot triggering. Roland also designed and now manufacture the first ever electronic hi-hat which operates exactly like its acoustic counterpart.

Tuned percussion controllers are an obvious choice for transmitting MIDI data. Kat Electronic Percussion manufacture the MalletKAT, an interesting device for the mallet players (musicians skilled at playing the vibes and glockenspiel). Like most high end percussion controllers it’s available with or without sounds, is a three-octave design (expandable to five) and features keyboard splitting and layering.

But the MalletKAT is more than just a musical performance controller and highlights not only the multitasking capabilities of modern MIDI controllers, but also, the general direction in which the major manufacturers appear to be heading. With software like Reason or Live DJs can also use MalletKAT, to access controller number events and trigger loops, tweak filters and control parameters like pitch bend. And because it functions as a real time controller (the black keys define which knob, and the white keys send out values), you can perform all these tasks while continuing to play the instrument in the normal way.

Wind controllers

Wind players, of course, blow their instruments and control the sound using the face muscles around their mouths. Wind controllers enable them to control a synthesiser or sound module in exactly the same way, by using sensors to convert their breath and lip pressure into MIDI data.

Like the early drum pad synths, the first ones made were not MIDI instruments at all but analogue controlled synthesisers. The best known and most successful model to appear in the seventies was the Lyricon. Based on the playing principles of the saxophone it was supplied in a case containing both the synthesiser and the controller.
When production of the Lyricon ceased Akai took over and reinvented it as the EWI (Electronic Wind Instrument) and the EVI (Electronic Valve Instrument). The EVI appears to have been dropped (bad news for experimental brass players) but the EWI lives on today and is still considered by many to be the most expressive way of controlling a synth. Why? Because unlike a digital controller, where the data is step-quantised in discrete steps, the EWI uses control voltages. Data transmission is completely smooth resulting in a more expressive and subtle performance.

Yamaha however, do manufacture a true MIDI wind controller. The WX5 (below) has two finger modes – saxophone or flute – and a choice of single reed (for sax or clarinet players) and recorder type mouthpieces. It’s not as subtle as the EWI but at least you can connect it directly to MIDI tone generators. It’s also very user friendly and suitable for pros and beginners alike.

Foot pedals

MIDI controllers are not restricted to transmitting just performance data. MIDI foot controllers for example are a godsend to guitarists with racks of gear to control. After all you can’t just stop in the middle of a blistering guitar solo to fiddle around selecting presets. Using MIDI foot pedal controllers instead makes stage performance much easier because you can store program changes and controller data in banks of presets ready to send at a moments notice. Expression pedals too can be used to control MIDI channels, control numbers and so on. Even MIDI note numbers can transmitted by foot switches on models like the Behringer FCB1010. This particular model even has two programmable, relay-controlled foot switch jacks for switching guitar amp channels via MIDI.

Control Consoles

Music software packages are wonderful things. But for many, fine tuning faders and knobs with a mouse is a tedious operation. The problem is easily solved by using a MIDI Control Console with real knobs and automated faders.
The two big names in the sequencing world, Steinberg and Emagic, both made their own proprietary hardware controllers for a while (Houston and Logic Control). However, these were discontinued leaving companies like JL Cooper and Mackie to get on with the job. The Mackie Universal is set up for Logic by default but also enables you to control Pro Tools, Cubase, Digital Performer, Sonar and other major audio software.

JL Cooper manufacture a similar product, the CS-102 (below) and an intriguing miniature control surface, the CS-32 Minidesk. It actually fits in the palm of your hand (great for mobile laptop recording) and can be used to control most popular audio software.

Light controllers

A Ten Minute Master on MIDI controllers wouldn’t be complete without mentioning the usefulness of MIDI for controlling stage lighting. For large productions console controllers can be used ‘front of house’ but for smaller gigs a member of the band can do it using much cheaper devices like the Ryger M2LPRO. It converts MIDI messages transmitted from a keyboard or sequencer into eight controlled mains outputs. Using a keyboard, a musician can assign an unused octave to change lighting scenes and control dimming and crossfade effects.

Tech Terms

Continuous Controller – channel specific messages (also known as MIDI Controllers and Control Changes) used to control synth parameters other than the notes themselves. Some commonly used controllers are Modulation (01), Breath Control (02), Volume (07), Pan (10), Expression (11) and Sustain Pedal (64).

Note On - messages used to convey which note is pressed on a musical keyboard along with velocity information (how hard the key was struck).

Note Off – messages used to convey when a note on a musical keyboard is released.

Program Change – message used to specify which sounds are played on a given Channel.

More information

The MMA (Midi Manufacturers Association) site is an absolute must for anybody interested in the latest MIDI technical developments.

Copyright 2006 Keith Gemmell

Analogue recording dominated the music industry for the best part of fifty years and just as it reached a state of near perfection, ironically, it suddenly disappeared? Well not completely, because many musicians and engineers still use it, but as far as the manufacture of professional multitrack tape recorders goes it’s as good as dead. However, pro reel-to-reel tape recorders continue to fetch high prices on the second-hand market. For example, a Sony APR24 will set you back around £5,500; a Studer A827, nearer £16000. So what’s the attraction?

The superb Studer 827 Gold Edition is last of the great 24-track 2-inch analogue tape recorders; seen here complete with autolocate and meter-bridge.

Head to head

On a pro reel-to-reel machine tape passes over three heads: erase, record and play (in that order) and the waveform is stored as a magnetic field. This is a continuous process and many top engineers claim smoother results and better highs compared to digital recorders which sample audio as slices, a non-continuous process (although as sample rates continue to improve, the argument begins to pale). Dyed-in-the-wool rock producers also claim better reproduction of the complex waveforms typically associated with crash cymbals, distorted guitars and instruments which produce multiple harmonics like 12-string guitars and acoustic pianos.

But curiously, it’s the limitations of magnetic tape that contribute to the mystique of analogue recording and enhance its desirability. For example, recording high signals (meters in the red) results in ‘tape saturation’, a mild distortion, sometimes used creatively, to thicken the sound of drums and guitars. ‘Tape compression’, a result of magnetic tape’s limited dynamic range is also considered desirable on vocals and drums.

Other side effects are less desirable; tape hiss (caused by non-aligned magnetic particles) being the worst offender. The lower your recording levels, the louder the noise. Wow and flutter (slow and fast tape speed variation) can also cause problems. However, analogue recorders underwent continuous refinement during their life span and in the 70s and 80s these problems were all but eradicated with improved tape transport mechanisms and high tape speeds.

As you probably know, audio is recorded onto tape as tracks. And wider tracks yield better signal to noise ratios than narrow ones. So, the wider the tape the better your results are likely to be. High-end multitrack recorders use 2-inch tape to record either 8,16, 24, 32 or 48 tracks although 24 is the most common (some studios use two 24-track machines together, one as a slave). These machines run tape at speeds of 7.5, 15 or 30 ips (inches per second). The faster the speed, the better the sound quality with less hiss and increased pitch stability. Other common tape widths are 1-inch, for 8, 16 or 24 tracks; 1/2-inch for 4, 8 or 16 tracks and 1/4-inch for 2-track stereo and 4 and 8-track recording.

Hiss off

Semipro machines like the Tascam MSR 16 (16 tracks on 1/2-inch tape) run at a maximum speed of 15 ips, half that of a pro tape deck, and as a result, add tape hiss to each track recorded. An accumulative process it’s more noticeable at mixdown with each track contributing to the overall noise. To solve the problem manufacturers added Dolby and dbx noise reduction units to their machines. Audio is compressed (encoded) during the recording process and expanded (decoded) during playback. Put another way, the compressor boosts the quiet passages (music) during recording and the expander reduces the volume of noise (tape hiss) when the quiet passages (music) are played back. During loud passages of music nothing happens because the hiss can’t be heard anyway. Both Dolby and dbx systems were very effective and it’s a matter of personal taste as to which sounds best. Pro studios though usually record without noise reduction at a speed of 30 ips.

The Tascam MSR-16 offered affordable 16-track recording on 1/2-inch tape during the late 80s and early 90s

When setting up an analogue recording system it’s common practice to match the recording console’s meters with those on the tape recorder. There’s a simple reason for this. Once done, you only have to watch one meter, usually the console’s. You do this using a steady 1kHz tone provided by a frequency generator.

Analogue tape recorders are sophisticated pieces of machinery, requiring periodic maintenance and regular cleaning. A layer of tape oxide quickly builds up on the heads and tape guides which, if not removed, will interfere with the tonal quality of the recorded sound. Cleaning them before every session is vital. Fortunately it’s an easy job requiring just cotton buds and a cleaning solvent.

Every few days or weeks, depending on usage, the tape path has to be demagnetised. A magnetic field builds up over time and causes all kinds of nasty problems including high frequency loss, and extra tape hiss. Using a degausser each tape head and guide has to be demagnetised one at time. Although not a difficult task demagnetising is a rigourous process and has to be performed to-the-letter to avoid potentially disastrous results and costly repairs.

Get in line

Every so often tape machines need aligning. Again the time interval depends on the use. It’s a skilled job involving manual adjustment of the heads, to line them up in a position exactly relevant to the tape path and calibration of the electronic circuitry. In a busy commercial studio this is carried out by a specialist maintenance engineer armed with an expensive test tape. Specific frequencies are recorded to these tapes which are used to precisely set up the machine. In the heyday of analogue recording a busy studio would do this before every session.

Analogue tape recording is something of an art. Conventions and procedures abound and there are right and wrong ways of doing everything. For example, it’s traditional to wind tape off the machine at playback speed to avoid unevenly spooled tape which is what happens if you use fast-wind mode. Apart from looking untidy, unevenly spooled tape is easily damaged at the edges and remember, in a commercial studio the clients are looking on. Tapes are also stored tail out (on the take-up reel) to help minimise print-through, the phenomena whereby you hear a faint pre-echo just prior to the start of a recording.

Work on a song invariably takes place in different studios and it’s common practice to splice on a couple of meters of green leader tape before the recording, to reduce static. A red trailer tape is spliced on at the end.

Splice and dice

In the old days editing was a tedious process involving demagnetised razor blades (single-edged of course), splicing tape and editing blocks but skilled engineers were remarkably good at it. Nowadays analogue recorded works are more likely to be transferred to ProTools for editing.

Visit the world’s top studios and as well as the ProTools rig you’ll find a couple of Studer 24-tracks tucked away in the machine room. These dinosaurs are kept alive because many of their clients demand them, just as they also expect classic Neumann mics and so on. But are the recordings made on these big old beasts any better than those made using state-of-the-art digital equipment? The arguments for and against will rage for some time yet but maybe the answer is a matter of choosing the recording format to suit the style of music. After all, many different styles of popular music now coexist which wasn’t always the case in previous eras.

Many top rock and acoustic musicians still opt for analogue recording and transfer the material to digital for editing and mastering whereas pop acts and producers of loop based music use the digital medium throughout. But as far as the consumer is concerned, it’s a digital world. Affordable analogue recorders for home use are not made anymore.

However, if you’re interested in recording the old fashioned way you can still buy second hand machines. At the time of writing a Fostex E16 (£310), Fostex R8 (£150) and a Tascam MSR 16 (£550) could all be found on e-Bay. Be careful though, overworked multitracks from commercial studios are best avoided. But lovingly-cared-for machines from home studios do turn up and can be easily integrated into your digital system using SMPTE time code. And compared to many digital machines they’re easy to use and recording levels are not so critical.

Running costs are a major consideration. For semi-pro recording, a reel of 1/4-inch Quantegy 456 (formerly Ampex) costs around £20. 1/2-inch and 1-inch tape 456 tape costs £35 and £55 respectively. 2-inch tape is very expensive and a reel of 456 costs around £116. The best, Quantegy GP9, costs £150; rock star prices.

To mark the end of an era, sadly, Studer have decided to cease production of analogue tape machines but if you happen to have £23000 to spare you can still buy one of their A827 Gold Edition models (while stocks last) complete with autolocate and meter-bridge. The future of recording is digital, without question, but discerning musicians, producers and engineers will also be recording on machines like these for many years to come.

Tech Terms

Gap – On a tape recorder’s record-head, a thin break in the electromagnet that contacts the tape.

Azimuth – On a tape recorder this is the angular distance measured between the head gap and the tape path. It should be 90 degrees. But it alters over time and the mechanical adjustment of the record and playback heads is known as Azimuth Alignment.

Splice – As a verb, the joining together of two lengths of tape with a third blank piece of leader tape. As a noun, the the taped joint itself.

Splice block (or editing block) – metal block for securely holding tape whilst you’re editing it with a razor blade.

More Information

Showcase International’s list of the world’s top recording studios. Follow the links and you might be surprised at just how many of them use 24-track analogue tape recorders.

Copyright 2006 Keith Gemmell

91. Software monitoring

If you record using a DAW such as Logic or Cubase you have three monitoring options. Which one you choose depends on your hardware and your preferred method of working.

1. Monitoring via the software – the input signal is mixed with the audio playback. Pros: you can hear the music back with panning, effects and EQ. Cons: Latency – the signal you hear back will be slightly delayed.
2. Direct Monitoring (ASIO 2 compatible hardware needed) – the audio hardware handles the monitoring and sends the input signal directly out again. Pros: Latency free. Cons: You’ll not hear any EQ or effects because the signal doesn’t actually pass through the software. The software just ‘controls’ the monitoring.
3. External mixer – the most straightforward way. Send a signal to your main studio speakers and an additional feed from the mixer’s auxiliary sends to your sound card. If the mixer has direct outputs use those instead.

92. Short and thick

If possible, place your power amp close to the speakers. Short cables with thick conductors are more efficient than long thin ones.

93. Form a triangle

You and your monitor speakers should form a triangle – place them the same distance from each other as they are from you.

94. Label everything

You’ll need to identify channels quickly on your mixer. Use masking tape along the bottom of the faders and label the channels. Your mixer may have a plastic strip already. If so, use a chinagraph pencil – they’ll write on anything.

95. Keep track

Keep a track sheet and a log if you’re recording with a hardware recorder. It might not seem important at the time but you’ll need it later, for the mixing session.

96. Bouncing tracks

With a little ingenuity, you can make cheap, good quality demos using a 4 track cassette portastudio. By mixing two or three tracks together on to an empty track and erasing them afterwards, you can record up to nine tracks.

97. What’s that hum

The last thing you need is a persistent hum on your recordings. Check that your signal and mains cables run parallel with each other. If they must cross, make sure they do so at right angles.

98. Live four track

You can produce a reasonable live recording from the main output of your band’s PA mixer connected to two tracks of a four track portastudio. Use the other two for a stereo pair of mics, placed out front, for ambience.

99. Live digital recording

Plug your digital multitrack into your PA mixer’s inserts and use the input trims to set your levels. Do this at the sound check because the faders will need adjusting to compensate for any changes made to the live sound balance. Take it home, overdub, edit, do whatever you like and mix it.

100. Go it alone

Ignore everything above (except the 3 to one rule) and do your own thing. Remember, when recording the end result always justifies the means.

Copyright 2006 Keith Gemmell

81. De-mag, de-frag

A lot of musicians still use analogue recorders. To get the best results clean the heads and tape guides regularly, with isopropyl alcohol and cotton buds. Both can be found at your local chemist. Also, invest in a demagnetising tool and demagnetise the heads occasionally.

On the other hand, if you record to hard disk your software will find and playback your audio files much quicker if you defragment the hard drive regularly.

82. Get connected

Lying on the floor, groping blindly around the back panels of mixers, recorders, effects processors and sound modules every time you need to make a patch is hard work – buy a patchbay and connect everything to it. The Behringer Ultrapatch Pro PX2000 is very reasonably priced.

83. Which sample rate?

If you’re using a DAW, you’ll need to decide on a sample rate and stick with it throughout your project otherwise you’ll end up with very odd sounding recordings. So which do you choose?

Basically, the higher the rate the better the sound quality and what’s actually available depends on your audio hardware. For example, if you have a popular sound card from the M-Audio Delta range you’ll have a choice of anything between 8 kHz and 96 kHz. Most people work with 44.1 kHz which is the happy medium. If your project is destined for a CD, that’s how it will end up anyway. By all means use 96 kHz, if you have the capability, but remember it will use more disk space and processing power. There’s little point though in going lower than 44.1 kHz because audio quality will deteriorate noticeably from 32 kHz down.

84. Which record format?

Most sequencers support 16 bit, 24 bit and 32 bit recording. If your audio hardware only supports 16 bit there’s nothing to be gained by selecting 24 bit except larger files with the same audio quality. However, if your hardware supports it, choose a higher resolution. Bear in mind though that higher resolutions result in larger audio files, thereby putting more strain on your computer.

85. Your own input

Many musicians mistakenly assume that when recording with sequencers you set the input levels using the software mixer’s channel faders. Of course, if you think about it this can’t be so. The fader only affects the playback of the audio once it’s recorded. So when recording, remember, the level meters show the signal level at the ‘input’ selected for the audio channel. You can adjust this level in one of the following ways:

By adjusting the output level of the sound source or your external mixer.

By using your audio hardware’s own software mixer to set the input levels, if this is provided.

86. Correct mistakes early on

Finalise sounds, correct timing problems and sort out wrong notes as early as possible when recording a rhythm track. Tiny mistakes may seem insignificant to begin with but they almost always increase in magnitude as the session progresses.

87. First takes are best

The first couple of takes are almost always the best. Musicians go off the boil after three or four takes so make sure everything is right before you start recording.

88. Keep moving on

Keep moving when overdubbing vocals or instrumental parts. Stopping for every tiny mistake will break the performer’s flow and undermine their confidence. Record several versions and edit them together afterwards.

89. Punch In/Out

Playing an instrument and recording your performance at the same time can be an unwieldy, and potentially dangerous task to say the least (pressing record buttons on and off and so on). Fortunately the whole process can be automated in Cubase and Logic.

For example – you need to drop in and repair a guitar solo between bars 17 and 21. All you have to do is set the Locators to encompass those bars and activate the Punch In, Punch Out or Autodrop buttons on the Transport panel. Now scroll back to a point before the drop in – let’s say bar 13 – and activate playback. When the cursor reaches bar 17 the sequencer begins recording and when it reaches bar 21 it stops.

The same process can be set up on digital hardware recorders and analogue recorders like the Tascam MSR-16.

90. Well stacked

One of the best things about recording with a sequencer is the ability to cycle record. Cubase for example, allows you to stack multiple takes on a single audio track. With ‘Stacked’ selected and ’Punch In’ and ‘Cycle’ activated cycle recording takes place as normal. When you’ve finished recording, all the takes are stacked on the one audio track, each with a different number. Now you can chop them up into smaller events, choose the best of the bunch and quickly assemble a perfect take.

Copyright 2006 Keith Gemmell