For the musician, setting up a small home multitrack facility can be fraught with problems, most of which are financial.

It's easy if you have a bulging bank balance; you just rush out and buy piles of new, hi-tech equipment. For the rest of us, starting with just a few hundred pounds and needing to produce professional results, reading reviews of the latest £1000 digital reverb can easily produce feelings of inadequacy. We hope to show that there's a lot to be said for starting with very little gear, really learning how to get the best from it, and only gradually buying more/better items of equipment as you come across the need for them. We found that you can get good enough results to achieve some success with very limited resources, and these are some of the pitfalls and tips we discovered on the way.

First Mistakes

After years of running live bands, lugging masses of gear around the country to often ill-attended gigs and waiting hours for musicians to turn up at rehearsals, we thought we'd set up a 4-track facility at home, play all the instruments ourselves and be able to work on our music in comfort any time we like. Not trusting the ten fairly recently introduced cassette-based multitrack recorders (wrongly, as it happens) but having a budget of only £450, after weeks of studying the small ads we opted for a second hand Teac A3340S and a Teac Model 2 6:4 mixer.

Woefully ignorant about recording gear, we omitted then to look at the heads and guides; a mistake which eventually added another £200 to the cost. We were impressed by the demonstration the private seller gave and so, blissfully unaware of the impending demise of the heads, purchased the package.

Six months later, during a long, tortuous session the tape machine began to shred oxide from the edge of the master tape and what remained of the tracks suddenly sounded drastically lacking in top end. That's when we discovered how to check out the head wear on a second hand machine! New heads have an even curve where the tape passes over them, so what you're looking for is any flattening of the curve or any obvious damage. Harman UK's engineer tells that modern tape formulations can also sometimes cause a 'rippling' effect on the head face. It's easier to recognise a worn head if you know what a perfect one looks like, but viewed under good lighting a distortion of the reflection from the metal will usually show up any such flattening. Worn or badly aligned tape guides, though relatively inexpensive to replace, can cause heads to wear prematurely, often unevenly. Characteristically, worn heads exhibit poor high frequency response, so if all looks satisfactory, ask the seller to record some test tones. These should be from a signal generator or the test oscillators built into some desks, but failing this you may be able to produce sufficiently high frequency tones from a synthesiser. Make sure the tones are unmodulated and as pure or close to sine waves as possible. Record tones of about 8-10kHz (cassette machines) or 15-20kHz (reel-to-reel) onto all tracks with the meters showing 0VU. On playback any difference in levels shown by the meters should be within the manufacturer's specification for frequency response; usually about ±3dB over the range covered. If there is a droop in response at these higher frequencies, this may well indicate head wear. Make sure that the heads are clean, demagnetised and that the tape is the correct type for the machine, or you could get totally misleading results. Magnetised heads can cause a loss of top end, increased noise levels, crosstalk, distortion, and in extreme cases even pops and crackles. If you catch head wear before it becomes too serious, it's sometimes possible to have the heads re-faced, which restores the correct profile and can extend their life. This is far cheaper than replacement, though only a temporary cure.

"...there's a lot to be said for starting with very little gear, really learning how to get the best from it, and only gradually buying more/better items of equipment as you come across the need for them."

Mixing

With the machine and mixer connected to our very average hi-fi system for monitoring, we began recording. After initial amazement at the quality attainable on such equipment, the limitations of the very simple mixer soon became apparent. The only equalisation available consisted of fixed frequency switched high-pass and low-pass filters. Whilst these proved quite useful for removing unwanted bottom end and removing tape hiss on tracks containing little high frequency information to mask the noise, there was no means of boosting any frequencies to add top, for example. However, we found that cutting out a lot of the lower frequencies had the effect of allowing what high end information there was to dominate the mix, resulting in a subjectively brighter sound. To obtain more control, we plugged synths, guitars and microphones into our stage amplifiers, taking the line output from the amps into the mixer. This, of course, increased the noise being recorded onto tape, but allowed us to use the amplifier's simple bass and treble controls to EQ sounds. By employing pre-emphasis; boosting the treble control (centred around 10kHz) when recording sounds, then switching in the mixer's low-pass filter (also at 10kHz) on playback, thus reducing tape hiss, it was possible to obtain a brighter, quieter sound than before, cancelling most of the noise introduced by the amplifier.

Monitoring

Mainly due to the very inaccurate hi-fi speakers used for monitoring, it was all too easy to get the tonal balance completely wrong. Mixes that were apparently good on cassette sounded bass heavy and dull when played over other systems. Differences caused by the inevitable misalignment between cassette machines, the narrow track format making head azimuth critical, were part of the problem, but not much could be done about this. Much cross referencing on various systems and taking into account colouration introduced by the monitors, improved the situation.

"...it's sometimes possible to have the heads re-faced, which restores the correct profile and can extend their life."

The amp's built-in cheapo spring reverb proved quite acceptable on most sources, despite its limited bandwidth, going 'boing' only when subjected to percussive sounds. Heavy reverb treatments resulted in an unpleasant, metallic, unnatural sound with resonances produced by the short two spring assembly becoming obvious. A useful warning though, when recording the effect onto tape like this is that the final mix could end up being swamped in muddy reverb, which is classic bad demo stuff! For alternative effects we tried miking sounds in the bathroom, down the hall, inside cupboards, and so forth. It was impossible to obtain long, smoothly decaying sounds reverb because the rooms were just not large or live enough. Fortunately, however, many of the sounds currently fashionable due to the emergence of digital room simulators could be produced, ie. short, hard, sometimes very coloured ambient treatments.

The sound was much cleaner when track bouncing was avoided. Instead at least two instruments were recorded 'live' onto each track. Early attempts resulted in mixes where initial tracks disappeared under the overdubs, but gradually it became possible to judge how much to over-emphasise instruments, particularly snare drum and bass, so that they 'sat' correctly in the final mix. If we had to, only one previously recorded track was bounced while a 'live' first generation overdub was added. This required a certain amount of pre-planning as some sounds deteriorate more than others when bounced, and also the Teac A3340S does not allow bouncing onto adjacent tracks without horrendous feedback occurring unless you're very lucky. Using the correct tape for the machine, then really pushing levels onto tape (we found that constant, non-transient sounds such as synthesised strings, could be recorded with the meters permanently reading well above 0VU) helped minimise tape noise. The tape compression incurred even improved some sounds, particularly toms.

Talking of tape noise, we ended up monitoring at neighbour-upsetting levels or over good phones because, although when listening at quieter volumes the signal to noise ratio remained constant, the tape and system noise would drop below our hearing threshold, giving a false sense of security and allowing undetected gunge into the recordings. Similarly, cheap phones with poor bass response can fail to show up nasty mains hums and buzzes.

Experimenting

A very useful method of creating ADT and flutter echo effects using only the Teac was accidentally discovered during one session. Making sure you have two spare tracks, record a sound simultaneously onto them, with one track in the sel-sync mode and the other set to monitor off-tape. Then bring this second off-tape signal (being monitored from the playback head and therefore slightly delayed) back through the mixer, routing it onto the first track. The result is a mono double-tracked signal on the first track, leaving the second tack free for further overdubs. A variation using only one track, is to record the sound, monitor it off tape, routing the monitored signal back onto the same track. This produces a fast, repeating 'flutter' echo, rather like turning up the feedback control on a DDL. If you have the repeats at too high a level, though, the system will go into oscillation, but even this can sometimes be used as a special effect. Halving the tape speed will, of course, double the length of the delay between repeats.

"It seemed as though we had a DDL, digital drum machine and Fairlight for £200!"

Drum sounds had been produced up to then using a very cheap Boss DR110, improving the snare drum by using its trigger to fire a burst of filtered noise from a synth. We then added real toms and cymbals, miked using Tandy PZM (which incidentally are ridiculously good for the price, though it's worth using the higher voltage batteries for more output and less susceptibility to distortion at high SPLs). On pieces that really needed a full acoustic drum kit, we took the 4-track machine into a local commercial studio for a couple of hours, having previously recorded a guide track for the drummer. The drums could then be put down quite cheaply using a decent multi-microphone set up and overdubs completed at home.

Making the Most of...

An H&SR review of the Boss DE200 DDL prompted the next purchase. For about £200 it provided ADT, echo, chorus, and flanging effects at a reasonable bandwidth but, best of all, enabled us to sample drum sounds from records. Initially we only used it to replace one sound, usually snare, on the DR110 firing the sample from the drum machine trigger out. The improvement was so dramatic that we soon dispensed with the DR110 altogether, building up entire drum tracks from samples bounced together. This was diabolically difficult and very time consuming, physically 'playing' each individual drum sound from a button. It would have been far easier if the DR110 could have been synced to tape and used to trigger the sounds. However, it was well worth it as our demos now suddenly started to sound something like 'real' records! Samples such as orchestral 'parumphs' were also possible. It seemed as though we had a DDL, digital drum machine and Fairlight for £200! Pseudo-stereo (dry sound left, effect right), ADT and chorus effects at mixdown helped to give space to the music, creating a more interesting stereo picture with only 4-tracks to make up the image.

Recordings made on this relatively modest equipment resulted in us winning second prize in the H&SR/TDK jingle competition. Encouraged by this success, we recorded further jingles and themes, resulting in a commission to produce an album for a library music company and other offers of work to be taken up when this is finished. This proved to be only the start of our problems! Now we had to make the transition from demos to 'real' records...