Despite the widespread use of reel to reel recorders in professional recording studios, the compact cassette system has gained universal acceptance as the medium for home recording and replay of every kind.

Even so, many users are often ignorant of the basics of good sound recording on cassette and this in part is due to the lack of information on the subject. Compared to a reel to reel machine the cassette player has some inherent limitations. Most cassette decks have two tape heads, one for erase, and the other for record and replay. This makes it impossible to monitor the actual sounds on the tape when recording. The electronics also have to be something of a compromise when one head is used for both recording and replay. On less sophisticated decks there can be complex mechanical switches which are prone to become stiff and noisy in operation.

Further, the low tape speed, 1 7/8" per second, means that the high frequency response at high levels is unlikely to be as good as a reel to reel. On the plus side, cassettes are readily available at reasonable cost and very convenient to use. With careful recording techniques the medium is capable of producing excellent results.

Getting the best from your cassette tape is often more difficult than one would expect. For a start, most recorders use standard accuracy components in the vital equalising sections. This means that there may be some response variations even between different samples of identical recorders.

Another restriction imposed on the cassette from the start is the fixed speed. When Philips patented the compact cassette in 1963, they allowed other manufacturers to produce machines and tapes under licence providing that the speed and basic format were retained. Sony's Elcassette, which was launched in the late seventies, used 3¾" per second but has now fallen by the wayside despite its greater potential.

From Dolby to dbx

Obviously something had to be done about the abysmal signal to noise ratio. All tape recording efforts can be regarded as an uneasy compromise between losing the low level information in noise and the peaks in overload. With cassette tape this problem is more acute because the signal to noise ratio is around 50dB. In comparison a well recorded LP has a range of 60dB. The Dolby noise reduction system was developed and now widely adopted for improving the signal to noise ratio.

The Dolby system works by boosting high frequency response of low level signals on record and by cutting the response on replay. The net result is that the signal emerges with the same tonal balance as it was recorded, but the noise is reduced by the high frequency (HF) cut as well. Although the situation has improved lately, the most common fault with the use of this system is incorrect level matching between the record and replay sections. As a result, many recordings made with Dolby tend to sound less bright, with the HF severely cut compared to the original. Of course, not all of the blame for a cassette's sound can be placed at the door of Dolby since a lot of HF loss is due to incorrect recording levels.

The Dolby system has another limitation which can be annoying. If Dolby is turned off during replay to gain a better HF response, an audible pumping sound can often be heard. With the passing of time several other noise reduction systems have been developed, the most successful of which is dbx. Dbx operates in a completely different manner to Dolby, operating on all of the signal.

Essentially the audio signal is square rooted by the processor on record and squared on replay. As a result, a signal with a dynamic range of 60dB will be recorded with 30dB range and this will easily be accommodated on cassette tape. On replay the original dynamics will be retained. Since the device operates on the ratio between signal levels, no particular matching of input/output levels is required.

Tape coatings

To make things more difficult for the recordist, many different types of tape coatings have been devised. The market has now narrowed the range down to four. With great originality these are labelled Type 1 to Type 4!

Type 1 has a coating of Ferric Oxide, Type 2 of FerriChrome, Type 3 is Chrome, whilst Type 4 is pure Metal (Iron). The original Type 1 is the most common and cheapest but tends to be the most noisy. Type 2 is an improvement giving better HF and lower noise, but at an increased cost. Type 3 also gives better HF and noise figures.

Pure metal tape is capable of really impressive results providing the cassette deck is capable of using it and is presently the most expensive tape available.

Testing tapes

It is unlikely that anyone would seriously contemplate using Type 1 tape for making a demo, for example, but how do you choose the best value for money. The best way is to purchase a selection of the different brands of tape you intend using. To test the frequency response of the tapes an FM tuner is required. Set your equipment up as follows.

Remove the aerial from the FM tuner and tune to a part of the band where there are no stations. You should also switch off the muting switch if one is fitted, so all you should be receiving is interstation noise. This noise has a spectrum which completely covers the audio spectrum and beyond. Check the meters and set the record level of the cassette deck to -20dB with this signal, recording about a minute or so. Now record a further section of silence.

Repeat this procedure with all the tapes at your disposal. On playback select the tape which sounds most like the original noise source. You will find that all the tapes will sound subtly different. The best will have a top response which is flat. If you find one with a very toppy response it's worth experimenting with it further too, since such tapes can often sound good on music.

Taking the flat tapes, compare the silent sections. The quietest tapes here will be the best match for your machine. All these tests should be done with Dolby switched off and any ancillary set for a flat response. This basic test can be used with any tape machine, although some better decks will have variable bias and equalisation.

Variable bias can theoretically enable you to set your machine for optimum results with any tape, although it can be time consuming in practice. However, a lot of recordists may not understand how bias works.

The transfer characteristic of magnetic tape is not linear as can be seen in Figure 1. In order to record music with any fidelity at all, it is necessary to superimpose the incoming audio signal onto a supersonic signal. This supersonic carrier is called BIAS and the linearity or otherwise of a recording depends critically on the bias used. Too much bias and the treble response goes down. Too little and the treble is unnaturally boosted. The perfect recording is obtained somewhere between these two extremes. The bias signal cannot itself be recorded because the wavelength of the signal is too small compared to the recording head gap and the magnetic particle size.

Interestingly this property of bias is used in the new Dolby HX system. HX stands for 'Headroom Expansion'. What the system does is to reduce the bias dynamically with the level of high energy frequency components. By reducing the bias when these are present they are recorded faithfully. That is, they are not muted by the inevitable HF compression of the tape.

Finally, a cautionary tale. Several years ago when I was repairing audio equipment, there was a favourite cassette player fault. The poor punter would walk in complaining that his recordings were distorted and weak. One would look sagely into the cassette player and tell him he obviously had a serious fault that required repairing. In fact, nine times out of ten our solution would be to rush for the nearest can of switch cleaner. A quick wipe with a soft cloth impregnated with the cleaner on the heads would eradicate the problem simply caused by tape coating debris!