.

In the second "Hi-Fi Choice on cassette decks", I pointed out that many of the then latest designs were quite considerably better than the average of the first book. After examining all the results of both the subjective and objective tests on the 36 new models reviewed in this third volume, it is quite clear that a similar improvement has taken place and that the most recently designed decks show many areas of improvement over the average better models reviewed in the last volume, some of which are included in this edition.

Quality control in particular would now seem to be better, for very few decks gave other than reasonable Dolby-out responses in ferric and chrome positions, although quite a few were not optimised too well for ferrichrome.

It is particularly interesting that manufacturers are now recommending types of tape that are suitable for use with their decks more openly, although a few still fight shy because of the politics involved, which is most unfortunate.

Facilities provided are very much the same, but logic control has been employed in many new models and a few more manufacturers now incorporate variable bias and/or equalisation user controls (Trio, JVC, Pioneer, Aiwa, Neal and Technics).

Before discussing the more general conclusions, I shall deal with each part of the cassette deck separately, so that general trends in different areas can be established.
.

Comparatively few machines had really adequate microphone input sensitivity, these including the Tandberg models 320 and 340, Sony TC158SD, TCK7/8B, Nakamichi 1000 II, Eumig, B & O 5000 and Uher CR240.

Unfortunately, the compromise of using the same preamplifier for the microphone and DIN inputs makes it virtually impossible to have good microphone input sensitivity whilst avoiding DIN input noise degradation.

Clipping margins are better on the microphone inputs than before in general, but as expected, most models having better than average sensitivity have this at the expense of reducing the clipping point, so that very loud material might not be able to be recorded with sensitive microphones without problems. Decks having poorer than average clipping margins include the two Tandbergs,
Toshiba PC4360, Trio KX1030 and B&O 5000 (very poor).

A few machines enabled a mono microphone to be plugged into one channel whilst putting the output of this onto both channels for recording two-channel mono. Models which provided this feature included Eumig, Nakamichi 1000II, Neal 302, Teac 303 and Toshiba PC5460.

This may be useful for those who don't want to buy two microphones, especially if the microphone input is being used for speech recording for cassette 'letters'. A few models usefully incorporated a switch for altering microphone input sensitivity, and the Uher CR240, Eumig and Sony 158 must be particularly commended for this.

DIN inputs again presented problems on the majority of decks, although the European machines worked comparatively well on these inputs. Some Japanese manufacturers have nevertheless learnt their lesson, for I have criticised the DIN input problem many times over the years and in both previous Hi-Fi Choice books on cassette decks.

So JVC, Toshiba and Nakamichi must be commended for providing DIN inputs with virtually no noise degradation, enabling these to be interconnected with typical DIN receivers, or the DIN sockets on most normal receivers, without noise problems.

Most machines incorporated the correct replay pin muting when the machines were on record, but too many did not, and some HF feedback and crosstalk problems might arise on such models. In one case, line-out phono sockets muted as well and this can be most inconvenient, namely on the Philips 2538.
Distortion tended to vary from excellent to just adequate on the DIN input and comments are made in the reviews where relevant.

Many machines incorporated input switching and this should theoretically allow the line input to feed straight to the record level faders without any hiss problem being encountered.

Too many machines fed the line inputs on to the DIN input via high value series resistors however, which compromised the performance of the phono inputs (gemeint sind die Cinch Buchsen), and comments are made where relevant. Even on the switchable models too much gain follows the record level control in many cases, so if these are used at positions above half-way, hiss will increase on the line input, and this frequently receives comment.
.

When the input gain controls are all at minimum, input amplifier hiss immediately prior to the record Dolby circuits should be at an absolute minimum; a few models did not have the noise floor as low as I would like to have seen it, although this was only a problem if the machines themselves could not accommodate high recording levels to offset the hiss problem.

It will be seen that models having high gain after the record level control will also have excellent clipping margins, since the front-end preamplifier will not have to give as much gain and will thus accommodate a wider input range. This is reasonable up to just below the limits where hiss can just become noticed, but of course it is unreasonable if the clipping is improved at the expense of hiss.
.

Amongst models having particularly bad input noise degradation the Marantz 5010 is frankly the most hissy cassette deck that I have yet examined. In the Eumig Metropolitan CCD, affect each other so that if the DIN input is mixed with a line input for example, the degree by which the levels affect each other is not quite tolerable; in other models though, any slight loading effects were so minimal as to be hardly worth mentioning.

In order for many Japanese models to incorporate a DIN socket, both the noise floor and line input performance has often had to be degraded, so it seems quite reasonable to propose that DIN sockets are just left out of some machines so that optimum performance can be gained from the line and microphone inputs.

Surely this is better than to degrade all inputs for the sake of accommodating a DIN one on a cheaper deck in which the manufacturers feel that they cannot afford to switch sensitivities properly. Akai sensibly omitted a DIN socket on their model GXC 725D.

All machines incorporating Dolby processing have to include mpx filtering by Dolby licence agreement, but quite frequently this is switchable. In the subjective tests we noticed with great interest that when we recorded pink noise at a high level with mpx filtering in, replay was often more toppy at HF and EHF, and thus more correct than when recorded with mpx filter out.

This is because frequencies above the mpx filtering shoulder are causing general tape saturation on record, and if HF and EHF are not recorded and reproduced accurately by the tape, they will be expanded down on replay, and thus become apparently down.

When the mpx filter is switched in, these EHF frequencies will not saturate the tape, and thus on replay HF frequencies will be reproduced more accurately. This was particularly noticeable on the Tandberg model 340A which produced amazing results with filtering switched in on pink noise.

Depending upon the program content, however, we sometimes preferred the filter in and sometimes out, and this also very much depended on the amount of cut produced by the filter at 15kHz. Some filters cut as much as 4dB, whilst others were less than ldB down at 15 kHz, but all reduced the 19kHz response very substantially indeed.
.

Many manufacturers employing a combined DIN/mic amplifier use too little loop gain in the preamplifier, and therefore distortion levels are not reduced sufficiently when feedback is applied.

This will not usually be troublesome if the DIN input is at the normal standard DIN level of 1uA, but frequently users will want to interconnect the DIN socket with equipment which is not designed accurately to DIN source standards itself.

Very few receivers incorporating DIN sockets give other than standardised DIN levels, but it is probable that the DIN standard will, in future, be improved so as to deliver higher source levels, ie 2mV/kohm equivalent to 2uA. This level is some 6dB higher and some machines began to show increased distortion here.

Whilst DIN inputs always had an acceptable clipping margin for normal DIN levels, higher levels often showed increased distortion, despite the clipping margins being comparatively good. It seems to me that mic/DIN input circuitry should be built around 3 rather than 2 transistors to give generally decreased distortion. If DIN sockets are a must for the manufacturer then he should optimise them properly and the best way to do this is to have gain switching in the preamplifier (see technical section).

One might think that a record level control should be effective, and no problems should be experienced. However, we all found machines that used controls many centimetres apart were extremely irritating in use, requiring two hands to fade a signal up or down without image shifting.

Split concentric controls are the obvious answer here, although the Neal solution is a reasonable alternative, employing a ganged stereo control and a separate balance one, but we did not like this economically unfortunately, since the balance control was not centre indented, which could have provided a very rapid centre balance point without visual examination.

Both the Pioneer 1000 and Technics 615 had level controls with many steps,
and this we very much liked economically, although it is presumably more expensive to include; Technics in particular must be commended for incorporating a stepped control on a fairly inexpensive machine.

In some cases earth loops were noted if a stereo microphone having a common earth between left and right channels was plugged into a deck. If only the left or right channels were plugged in, or if two seaprate microphones were used, the earth loop problem vanished and this must be watched. Some machines hummed quite badly on the microphone inputs.
.

A few machines incorporated very inadequate metering, such as the B&O 5000 and Eumig models where it was very difficult to set peak recording levels.

Many machines incorporated conventional meters which were not peak reading types, and which were not complemented by peak reading lights, or had lights which were set to operate at inappropriate levels, and with these it is necessary to learn by experience how much to let the meters read on peaks of different types of program.

Such "VU" meters are normally set so that +3dB corresponds to Dolby level. Since the average programme may under-read by up to 8dB on peaks, the needles should not be allowed to go higher than zero dB or so, which should be equivalent to peaking at approximately 5dB over Dolby level.

Long continuous sounds or highly compressed music, such as some modern pop music or classical church organ music, will under-read much less but speech and sharp percussive sounds should be reading substantially below zero dB but will be intrinsically peaking much higher. Our favourite meters of all were the Sony TCK8B liquid crystal types, but I must also commend the Aiwa 6800, and Technics 631/M85.

Some manufacturers incorporated high frequency boost in the metering circuits, and I do not welcome this since it may scare one into under-recording if using a tape which has excellent HF performance. On the other hand, if the equalisation is not enough you may well over-record on a poorer tape type.

I prefer to know the peak value of the actual signal before processing and equalisation, and judge by experience how far I can go on the tape. It will not take more than three or four trials to establish optimum peak levels, and if these levels in general are kept to, all the tapes should sound equally loud and of similar quality, provided that the source is good. Peak reading lights are useful since they will tell you if you have exceeded a high recordifig level; sometimes it will be found that letting the highest level light flicker very occasionally will be satisfactory.

It is always useful to be able to establish how much level is present on a recorded cassette so that one can tip off ones friends that they are over-recording, etc. Most recorders allow their meters to read at a fixed gain level on playback so that +3dB, for example, represents a continuous signal at Dolby level.

Some machines however connected the metering on replay after the output volume control, so that the metering level was meaningless as far as the tape modulation is concerned. In this situation they would only tell the user how much volume was being sent from the recorder, and this information is only very rarely needed.
.

The hiss levels on many of the modern decks are, frankly, disappointing, and this can be partly due to the replay amplifier being more noisy than it should be, thus contributing extra hiss to that intrinsically produced by the tape itself.

Several recorders employed replay heads with very fine gaps in order to reproduce accurately frequencies that very few of us normal folk can even hear. Most of these fine gap replay heads have lower outputs and thus require more amplification.

Typical examples of these machines are the Nakamichi 1000 II and Technics M85 models. In such cases, replay hiss can be further affected by too much HF being reproduced, such as on the Tandberg 320, and again the Nakamichi 1000 II. Not only will excessive HF on replay produce more hiss but it will also make prerecorded cassettes sound far too toppy and the Dolby B circuits will not track properly in such cases.

Replay amplifier distortion was only rarely a problem with normal tape types, but it should be borne in mind that when and if iron pre-recorded cassettes are marketed, the maximum levels on the tapes are likely to be substantially higher so that clipping margins in addition to amplifier distortion become important.

Only the Hitachi 220 had a poor clipping level, and this would otherwise have been a 'best buy' (the Philips 2538 and Uher 240 were also not too good here).
.

Some Dolby 'B' circuits introduced too little or too much noise reduction and whilst this was sometimes due to the selection of Dolby 'B' ics, which should possibly have been rejected, sometimes it was due to inadequate setting up, or in the case of recorded noise, more probably due to noise entering the record Dolby from the record amplifier.

In general, replay responses were decidedly better on the latest models and only very rarely was the HF clearly down at 10kHz. Replay head azimuth settings though were sometimes very inaccurate, the worst errors being noted on the B&O 5000 (very difficult to correct) and Hitachi D220. If the replay azimuth is out hardly any difference will be noticed on recordings made on the machine, but pre-recorded cassettes or cassettes made by friends (assuming their machines are correct) will reproduce with muffled HF and sometimes swishing transient images.
.

All the machines in this survey are equipped with headphone sockets of one form or another. Almost all machines incorporated stereo jack sockets, but sometimes a manufacturer will use an annoying socket such as the DIN type and a matching plug may be difficult to obtain and even more difficult to understand!

In each review comments are made as to the recommended impedance levels which should work satisfactorily. In most instances the clipping margin on the headphone output is not as good as that on the line output and so it is well worth following the recommendations.

The amount of volume that most people require has been estimated in the laboratory, and in almost all cases the decks would give sufficient volume into one type or another; a few, however, were totally inadequate on any type of headphone. There is no technical reason why 8 ohm, 25 ohm or 600 ohm models should not be used, apart from that of compatibility with the recorder.

Don't forget that if the headphone volume is controlled by the replay volume control, adjusting their level will change the signal sent to the amplifier, which may be difficult if tapes are being copied.
.

Almost all the decks in the survey include line in/out phono sockets and at least one 5-pole DIN socket. Receivers and amplifiers usually include
tape recorder sockets of both phono and DIN types, and unless otherwise stated in the reviews, always stick to using the phono socket interconnections where possible.

DIN interconnections can be useful since all the signal paths are made within one multi-wire cable having a DIN plug at each end, but the imput levels at the deck end are often so low that these inputs can pick up bad radio-frequency interference; if there are any powerful professional or amateur transmitters close by, this will almost certainly cause trouble with the DIN input.

Many machines had satisfactory DIN inputs nevertheless and so they should normally give good performance when interconnected with the DIN socket on the receiver or amplifier. Whilst recording from an input connected to the DIN socket, the replay pins of this socket should be muted, but several recorders did not provide such muting and this can create problems with breakthrough signals.

Sometimes the DIN output is at a fixed level whilst the phono sockets are adjustable, and this can be quite useful. Make sure that each channel is individually screened from the others to prevent bad HF crosstalk if making up any DIN leads.
.